text
stringlengths 66
377k
|
|---|
Inflammation-Associated Microsatellite Alterations Caused by MSH3 Dysfunction Are Prevalent in Ulcerative Colitis and Increase With Neoplastic Advancement
OBJECTIVES:Inflammation-associated microsatellite alterations (also known as elevated microsatellite alterations at selected tetranucleotide repeats [EMAST]) result from IL-6-induced nuclear-to-cytosolic displacement of the DNA mismatch repair (MMR) protein MSH3, allowing frameshifts of dinucleotide or longer microsatellites within DNA. MSH3 also engages homologous recombination to repair double-strand breaks (DSBs), making MSH3 deficiency contributory to both EMAST and DSBs. EMAST is observed in cancers, but given its genesis by cytokines, it may be present in non-neoplastic inflammatory conditions. We examined ulcerative colitis (UC), a preneoplastic condition from prolonged inflammatory duration.METHODS:We assessed 70 UC colons without neoplasia, 5 UC specimens with dysplasia, 14 UC-derived colorectal cancers (CRCs), and 19 early-stage sporadic CRCs for microsatellite instability (MSI) via multiplexed polymerase chain reaction capable of simultaneous detection of MSI-H, MSI-L, and EMAST. We evaluated UC specimens for MSH3 expression via immunohistochemistry.RESULTS:UC, UC with dysplasia, and UC-derived CRCs demonstrated dinucleotide or longer microsatellite frameshifts, with UC showing coincident reduction of nuclear MSH3 expression. No UC specimen, with or without neoplasia, demonstrated mononucleotide frameshifts. EMAST frequency was higher in UCderived CRCs than UC (71.4% vs 31.4%, P 5 0.0045) and higher than early-stage sporadic CRCs (66.7% vs 26.3%, P 5 0.0426). EMAST frequency was higher with UC duration >8 years compared with £8 years (40% vs 16%, P 5 0.0459).DISCUSSION:Inflammation-associated microsatellite alterations/EMAST are prevalent in UC and signify genomic mutations in the absence of neoplasia. Duration of disease and advancement to neoplasia increases frequency of EMAST. MSH3 dysfunction is a potential contributory pathway toward neoplasia in UC that could be targeted by therapeutic intervention. SUPPLEMENTARY MATERIAL accompanies this paper at
# Introduction
Defects in DNA mismatch repair (MMR) function are associated with human cancer development and progression. Germline mutations in the MMR genes MSH2, MLH1, PMS2, and MSH6 are causal for Lynch syndrome, an inherited condition in which patients may develop colorectal cancer (CRC) (constituting ;3% of all CRCs) and cancers of the rest of the gastrointestinal tract, female reproductive and urinary tracts, and specific skin and CNS tumors. Somatic inactivation of MLH1 via biallelic hypermethylation of MLH1's promoter is observed in ;15% of sporadic CRCs, and biallelic somatic mutations of MSH2, MLH1, PMS2, or MSH6 are the apparent cause for Lynch-like syndrome in 1%-2% of all patients with CRC. The common finding in these patients with germline or somatic MMR deficiency is the presence of microsatellite instability-high (MSI-H) in tumor DNA, defined by an NCI Consensus Panel as at least 2 frameshift mutations identified with the use of a panel of at least 5 mono-and dinucleotide microsatellite markers. In general, the outcome of patients with an MSI-H tumor is better than a patient with a microsatellite stable (MSS) tumor, and patients with MSI-H tumors can benefit with further increased survival through the use of immune checkpoint inhibitors due to hypermutated cancer genomes that drive immune-responsive neoantigens generated from exon coding mononucleotide microsatellite frameshifts. Another form of MSI is termed elevated microsatellite alterations at selected tetranucleotide repeats (EMAST) that has been observed in a variety of cancers including CRCs. EMAST (in the absence of MSI-H) is caused by an IL-6-induced nuclear-to-cytosol shift of the MMR protein MSH3, triggering subsequent di-, tri-, and tetranucleotide and longer frameshifts of genomic microsatellites in the nucleus. EMAST is observed in ;50% of all CRCs. Patients with EMAST CRC, unlike patients with MSI-H CRC, demonstrate poor survival compared with patients without EMAST CRC and have advanced-stage disease and frequent metastasis. Because the mechanism of MSH3 dysfunction is due to intracellular displacement of MSH3 by proinflammatory cytokine signaling, EMAST can also be termed inflammation-associated microsatellite alterations. This term differentiates the isolated MSH3 dysfunction observed in inflammation-associated microsatellite alterations from secondary MSH3 mutations as a result of MLH1 deficiency in sporadic CRCs, a scenario where a tumor would manifest mononucleotide frameshifts in addition to di-, tri-, and tetranucleotide instability. In particular, tumors defective for MLH1, MSH2, or PMS2 would demonstrate mono-, di-, tri-, and tetranucleotide frameshifts, whereas tumors defective for MSH6 would manifest mostly mono-and some dinucleotide frameshifts, and tumors with isolated MSH3 dysfunction would demonstrate di-, tri-, and tetranucleotide instability and no mononucleotide frameshifts. With MSH3 deficiency, affected cells will experience a defect in MMR function and increased DNA double-strand breaks (DSBs) that can lead to aneuploidy and loss of heterozygosity (LOH) events, generating a complex DNA repair deficit. One or both of these DNA repair deficits may contribute to the pathogenesis of tumors, including metastases.
The observations of MSI-H and EMAST have largely been among cancers. Noncancer tissues, presumably due to intact MMR function and stable normal genomes, generally do not manifest MMR deficiency. Rarely have MMR defects been found in noncancer tissue. We have previously observed evidence for MSH3 dysfunction within hamartomatous polyp epithelium, whose polyp is a non-neoplastic lesion that possesses an expanded inflammatory lamina propria. With this last observation coupled with the recent characterized mechanism of proinflammatory cytokine-induced MSH3 intracellular displacement, we sought out any evidence of MSH3 dysfunction in ulcerative colitis (UC), an inflammatory bowel disease condition in which tissues contain several proinflammatory cytokines including IL-6. UC can progress to CRC in some patients, with the greatest risk factor being disease duration over 8 years, but also include the extent of UC, presence of primary sclerosing cholangitis, family history of sporadic CRC, severity of bowel inflammation, and young age at onset for UC. Previous reports have examined MSI using mono-and dinucleotide markers on UC specimens with varied results, but generally conclude that UC and UC-associated neoplasia are not commonly an MSI-H-driven process. Some reports show that UC specimens lack mononucleotide instability but occasionally possess dinucleotide instability, which raises the possibility of isolated MSH3 dysfunction. Several reports have hypothesized because there is little evidence for loss of MMR protein expression in UC that the repeated bombardment of tissue by inflammation might alter MMR function in the absence of mutation or loss of protein. Here, we examined whether our previous discovery of inflammation-associated microsatellite alterations and MSH3 dysfunction among CRCs is operative in non-neoplastic but inflammatory UC. Such finding would extend the role of inflammation as a continuing cause of DNA damage before any neoplastic transformation.
# Methods
## Cell lines
The human colon cancer cell lines HCT1161315 (proficient in MSH3), G5 (deficient in MSH3), and DLD1 (deficient in MSH6) have been described previously. HCT1161315 and DLD1 were grown in Dulbecco Modified Eagle Medium supplemented with 10% fetal bovine serum. G5 cells were maintained in Dulbecco Modified Eagle Medium with 10% fetal bovine serum and 0.6 mg/mL of puromycin.
## Clinical samples
A total of 70 human UC clinical tissue samples, 14 UC-derived CRC tissue samples, and 5 UC with dysplasia tissue samples were collected from Osaka University Hospital and Mie University Hospital. Fifty-six of the 70 patients with UC had adequate demographic clinical information, such as age, sex, disease duration, and types of UC (pancolitis or left-sided colitis) and limited data on patient therapy usage (some data on steroid and 5-ASA usage and little or no data on use of azathioprine, 6-mercaptopurine, and biologic therapies such as anti-TNF-a, vedolizumab, and tofacitinib). Although we do not have the presurgery Mayo Disease Activity Index scores for the 70 patients with UC, all patients had disease severe enough to warrant surgery (colectomy) based on clinical practice because this group did not have dysplasia or cancer. Nineteen early-stage CRC (AJCC stage 0 and I) samples were collected from Osaka National Hospital. Ten normal human colon tissues were collected from Osaka University Hospital as controls. These patients did not have UC and had partial resections for reasons such as obstruction, volvulus, and diverticular disease. All institutions had IRB approval to conduct this study. All UC-related samples were from archived formalin-fixed, paraffin-embedded specimens taken from the colectomies that were performed.
## Dna extraction
All epithelial tissues were microdissected from paraffin-embedded 10-mm tissue sections. DNA was isolated and purified from the microdissected tissues as previously described. DNA from noninflamed submucosa and/or serosa tissues were used as controls.
## Multiplex polymerase chain reaction and determination of msi and emast
We established a new multiplex polymerase chain reaction (PCR) combining Supplementary Digital Content 1, http://links.lww.com/ CTG/A129. We used 2 mononucleotide (BAT25 and BAT26), 5 dinucleotide (D2S123, D5S346, D17S250, D18S64, and D18S69), and 7 tetranucleotide microsatellite sequences (D9S242, D20S82, D20S85, D19S394, D8S321, MYCL1, and RBM47) as previously described. We defined EMAST as at least 1 tetranucleotide marker with frameshift without any frameshift of a mononucleotide marker and combined this group with MSI-L samples due to both MSI-L and EMAST being caused by MSH3 dysfunction.
# Western blot analysis
Whole-cell protein was extracted from cell lines and separated by sodium dodecylsulfate-polyacrylamide gel electrophoresis gel. The separated proteins were transferred to nitrocellulose membranes. Anti-human MSH6 mouse monoclonal antibody (clone 44, Cat No. 610919, dilution: 1:500; BD Biosciences, San Jose, CA), anti-human MSH3 mouse monoclonal antibody (clone 52, Cat No. 611390, dilution: 1:500; BD Biosciences), anti-human MSH2 mouse monoclonal antibody (clone G219-1129, Cat No. 556349, dilution: 1:200; BD Biosciences), anti-human MLH1 mouse monoclonal antibody (clone G168-728, Cat No. 554073, dilution: 1:500; BD Biosciences), anti-human PMS2 mouse monoclonal antibody (clone A16-4, Cat No. 556415, dilution: 1:500; BD Biosciences), and anti-a-tubulin antibody (dilution 1:8,000; Sigma-Aldrich, St. Louis, MO) were used as primary antibodies for the detection of specific proteins. Goat anti-mouse antibody (dilution: 1:1,000; Cell Signaling Technology, Danvers, MA) was used as a secondary antibody. Signals were detected using SuperSignal West Pico (Thermo Scientific, Waltham, MA) and captured by an ImageQuant LAS4000 system (GE Healthcare, Chicago, IL).
## Immunohistochemistry
Immunohistochemical staining was performed using 5-mm-thick paraffin-embedded sections on the 10 normal colons collected and 10 randomly selected UC colons out of 70 collected. After deparaffinization in xylene and dehydration in graded ethanol solutions, tissue sections were heated at 121°C for 15 minutes in TE9 buffer (pH 9.0). Slides were incubated with 3% hydrogen peroxide for 20 minutes to quench endogenous peroxide activity. Samples were washed with PBS and then incubated with immunohistochemistry (IHC) blocking diluent (Leica Biosystems, Wetzlar, Germany) for 20 min, followed by incubation with primary antibody at 4°C overnight. After washing with PBS, the sections were incubated with secondary antibody for 30 minutes. Samples were developed using 3,3-diaminobenzidine tetrahydrochloride and contrasted using hematoxylin. Anti-human MSH3 rabbit monoclonal antibody (clone EPR4334 (2), Cat No. ab111107, dilution: 1:1,000; Abcam, Cambridge, MA), antihuman MSH6 rabbit monoclonal antibody (clone EPR3945, Cat No. GTX62383, dilution: 1:200; GeneTex, Irvine, CA), and antihuman IL-6 mouse monoclonal antibody (Cat No. ab9324, dilution: 0.25 mg/mL; Abcam) were used as primary antibodies.
# Statistical analysis
The StatView 5.0 program (Abacus Concepts, Piscataway, NJ) was used for statistical analysis, Student t test, Fisher exact test, Mann-Whitney U test, and 2 3 2 contingency table analyses. P , 0.05 indicated a significant difference.
# Results
Multiplexed PCR accurately determines MSI and EMAST status and confirms MSH3 deficiency as cause of EMAST We set out to create a multiplexed PCR method for detecting all forms of MSI, capable of determining MSI-H, MSI-L, EMAST, and MSS efficiently and simultaneously. Furthermore, we wanted to confirm previous observations that deficiency in MSH3 coincides and causes EMAST. We developed primers to successfully amplify 2 mononucleotide, 5 dinucleotide, and 7 tetranucleotide microsatellite markers that have been individually previously used for detection of MSI and EMAST (see Supplementary Digital Content 1, http://links.lww.com/CTG/ A129) in a multiplexed fashion using 4 groups of PCR reactions followed by fragment analyses (see Supplementary Digital Content 2, http://links.lww.com/CTG/A130, which shows example fragment chromatograms of the 14 markers). We then used cell lines in which the MSH3 and MSH6 status was known to assess microsatellite frameshifts in the presence or absence of those DNA MMR genes. Cell lines were cloned, and each isolated cloned colony was subject to the multiplex PCR, enabling a purity that is often absent from the heterogeneity of a tumor. HCT1161315 cells are derived from MLH1-and MSH3-deficient parental HCT116 CRC cells where 1 copy of human chromosome 3 (correcting MLH1 deficiency) and 1 copy of human chromosome 5 (correcting MSH3 deficiency) have been transferred. Western blots confirm expression of MLH1 and MSH3 in this cell line. The cell line G5 is derived from HCT1161315 cells by transfection of a tetracycline-regulated retroviral vector that encodes shRNA against MSH3. Western blots confirm acquired absence of MSH3 expression in G5 cells. DLD1 CRC cells are known to be MSH6 deficient, and Western blots confirm the absence of MSH6 expression. Fragment analyses of cell clones demonstrate patterns of MSI based on the presence or absence of MSH3 and MSH6. Fully MMRproficient HCT1161315 cells demonstrated no mononucleotide frameshifts, no dinucleotide framseshifts, and 1 tetranucleotide frameshift among 96 cell clones. In contrast, MSH3deficient G5 cell clones show no mononucleotide frameshifts but possess excessive dinucleotide and tetranucleotide frameshifts. MSH6-deficient DLD1 cell clones demonstrate excessive mononucleotide frameshifts with 1 dinucleotide frameshift and no tetranucleotide frameshifts. Our data confirm that MSH3 deficiency causes dinucleotide or greater MSI (EMAST and MSI-L), whereas MSH6 deficiency causes dinucleotide or lesser MSI (MSI-H). Our multiplexed PCR approach is accurate in simultaneously detecting frameshifts at mono-, di-, and tetranucleotide microsatellite markers to allow determination of the MSI and EMAST status of human DNA.
## Uc tissues demonstrate reduced msh3 expression
We have previously shown in CRCs and CRC cells that IL-6 signaling triggers a nuclear-to-cytosol displacement of MSH3. This mechanism seems to coincide with loss of nuclear MSH3 expression in CRCs. Although this phenomenon has largely been seen among cancers that have surrounding inflammation as a source of IL-6, we have previously reported absence of MSH3 expression within the epithelium of hamartomatous polyps, which contain inflammation but are not neoplastic. We wondered if the elevated neoplastic risk condition of UC, a type of inflammatory bowel disease whose affected tissues exhibit several proinflammatory cytokines including IL-6, could harbor loss of MSH3 expression similar to our previous findings in CRCs, perhaps as 1 mechanism that could cause DNA damage and elevate neoplastic risk. We performed IHC on 10 randomly selected samples from our 70 resected UC colon specimens and 10 normal colonic tissues for expression of MSH3, IL6, and MSH6. For MSH3 and MSH6 expression, 1000 cells were counted in 10 separate 3100 microscope fields, yielding total counts of approximately 10,000 cells for which to assess nuclear expression. IHC demonstrated uniform nuclear MSH3 expression in normal colonsthat was reduced and/or heterogeneous in UC colons. In comparing nuclear MSH3 expression among epithelial cells between UC and normal colons, UC colons showed significant loss of nuclear MSH3 expression in over 5% of epithelial cells (typically in apical portions of crypts) compared with 0.1% for normal colons. Eight of 10 UC colons showed elevated tissue and lamina propria IL-6 expression compared with 2 of 10 or normal colons. The elevated IL-6 expression among UC colons is consistent with our previous data in CRCs on this being coincident with reduced MSH3 expression. As a control, we also performed IHC for MSH6 expression, which does not undergo nuclear-to-cytosol trafficking by IL-6 signaling. As shown in-i, there is no difference in MSH6 expression between normal and UC colon epithelium. Our data demonstrate elevated expression of IL-6 among UC colons that is coincident with reduced MSH3 epithelial expression.
## Uc and uc-associated neoplasia demonstrate inflammationassociated microsatellite alterations (emast)
With the observation of elevated IL-6 expression that is coupled with reduced MSH3 expression in UC samples, we set out to determine whether the biomarker for absent MSH3 function, EMAST, is present in UC tissues. Such a finding would demonstrate that DNA damage via frameshift mutation would be present in UC, a novel finding.
Because of previous findings that the mechanism for MSH3 displacement is due to proinflammatory IL-6 signaling, EMAST that is observed because of isolated MSH3 dysfunction can be termed inflammation-associated microsatellite alterations. This term sets it apart from other causes of MSH3 dysfunction such as secondary frameshift mutation of MSH3 as a result of MLH1 hypermethylation (and would demonstrate both EMAST and MSI-H in the tissue). We used our multiplexed PCR method on 70 resected UC colons that had no evidence of neoplasia. Among the 70 cases, none showed frameshift mutations in mononucleotide microsatellite markers, 5 cases (7.1%) showed dinucleotide MSI, and 22 cases (31.4%) demonstrated tetranucleotide frameshifts, all consistent with our observed reduced MSH3 expression. This is in contrast to no microsatellite frameshifts among mono-, di-, or tetranucleotide markers for normal colon tissue. Thus, even in the absence of neoplasia, UC tissues demonstrate ongoing DNA damage and mutations that are likely due to the presence of proinflammatory cytokines.
We further examined 5 UC colons with evidence of dysplasia, an important finding and decision point in the clinical care approach from surveillance of patients with UC, as dysplasia is a precancerous condition. Using our multiplexed PCR method, 3 of 5 UC colons with dysplasia (60%) demonstrated tetranucleotide frameshifts, with no mononucleotide or dinucleotide instability. This again is consistent with isolated MSH3 dysfunction.
We wanted to compare UC-derived CRCs with UC without neoplasia and to early-stage sporadic CRCs for inflammationassociated microsatellite alterations. To this end, we performed our multiplexed PCR on 14 resected UC-derived CRCs and demonstrated no mononucleotide instability among the samples, 5 (35.7%) samples with dinucleotide frameshifts, and 10 (71.4%) samples with tetranucleotide frameshifts. Again, this is consistent with isolated MSH3 dysfunction, and the frequency of EMAST among UC-derived CRCs was significantly higher than the frequency among UC samples without neoplasia (P 5 0.0045), suggesting progressive accumulation of frameshifts with advancing tumorigenesis. The high frequency of EMAST among UC-derived CRCs is comparable to multiple previous reports of sporadic CRCs, which demonstrate EMAST among 50% of CRCs. However, our previous report showed that the frequency of EMAST was dependent on CRC stage, with 62% of EMAST CRCs in stage III/IV and 37% of EMAST CRCs as stage I/II. UC-derived CRCs are usually diagnosed at early stage because of constant clinical surveillance. Indeed, 9 (64%) of our UC-derived CRCs in our cohort were stage 0-I; however, both early-stage (6/9, 66.7%) and later-stage (4/5, 80%) UC-derived CRCs demonstrated EMAST, suggesting that its development likely occurs before advancement of stage. In comparing 9 early-stage UC-derived CRCs with 19 early-stage sporadic CRCs, EMAST was significantly more frequent in the UCderived CRCs (66.7% vs 26.3%, P 5 0.0426). Interestingly, among the early-stage sporadic CRCs, we identified 1 mononucleotide frameshift for which its presence signifies at least 1 MSI-H tumor that is likely due to hypermethylation of MLH1 typically seen in up to 15% of sporadic CRCs. We observed no mononucleotide frameshifts in any of the UC tissues without neoplasia, UC tissues with dysplasia, or UC-derived CRCs.
Inflammation-associated microsatellite alterations (EMAST) are more common with long-standing UC UC is an inflammatory condition for which neoplasia risk increases over time, particularly after more than 8 years from initial diagnosis. We wanted to compare the frequency of EMAST between those with short-duration UC and those with long-standing UC (LSUC). We had clinical demographic information on 56 of our 70 samples of patients with UC, and within this cohort, the median disease duration was 15 years (interquartile range 11-19 years). We compared samples of patients with LSUC that had a disease duration of .8 years from diagnosis with samples of patients without LSUC that had a disease duration of #8 years. As shown in, patients with LSUC had an average disease duration of 13 years, whereas patients without LSUC had an average duration of UC of 2 years (P , 0.0001). Patients with LSUC were significantly older (average 42 vs 33 years, P 5 0.0094), but showed no difference in sex composition or extent of colonic disease (left-sided vs pancolitis). Assessing EMAST via our multiplexed PCR, patients with LSUC had significantly higher frequency of EMAST than patients without LSUC (10/25, 40% vs 5/31, 16.1%, P 5 0.0459). These data suggest progressive accumulation of frameshifts from MSH3 dysfunction with longer duration of UC.
Between 5% and 10% of microsatellite unstable DNA needs to be present to reliably detect microsatellite frameshifts for application in UC surveillance biopsies Patients with UC are surveyed for dysplasia and/or neoplasia by colonoscopy and protocol-based biopsies from throughout the colon after 8 years of disease duration. To examine the potential clinical utility of detecting EMAST from UC biopsy material, we (i) determined the approximate mix of frameshifted DNA within non-frameshifted DNA to reliably detect the frameshifts and (ii) assessed the amount of DNA from biopsy forceps taken from resected UC specimens and whether frameshifts could be detected in that amount. We mixed increasing amounts of DNA from G5 cells, which are MSH3 deficient, in a background of HCT1161315 cells, which are MMR proficient (see data from. Because these 2 cell lines are isogenic, polymorphic allele lengths were generally identical for comparison. As shown infor D8S321, a tetranucleotide microsatellite marker, detection of a frameshifted allele could be consistently detected reliably when the amount of frameshifted (or unstable) DNA approached 10% admixture. We measured the DNA amount from standard biopsy forceps taken from 14 biopsies of resected UC specimens. The median amount of DNA obtained from the biopsy forceps was 470.25 ng (interquartile range 232.1-637.5 ng) (see Supplementary Digital Content 3, http://links.lww.com/CTG/ A131, which shows individual sample DNA amounts and if sample was detected by fragment analysis). With utilization of our multiplexed PCR method, all samples demonstrated detectable yields at fragment analysis for all mono-, di-, and tetranucleotide markers (see Supplementary Digital Content 3, http://links.lww.com/CTG/A131). Our data indicate that the use of surveillance biopsies for LSUC is feasible for detection of EMAST, with estimation that, at a minimum, between 5% and 10% of endogenous frameshifted DNA is admixed in the sample.
# Discussion
In this study, we determined whether inflammation-associated microsatellite alterations/EMAST are present in the nonneoplastic and proinflammatory cytokine-enriched condition of UC. We demonstrate that (i) the UC epithelium contains cells with reduced nuclear expression of MSH3 (but not MSH6) in the setting of increased IL-6 expression, (ii) UC specimens exhibit inflammation-associated microsatellite alterations/EMAST consistent with MSH3 dysfunction, (iii) inflammation-associated microsatellite alterations/EMAST frequency increases from UC without neoplasia to UC with dysplasia and UC-derived CRCs, (iv) UC-derived CRCs display higher frequency of EMAST compared with early-stage sporadic CRCs, and (v) inflammation-associated microsatellite alterations/EMAST are more frequent within specimens from patients with LSUC (.8 years of duration) than patients with shorter duration. These novel findings signify strong evidence that ongoing DNA mutations occur in the non-neoplastic UC epithelium as a result of inflammation. A limitation of our study is the low number of available UC cases with dysplasia prohibiting differentiation analysis between low-grade and high-grade dysplasia. However, we might expect that EMAST is present in both forms of dysplasia because it is detected in UC samples without dysplasia.
EMAST is a biomarker for MSH3 dysfunction; it does not necessary mean that EMAST itself is causal for progression of neoplasia. In 1 study comparing primary CRC and their matched metastases, the presence of dinucleotide and longer frameshifts was not strikingly different between the primary and its metastasis; however, there were stark differences in LOH events, with LOH at specific sites enriched in metastases. This raises the possibility that MSH3's role in repair of DSBs may be the dominant factor in the pathogenesis of metastasis rather than its role in MMR. This hypothesis has not been fully tested for CRCs. In this study, we evaluated whether EMAST was present and operative in UC. We did not examine the presence or consequences of DSBs in UC. Aneuploidy has been observed in UC-derived CRCs and UC dysplastic lesions, and in 1 report, aneuploidy was found in the adjacent nonmalignant mucosa surrounding UC-derived CRCs. It is tempting to speculate that these observations are a direct result of MSH3 dysfunction, particularly from defective DSB repair. Our findings in this study at least suggest that this is a possibility.
Because EMAST is generated under the influence of proinflammatory cytokines and inflammation, reducing the inflammation would be predictive of reducing the frequency of EMAST (and potentially the future consequence of neoplasia). Epidemiological evidence suggests that prevention of flares and inflammation reduces the risk of future neoplasia in UC. In relation to MSI, previous studies only examined mono-and dinucleotide frameshifts, with most studies observing frameshfts at dinucleotide repeats, consistent with potential MSH3 dysfunction rather than any other MMR protein dysfunction. Although the use of 5-aminosalicylic acid (5-ASA), an anti-inflammatory compound used for mild to moderate UC, failed to reduce the frequency of detected dinucleotide frameshifts after 1 year of usage in patients with UC (56), 5-ASA did reduce frameshifts at mono-, di-, and tetranucleotide microsatellites in vitro including reducing the mutant frequency of TGFBR2 and ACVR2. Additional studies using biologics might afford an opportunity to evaluate any change in EMAST frequency in UC.
In this study, we ascertained a minimal frameshift mutant admixture for detection of EMAST. We believe that this is important as (i) it will be more reliable than MSH3 IHC in which only 5% of cells show loss of nuclear expression, (ii) most samples from UC colons are taken by biopsy, not resection, and (iii) EMAST could be a potential biomarker for UC and/or UC neoplastic progression in the future. We identified using cell lines that a threshold of between 5% and 10% of microsatellite frameshifted DNA will be needed to reliable detection. Biopsy samples of resected UC colons had adequate amounts of DNA for microsatellite analysis, and all 14 markers demonstrated detectable yields. These data indicate that surveillance biopsies of UC colons could be used for EMAST detection. Potential roles for EMAST in UC management might include (i) as an adjunct in the decision making for more frequent surveillance or decision to surgery with other parameters and (ii) as a biomarker for resolution of inflammation or effectiveness of treatment. We conclude that the biomarker EMAST is present in nonneoplastic UC colons and represents the biochemical failure by nuclear MSH3 to repair dinucleotide and longer microsatellite sequences. Additional MSH3 DNA repair roles may be abrogated in UC, but we did not test for them in this study. We acknowledge that a limitation of this study is that all of our samples were postsurgery samples, meaning that the UC patient inflammation was severe enough to require surgery. A prospective analysis using biopsies from patients with UC may be needed to more closely approximate typical UC clinical care and surveillance. . The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Potential competing interests: None to report.
## Conflicts of interest
|
Institutionalising health technology assessment: establishing the Medical Technology Assessment Board in India
# Abstract
India is at crossroads with a commitment by the government to universal health coverage (UHC), driving efficiency and tackling waste across the public healthcare sector. Health technology assessment (HTA) is an important policy reform that can assist policy-makers to tackle inequities and inefficiencies by improving the way in which health resources are allocated towards cost-effective, appropriate and feasible interventions. The equitable and efficient distribution of health budget resources, as well as timely uptake of good value technologies, are critical to strengthen the Indian healthcare system. The government of India is set to establish a Medical Technology Assessment Board to evaluate existing and new health technologies in India, assist choices between comparable technologies for adoption by the healthcare system and improve the way in which priorities for health are set. This initiative aims to introduce a more transparent, inclusive, fair and evidence-based process by which decisions regarding the allocation of health resources are made in India towards the ultimate goal of UHC. In this analysis article, we report on plans and progress of the government of India for the institutionalisation of HTA in the country. Where India is home to one-sixth of the global population, improving the health services that the population receives will have a resounding impact not only for India but also for global health.
# Introduction
India has experienced steady economic growth over the last decade, 1 with a consistently growing burden of chronic health problems in line with its growing capital wealth. [bib_ref] Assuring health coverage for all in India, Patel [/bib_ref] [bib_ref] Next generation maternal health: external shocks and health-system innovations, Kruk [/bib_ref] [bib_ref] Towards achievement of universal health care in India by 2020: a call..., Reddy [/bib_ref] The rising cost of healthcare has not been matched by a corresponding increase in the healthcare budget, which at 1.3% of gross domestic product (GDP), is one of the lowest in the world.With the goal of achieving universal health coverage (UHC) and the sustainable development goals in sight, 2 4 6 decisions on investing in cost-effective health technologies will be critical in improving population health and health equity in India.
Health technology assessment (HTA) provides a globally accepted and structured approach to synthesising evidence for cost and clinical effectiveness alongside ethical and equity considerations to form the basis for evidence-based priority setting and policy decisions. 7 8 HTA is currently not a formal component of healthcare decision-making in India. Decisions for allocation of health resources at both the national and state level are predominantly made on the basis of consensus opinion from expert committees. However, explicit priority setting is in line with increasing efforts of the Ministry of Health and Family Welfare (MHFW) to strengthen evidence-based service delivery through initiatives such as the institutionalisation of the National Health Accounts 9 and the development of evidence-based standard treatment guidelines.Discussion of HTA in India has begun to move beyond academia into official
## Institutionalising health technology assessment: establishing the medical technology assessment board in india
Laura E Downey, 1 Abha Mehndiratta, 1 Ashoo Grover, 2 Vijay Gauba, [bib_ref] Next generation maternal health: external shocks and health-system innovations, Kruk [/bib_ref] ► India is committed to universal health coverage (UHC) for its 1.25 billion population. ► Health technology assessment (HTA) is an important tool to make the priority setting process for allocating health budgets towards UHC more evidence based, fair and equitable. ► The Department of Health Research, Government of India, is taking active measures to formalise a system of HTA in India to form the basis of health policy decisions in the country. Activities to this end are underway and described for the first time in this paper. ► Formalising a system of HTA in India is a promising step in promoting the optimal utilisation of existing resources to ensure that the greatest amount of health is bought for every rupee spent in India and extending adequate healthcare services to the population towards the goal of UHC.
## Bmj global health
government policy, such as the 12th Five Year Plan 11 and recently approved National Health Policy,marking an important shift in the government's commitment towards a more effective resource allocation for health.
To further the HTA agenda in India and implement a formal evidence-based process by which health policy decisions are made in the country, the government of India has assigned the task of establishing the Medical Technology Assessment Board (MTAB) to The Department of Health Research (DHR), a medical research department within the MHFW. The MTAB will use HTA as a tool for decision-making regarding the allocation of national health resources. Institutionalising HTA in India represents a landmark shift in the way health resource allocation decisions are made in the country, requiring the buy in and cooperation of central, state and local decision makers and from the professional and paraprofessional medical community to maximise effective implementation.
In this paper, we highlight how HTA can assist the Indian government through the MTAB in effective priority setting for UHC; outline a roadmap for effectively embedding HTA in health decision-making and discuss current progress and activities underway, towards achieving this ambitious health sector reform through establishing the MTAB.
## The establishment of mtab in india
A roadmap for establishing the MTAB and institutionalising HTA has been codesigned by the Government of India, led by the DHR, with technical assistance from national and international research and policy organisations. This roadmap is broadly driven by the core principles laid out in the recent WHO West Pacific Region policy brief 'Factors Conducive to HTA development in Asia.The policy brief is the first of its kind to employ a detailed document review, informant interviews, focus group discussions and a survey questionnaire to identify the characteristics of and conducive factors to setting up successful national HTA agencies in the Asia Pacific Region. The study was conducted by a team of scholars from six countries in the Asia Pacific region (China, Indonesia, the Republic of Korea, Malaysia, Thailand and Vietnam), and data informants represented national government, academic institutions, non-governmental organisations and policy think-tanks. The evidence was synthesised and a thematic analysis was conducted to identify common principles that contribute towards successful HTA institutionalisation across Asia. We will outline here each overarching principle for the successful establishment of HTA in an Asian context highlighted by the WHO policy brief, and the corresponding plans for implementation in India through establishment of the MTAB.
## Build human resources and national capacity for hta
The DHR has recently undertaken a national capacity gap analysis, which was sent to over 50 institutes across country to facilitate understanding of both current capacity to undertake health economic analysis in the country and gauge interest in contributing to the MTAB programme of work. From the results of this survey, memorandums of understanding have been signed between DHR and centres of academic excellence across country to collaborate in this area. A recently published systematic review has highlighted a paucity of direct cost-effectiveness analyses being undertaken and published in India 14 ; however, important HTA-related research across the country, including epidemiological, service use and disease burden studies, could contribute as a necessary starting point towards this effort. Capacity building through in-country delivery of intensive health economic training courses, planned to take place in 2017 and 2018, will also provide a necessary basis from which to build local skill and expertise in this area to drive forward the HTA agenda. Future collaboration may be in the form of undertaking analyses, implementation and policy support, providing training or building modules on evidenced-based medicine and priority setting into medical and allied health education programmes. establish a core htA team or agency DHR has taken active measures towards hiring an MTAB secretariat as a dedicated workforce to oversee the running of this work programme. This focal HTA agency will be essential to drive the HTA work forward through developing key policy, process and methodology documentation and planning and coordinating all MTAB-related activities.
## Link hta to policy decision-making mechanisms and processes
The mandate to formally adopt HTA as part of Indian healthcare decision-making, specified in the 12th Five Year Plan 11 and National Health Policy, 12 is a positive start to establishing legitimacy and credibility of the MTAB. Initial buy in from policy-makers has been strong, best evidenced by public support of the initiative by both secretaries of state.Topic nomination and prioritisation will be conducted in consultation with state secretaries of health and the central MHFW to ensure that all future HTA analyses are policy-relevant and user-driven to facilitate immediate uptake into the health system. The way in which recommendations made by the MTAB will be implemented is yet to be fully determined, however, a number of options are presently being explored. The National Health Mission provides a unique portal between the centre and the states through national grants in aid,where guidance on cost-effective intervention and service delivery guidance could reduce waste and improve allocative efficiency of health spending. The same applies to the national health insurance scheme, Rashtriya Swasthya Bima Yojna, where the MHFW currently lacks evidence to support decision-making regarding cost-effective investment. [bib_ref] The need for better evidence to evaluate the health & economic benefits..., Nandi [/bib_ref] BMJ Global Health transparency of htA agency and processes MTAB will need to consistently demonstrate its value to both the government and the stakeholder community. Plans for undertaking multiple pilot HTA analyses are underway, where the methods, processes, data availability, technical capacity and evidence to policy continuum will be tested and documented. A conflict of interest policy will be developed to ensure transparency of any pecuniary or non-pecuniary interests of all involved in the analysis and decision-making, based on international best practice standards. [bib_ref] Historical development of health technology assessment in Thailand, Teerawattananon [/bib_ref] To gain buy in from the stakeholder community, the MTAB must adhere to a standard of full transparency, allowing the public an open window into the process manuals through guiding HTA development, the reference case for economic evaluation, the data and models considered by the MTAB and documentation outlining and potential conflicts of interest and how decisions were ultimately reached. This process of open participation, high technical rigour, and transparency of methods, data, and decision-making, will bring legitimacy and credibility to HTA in India and the MTAB recommendations.
## Take advantage of international collaboration during formative stage of development
The partnership between the government of India and the International Decision Support Initiative, a global network of leading government institutes, universities and think tanks which supports policy-makers in priority setting for UHC, has enabled a demand-driven and contextually appropriate national plan for institutionalising HTA in India, based on a synergy between strong local contextual knowledge and strong international evidence and experience. Successful implementation of HTA in Indian health policy will undoubtedly require an iterative trial and error process, where drawing from the experience of more developed international HTA agenices may also allow potential challenges to be recognised and prospectively addressed. In January 2017, a DHR-led group participated in a study tour to Thailand to understand how HTA has contributed to the country's achievement of UHC, 20 providing a promising platform for knowledge exchange and learning. Sustainability requires local ownership, and all parties involved in the early stages of this MTAB programme are committed to ensuring that this process is locally driven, owned and carried forward. the promIsIng future of htA In IndIA In a health system burdened by rising healthcare costs and inordinately high out of pocket expenditure, an established and legitimate HTA mechanism in India holds a promising future. The MTAB could provide a platform from which healthcare payers could engage in more strategic purchasing of health services, and offers a crucial mechanism for defining evidence-based criteria for the purchase of services through direct procurement, contracting or health insurance. Further, by specifying appropriate population criteria for which a given HTA recommendation applies, and linking provider payment to fulfilment of these criteria, the government is also afforded a stronger mechanism to enforce regulatory control on service provision and reduce inappropriate delivery of care. [bib_ref] Using health technology assessment for informing coverage decisions in Thailand, Mohara [/bib_ref] Such a mechanism not only reduces inappropriate practices and avoids long-term sequelae of inappropriate intervention but also reduces waste of financial resources and the associated opportunity cost.
Adequate public expenditure on health is recognised as an important contributor to the success of embedding HTA in health policy.While the underlying assumption of HTA is to spend money better, the authors recognise that public health spending in India remains low and can only stretch so far. The recently approved National Health Policy has endorsed the need to raise the level of public spending on health in India to upwards of 2% of GDP.The existence of an HTA mechanism would allow for well-planned and systematic deployment of newly allocated resources and reduction of potential waste should the government raise public health expenditure in the future.
The complex and fragmented Indian health system architecture poses the most significant challenge to the successful translation of MTAB recommendations for cost-effective service provision translation into practice.
Other key challenges relate to health system readiness, data availability and quality, dissemination of information, high out of pocket costs, buy in from the medical and allied health professional communities and the powerful private sector and the availability of mechanisms for monitoring and evaluation. The complex relationship between the central and state government also presents an important challenge for the implementation of any federal government initiative, with health as a state-driven subject. In most states, mixed service delivery prevails, in which public sector facilities and programmes deliver many services side by side with expanding social insurance schemes. [bib_ref] Universal health insurance in India: ensuring equity, efficiency, and quality, Prinja [/bib_ref] HTA needs to acknowledge, address and contribute to improved effectiveness of both private and public care. In the context of national health decision-making, there are key considerations related to the best choice of technologies to ensure that the poor who largely use public facilities receive quality care. However, HTA may also act as a lever to address many of these systemic issues by increasing the credibility, accountability and quality of public and prepaid healthcare services and consequently enhancing their utilisation. [bib_ref] Strengthening costeffectiveness analysis in Thailand through the establishment of the health intervention..., Tantivess [/bib_ref] [bib_ref] Applying rapid 'de-facto' HTA in resource-limited settings: experience from Romania, Lopert [/bib_ref] [bib_ref] India's largest hospital insurance program faces challenges in using claims data to..., Morton [/bib_ref] conclusIon As India strives to deliver UHC, institutionalising HTA to inform how the government invests in health will be increasingly important. HTA will provide a systematic approach to evaluate the properties, effects and costs of health technologies or interventions while considering equity issues and health impact. Institutionalising HTA BMJ Global Health in India will require government support, technical capability and capacity, awareness among stakeholders, including physicians and allied health professionals, fostering of the culture for the assessment and a fit for purpose health system that can support its implementation uptake of the recommendations from the HTA framework in India. By bringing together Indian academics, health and economic experts and policy-maker representatives, with international experts in the field, the DHR is actively paving a way forward for improving priority setting in India. This represents the beginning of the long path towards building a sustainable HTA framework to inform coverage decisions as part of India's UHC agenda. In a country that is home to one-sixth of the global population, improving the health of the Indian population will have a resounding impact, not only for India, but for global health in general.
Contributors This article was drafted based on the collective experience of the authors in planning and working towards establishing MTAB in India. This initiative is led by the Indian MHFW by SS, Secretary DHR and VG, Joint Secretary DHR. Technical advice and support is provided to DHR by the International Decision Support Initiative (IDSI). LD drafted the first version of the manuscript. AM, AG, VG, KS, SP, FAC, SD, YT, SK and SS provided editorial feedback and comments.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
## Data sharing statement this study contains no unpublished data.
Open Access This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http:// creativecommons. org/ licenses/ by/ 4. 0/
© Article author(s) (or their employer(s) unless otherwise stated in the text of the article) 2017. All rights reserved. No commercial use is permitted unless otherwise expressly granted.
|
LiST as a catalyst in program planning: experiences from Burkina Faso, Ghana and Malawi
Background African countries are working to achieve rapid reductions in maternal and child mortality and meet their targets for the Millennium Development Goals (MDGs). Partners in the Catalytic Initiative to Save One Million Lives (CI) are assisting them by providing funding and technical assistance to increase and accelerate coverage for proven interventions. Here we describe how the Lives Saved Tool (LiST) was used as part of an early assessment of the expected impact of CI plans in Malawi, Burkina Faso and Ghana.MethodsLiST builds on country-specific demographic and cause-of-death profiles, and models the effect of changes in coverage for proven interventions on future levels of mortality among children less than 5 years of age. We worked with representatives of Ministries of Health and their development partners to apply LiST to assess the potential impact of CI plans and coverage targets, generating a short list of the highest-priority interventions for additional scale-up to achieve rapid reductions in under-5 mortality.ResultsThe results show that in each country, achieving national coverage targets for just four or five high-impact interventions could reduce under-5 mortality by at least 20% by 2011, relative to 2006 levels. Even greater gains could be obtained in Burkina Faso and Ghana by scaling up these high-impact interventions to 80%.DiscussionLiST can contribute to the development of stronger programmes by identifying the highest-impact interventions in a given epidemiological setting. The quality of LiST estimates is dependent on the available data on coverage levels and causes of death, and assumes that the target levels of coverage are feasible in a given context while maintaining service quality. Further experience is needed in the feasibility and usefulness of LiST as part of the program planning process at district and subdistrict levels.
Background African countries are working to achieve rapid reductions in maternal and child mortality and meet their targets for the Millennium Development Goals (MDGs). Partners in the Catalytic Initiative to Save One Million Lives (CI) are assisting them by providing funding and technical assistance to increase and accelerate coverage for proven interventions. Here we describe how the Lives Saved Tool (LiST) was used as part of an early assessment of the expected impact of CI plans in Malawi, Burkina Faso and Ghana.
# Methods
LiST builds on country-specific demographic and cause-of-death profiles, and models the effect of changes in coverage for proven interventions on future levels of mortality among children less than 5 years of age. We worked with representatives of Ministries of Health and their development partners to apply LiST to assess the potential impact of CI plans and coverage targets, generating a short list of the highest-priority interventions for additional scale-up to achieve rapid reductions in under-5 mortality.
# Results
The results show that in each country, achieving national coverage targets for just four or five high-impact interventions could reduce under-5 mortality by at least 20% by 2011, relative to 2006 levels. Even greater gains could be obtained in Burkina Faso and Ghana by scaling up these high-impact interventions to 80%.
Discussion LiST can contribute to the development of stronger programmes by identifying the highest-impact interventions in a given epidemiological setting. The quality of LiST estimates is dependent on the available data on coverage levels and causes of death, and assumes that the target levels of coverage are feasible in a given context while maintaining service quality. Further experience is needed in the feasibility and usefulness of LiST as part of the program planning process at district and subdistrict levels.
# Introduction
Earlier papers in this issue have described the Lives Saved Tool (LiST) and how it can make state-of-theart evidence about intervention effectiveness available to policy makers and programme planners as a basis for sound decision making about how to allocate resources to achieve the Millennium Development Goals (MDGs). Here we describe early applications of the tool in the context of the Catalytic Initiative to Save One Million Lives (CI).
The CI aims to accelerate coverage of interventions proven to be efficacious in reducing mortality from major causes of mortality in children under the age of 5 years, resulting in rapid reductions in child mortality at scale.As part of the CI, UNICEF is implementing the 'Integrated Health System Strengthening' project with support from the Canadian International Development Agency (CIDA) to accelerate reductions in under-5 mortality in six countries in sub-Saharan Africa (Ethiopia, Ghana, Malawi, Mali, Mozambique and Niger). In a complementary effort being implemented under the CI 'umbrella', WHO, UNICEF and UNFPA are working with support from the Bill & Melinda Gates Foundation (BMGF) to accelerate coverage of proven interventions and reduce child mortality in Burkina Faso, Malawi and Mozambique. These efforts aim for reductions in under-5 mortality by 2011 that are at least 20% greater in the project acceleration areas than in other areas of each country where the CI is not being implemented. By late 2008, all countries receiving CI funds had developed specific plans for scaling-up interventions within the context of their respective national strategies for maternal, newborn and child health.
The Institute for International Programs (IIP) at the Johns Hopkins University is responsible for conducting independent effectiveness evaluations of CI activities at country level in collaboration with in-country research partners. These evaluations are prospective, require innovative evaluation designs and are guided by principles that differ from most traditional evaluations: (i) the evaluations are based on a stepwise approach that includes early assessments of the programme design and the likelihood of achieving impact; (ii) use of LiST to provide detailed input to programme planners allowing them to make changes to initial plans and increase the probability of impact; (iii) independence from programme implementers with responsibility for providing regular feedback to improve programme effectiveness; and (iv) designs that are flexible and realistic, recognizing that good programmes change over the course of their implementation in response to contextual factors and ongoing feedback from internal and external monitoring and evaluation.In 2007 and 2008, country teams from Burkina Faso, Malawi and Mozambique submitted proposals for support from the Catalytic Initiative that described plans to accelerate coverage for proven interventions and produce reductions of at least 20% in rates of under-5 mortality in the project areas by 2011, over and above those achieved in areas of each country not implementing the accelerated CI approach. Country teams proposed to supplement existing government plans by accelerating the scale-up of between 13 (Burkina Faso) and 20 (Malawi and Ghana) interventions simultaneously in large areas of each country, raising concerns among CI leaders (CIDA, BMGF, the Doris Duke Foundation, UNICEF and WHO) about the feasibility of achieving targeted increases in coverage during the time frame of the project. It was therefore proposed that Ministries of Health and their CI counterparts be invited to participate in applications of LiST with the aim of narrowing the focus of their project plans to a smaller set of interventions that together could achieve the mortality reduction targets of the CI. We report here on the first three of these country applications in Malawi, Burkina Faso and Ghana.
# Methods
## Iip prepared for the visits by building preliminary
LiST scenarios using the most recent available estimates of country-specific under-5 deaths by cause 3 and coverage targets as presented in the country-specific CI plans. In Malawi, model inputs were national coverage estimates from a Multiple Indicator Cluster Survey (MICS) conducted between July and November 2006 4 . In Burkina Faso, coverage estimates for most interventions were available from a 2006 MICS.The national plan in Burkina is based on two or more antenatal care visits rather than the global consensus indicator of more than four visits, and this indicator had not been calculated for the 2006 MICS. We therefore adjusted the 2006 estimate for one or more visits using the reported relationship between at least one and two or more visits reported from the 2003 DHS. In Ghana the scale-up will focus in the Central, Northern, Upper East and Upper West Regions, for which no recent coverage data were available at the time of the LiST workshop. National data sets had limited applicability due to the huge variability by geographic region in child mortality burden and intervention coverage levels. We therefore used coverage estimates from a 2007 survey
[formula] LiST AS A CATALYST IN PROGRAM PLANNING i41 [/formula]
conducted in the Central, Northern, Upper East and Upper West Regions. [bib_ref] The Accelerated Child Survival and Development program in west Africa: a retrospective..., Bryce [/bib_ref] The model was applied only to the Upper East Region because there was only time in the workshop to run one set of models. [fig_ref] Table 1: Context for LiST applications in Malawi, Burkina Faso and Ghana Ministry of... [/fig_ref] summarizes the participants and the context for the LiST application in each country. In each setting, the IIP team asked in-country partners to identify a few individuals familiar with the national plans and comfortable with computer applications to work with the IIP team to apply LiST using the interventions and targets specified in the country plan. This smaller group worked together to apply LiST and prepare a summary of the results for presentation to the Ministry of Health leaders and a larger group of stakeholders. Uncertainty estimates were not presented as this aspect of LiST is still under development.
Some interventions in the country plans could not be included in the LiST modelling exercise. [fig_ref] Table 2: Reasons why interventions present in country plans were not modelled in LiST... [/fig_ref] shows these interventions and the reasons for their exclusion. In addition, coverage levels for DTP, measles, tetanus and Hib (where implemented) vaccinations were estimated at over 80% in 2006 in all three countries; 7 the LiST applications did not assume any further scale-up for these interventions as targets had already been achieved. In Malawi, vitamin A coverage was 480% in 2006 4 and we assumed coverage would be sustained at this level through 2011.
Coverage targets play an important role in determining the relative contribution of each intervention to mortality reductions as modelled by LiST, and considering alternative scenarios with higher and lower targets is an essential part of a full LiST application during the pre-or early-implementation period of a programme. The early applications reported on here began with the intervention-specific coverage targets defined in the national plan for the CI acceleration, most of which were generated through a workshopbased application of the UNICEF-supported Marginal Budgets for Bottlenecks (MBB) tool.In-country presentations of the results of applying LiST to initial implementation plans focused on: (i) the total percent reduction in under-5 mortality between baseline (around 2006) and 2011 if national coverage targets were achieved; and (ii) the contribution of individual interventions to the modelled reductions in child mortality, permitting a comparison of the relative contribution of different interventions to the achievement of the plan's outcome objectives. Here, we also present the total and intervention-specific reductions in under-5 mortality that would result from achieving 80% coverage with the highest-impact interventions in each country to illustrate the maximum impact that could be achieved by accelerating coverage for the most effective interventions focused on the major causes of child death in each country. Or indoor residual spraying. Full implementation of the national plan in Malawi is projected to result in a 36% reduction in under-5 mortality-from 123 in 2006 to 78 in 2011. Four interventions account for 55% of this reduction: correct treatment of childhood pneumonia, diarrhoea and malaria, and the use of insecticide-treated nets for the prevention of malaria. The second scenario (achieving 80% coverage for these four interventions) would result in a lower level of mortality reduction (31%) than the current national plan in this 5-year period, because the plan sets a target of 85% for ORS and zinc in the treatment of diarrhoea, which is the single largest cause of under-5 deaths, and the scenario target of 80% represents a reduction from this level of coverage.
Full implementation of the national plan in Burkina Faso is projected to result in a 24% reduction in under-5 mortality-from 170 in 2006 to 129 in 2011. Five interventions account for over 80% of this reduction: treatment of childhood pneumonia, diarrhoea and malaria, the prevention of malaria at household level through either insecticide-treated nets or indoor residual spraying, and vitamin A supplementation. Achieving 80% coverage for these five interventions would reduce under-5 mortality from 2006 levels by 40%, saving a total of about 20 000 additional child lives relative to the current national plan in this 5-year period.
Full implementation of the national plan in Ghana is projected to result in a 26% reduction in under-5 mortality-from 107 in 2006 to 79 in 2011 in the Upper East Region. Five interventions account for over 75% of this reduction: treatment of childhood pneumonia, diarrhoea and malaria, the use of insecticide-treated nets for the prevention of malaria, and improving sanitation at the household level from a baseline of 18-70%. Achieving 80% coverage for these five interventions would reduce under-5 mortality by 37% in this 5-year period, saving a total of $500 additional child lives relative to the current national plan in the Upper East Region. A larger number of deaths would be averted in the 54 districts that comprise the CI intervention area, but available data are insufficient to support an estimate.
## Responses from government and partners
In all three countries, responses to the LiST results highlighted the need for country-specific estimates of intervention effectiveness as inputs to programme priority setting. LiST can stimulate reassessments of existing coverage targets for the interventions with the highest potential for rapid mortality reduction. The ability of LiST to produce estimates of the point at which increases in coverage for specific interventions would result in measureable decreases in mortality was also appreciated.
Some stakeholders, and especially donor representatives, were struck by the fact that many more interventions were included in the scale-up plans than was necessary to achieve the mortality reduction target, leading to discussions about how to fine-tune the plans to give greater focus to the highest-impact interventions and achieve the greatest impact, while at the same time increasing feasibility of implementation. In Malawi, where the results were presented at the start of a week-long workshop convened to plan for the independent evaluation of the CI, high levels of concern about achieving coverage targets while maintaining quality resulted in a reorganization of the workshop programme to allow a separate working group of MOH personnel and their UN counterparts to focus on how to refocus implementation plans and reduce barriers to rapid acceleration of coverage for these interventions. In Burkina Faso, the advance team included high-level representatives of the Ministry of Health, and several targets were changed to result in higher impact during the LiST application (original results not shown here) and prior to the presentation of the results to a larger group of stakeholders. A stakeholders' meeting in Ghana provided a forum for presentation and discussion of the LiST results. In all three countries, the LiST application and resulting discussions highlighted the fact that national mortality reduction targets could be achieved if coverage for the existing immunization schedule was maintained and programmes were scaled up to achieve national coverage targets for up to five additional interventions including correct treatment of childhood diarrhoea, pneumonia and malaria. The LiST applications also sparked discussions about how barriers to rapid increases in coverage could be addressed.
Participants in the advance teams in each country were able to use LiST independently by the close of the preparatory period. In all three settings, local individuals took the initiative to explore alternative scenarios that went beyond the estimation of impact for the CI plans and targets. Members of the advance teams in two of the three countries have applied LiST independently since being exposed to it in the workshop. The ability to try out alternative scenarios-setting higher or lower targets for particular interventions and examining the effects on under-5 mortality-was particularly appreciated.
## Discussion and implications for programming
Applications of LiST as a part of programme planning exercises in Malawi, Burkina Faso and Ghana showed that in each country a set of four or five proven interventions could reduce under-5 mortality by at least 20% in a period of 5 years if existing national coverage targets are able to be achieved without jeopardizing the quality of the interventions. Even greater gains could be obtained in Burkina Faso and Ghana if the five most effective rapid-mortality-reduction interventions in each country were scaled up to coverage levels of 80%. These findings assume that existing high coverage levels for childhood immunizations in these countries will be maintained.
These conclusions are limited in several important ways. First, the estimates of lives saved are presented here as point estimates despite obvious uncertainties resulting from variability in baseline mortality levels (estimated from survey-based birth histories), in cause of death distributions (modelled for most countries), in baseline coverage levels (also estimated through surveys or in some cases from information systems), and in parameter specifications (e.g. assumptions about effect of specific interventions on cause-specific deaths), among other factors. The magnitude of the differences among interventions in the numbers of lives saved are sufficiently large in most cases to make the implications for programming clear, but further work is needed to address this issue. Second, the estimates produced by LiST assume that interventions will be delivered at levels of quality sufficient to produce effects on mortality equivalent to those assumed in the model. Third, coverage targets for some interventions in these countries were unrealistically high. In Ghana, for example, the national plan calls for an increase in improved sanitation from 18 to 70%.
Despite these limitations, these first applications of LiST highlight important considerations for the design of programmes to accelerate child survival. In all three countries, many more interventions are included in national plans than are needed to achieve the planned impact. Given severe resource constraints, it may not be possible to implement all currently planned interventions at scale simultaneously. A tighter focus on the highest-impact interventions could reduce the complexity of intervention packages and increase the probability that coverage and impact targets are achieved. Governments and their partners should assess the trade-offs between the rapid achievement of mortality targets by accelerating coverage for the highest-impact interventions and slower gains across a broader range of interventions that are less effective in their contexts.
These exercises also make clear, however, that LiST is only a tool and cannot in and of itself strengthen maternal and child policies and programmes. At the time of this writing (June 2009), none of the three countries has changed their formal targets or plans for accelerating coverage to reflect the outcomes of the LiST application. MOH coauthors highlight the challenges of the policy process and the barriers to rapid action based on LiST results (Box 1). In Ghana, for example, revising targets within the current 5-year plan of work (2007-11) would require revisiting not only the complicated process of target setting, but also the 'triggers' for funding that donors have set based on those targets. Overcoming this and other barriers to evidence-based programme planning will require that LiST is embedded into the broader cycle of programme planning and monitoring-not an easy thing to do. Despite the difficulties in changing the formal country plans to reflect the LiST results, MOH representatives in all three countries did report that they have changed the implementation plans for the scale-up to give greater emphasis to the four or five interventions within their existing programmes with the greatest impact.
These early applications of LiST also highlight important gaps in essential data for programme planning and evaluation at country level. First, countries and districts need accurate information on the causes of child deaths in order to generate accurate LiST results that can be used to prioritize interventions and allocate resources. When feasible, these data should be collected routinely through verbal autopsies conducted as a part of nationally representative household surveys; where they are not available, the most recent cause-of-death profiles developed by WHO and the Child Health Epidemiology Reference Group can be used in their stead.Second, countries and districts need frequent measures of coverage for all the interventions included in their national plans. In Burkina Faso, the most recent available coverage estimates for some interventions was 2003. Efforts by national and international partners to conduct household surveys at more frequent intervals and to maintain consistent indicator definitions across surveys to allow trend analyses are important steps; making those surveys feasible and affordable for more frequent use at national and even district levels is a continuing priority.
LiST must not be viewed as a 'magic bullet' that will solve all problems in planning and implementing strong public health programmes in child survival. First, the LiST applications reported here occurred in the context of specific programmes focused on accelerating intervention coverage and reducing child mortality. There are other worthy public health goals for which LiST may hold less immediate relevance. Second, those applying LiST as a part of programme planning exercises, even within child survival programmes, should keep in mind that the resulting estimates of lives saved assume that interventions that are currently reaching high proportions of the population will need to be sustained. Third, the potential contributions of LiST as part of an ongoing programme management process must be documented further at national and regional levels, and further experience is needed at district and subdistrict levels where most operational planning actually occurs.
# Funding
This work was supported by the Bill and Melinda Gates Foundation through grants to the World Health Organization to support the rapid scale-up of proven interventions in Burkina Faso, Malawi and Mozambique and to the US Fund for UNICEF to promote evidence-based decision making in designing maternal, neonatal and child health interventions in low-and middle-income countries, and by the Canadian International Development Agency through a grant to the Institute for International Programs at the Johns Hopkins Bloomberg School of Public Health for real-time monitoring of child mortality under the Catalytic Initiative to Save One Million Lives. (1) Political commitment. National public health policies and plans are shaped by strong forces other than scientific evidence, including technical clarity about what changes need to be made and their costs and benefits to the population, and competition for attention and resources among disease-and intervention-specific advocacy groups, some of whom are backed by enormous resources. LiST can contribute to efforts to put maternal and child survival at the top of the priority list at national and district levels, especially once the costing module is available. (2) Consensus. National plans, and especially sector-wide approaches (SWAps), are the result of an intensive collaborative process between governments and donors, with release of funds increasingly tied to achievement of milestones and targets. Changes reverberate through this system and are therefore difficult to make once the plan is approved. (3) Timing. National plans run in three-to five-year cycles, and once set are difficult to change as described above. District plans are usually developed on an annual basis and can therefore change more rapidly if supportive national policies are in place. LiST applications should be embedded in the policy and program planning processes at national and district levels.
i46 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
[fig] Figure 1: shows trends in rates of under-5 mortality for the three countries if coverage levels for all interventions in the national plan are sustained at 2006 levels, and for two scenarios generated by LiST: (i) trends in mortality if coverage targets in the national plan are achieved; and (ii) trends in mortality if the minimum number of interventions in the national plan that could achieve a 20% reduction in under-5 mortality by 2011 are scaled up to 80% coverage. Under-5 mortality rates increase slightly in the null scenario (if baseline coverage levels are maintained) due to historical trends in nutritional status and HIV prevalence among women of reproductive age. [/fig]
[fig] Box 1: Challenges in incorporating LiST results into program policies, plans and targets Coauthors from Ministries of Health in Malawi, Burkina Faso and Ghana (HN, GYA, NN, SKF) identified three factors that are likely to affect the speed and extent to which LiST results are taken up by governments and their counterparts and reflected in changes in official plans and resource allocation: [/fig]
[table] Table 1: Context for LiST applications in Malawi, Burkina Faso and Ghana Ministry of Health, Burkina Faso (Extrapolation based on 2006 census). c 2008 estimates from the Ghana Statistical Service (http://www.statsghana.gov.gh/docfiles/gh_figures_2008.pdf). d 2006 estimates from the United Nations Population Reference Bureau. [/table]
[table] Table 3: presents the number of interventions included in the country scale-up plan, the number of interventions modelled and the overall level of child mortality reduction between 2006 and 2011 if national coverage targets are achieved. The table also presents the interventions that will have the greatest impact on under-5 mortality in each country if national coverage targets are achieved, representing a cumulative reduction of at least 20% relative to baseline, as well as their baseline and target levels of coverage. All projections assume that 2006 coverage levels for childhood immunizations are sustained through 2011. [/table]
[table] Table 2: Reasons why interventions present in country plans were not modelled in LiST application Family planning reduces the number of child deaths; evidence for an effect of birth spacing on child mortality is under review and is not currently reflected in LiST.Now incorporated into LiST.LiST AS A CATALYST IN PROGRAM PLANNING [/table]
|
Diagnostics and Treatment of Volkmann Ischemic Contracture in a Seven-Year-Old Child
Ischemic contracture is a rare, but debilitating condition following supracondylar fractures in children requiring intensive interdisciplinary conservative and surgical treatment to relieve symptoms. Early signs of the preceding compartment syndrome consisting of unusually intense, prolonged pain and secondary nerval deficits early after trauma should to be recognized to prevent its development.Keywords► Volkmann ischemic contracture ► compartment syndrome ► tenotomy ► supracondylar fracture ► children Abstract A 7-year-old boy presented 6 weeks after open reduction and crossed Kirschner wire (K-wire) fixation of a supracondylar humerus fracture. Previous treatments had restored skeletal anatomy without documented complications. However, the patient would not move the entire arm, including his forearm and hand. Any passive movement led to anxious adverse reactions, and there was partial numbness of all fingers. After intensive physio-and occupational therapy supported by nerve stimulation and psychological counseling, anxiety-related functional deficits of the shoulder and elbow resolved to reveal the severe Volkmann contracture of the right hand developed fully. Electroneurography, X-ray, magnetic resonance imaging of the forearm, and ultrasonography showed nonfunctional ulnar and a partially disturbed radial motor nerve distal to the elbow along with damaged flexor muscles of the forearm after compartment syndrome. In addition, damage to the median nerve at the elbow level was diagnosed. After intense conservative therapy, we partially resected fibrotic fascia of the superficial flexor compartment, freed ulnar and median nerves, and performed staircase-like releases of tendons and tenotomies. We achieved a full range of motion of all fingers and markedly improved the range of motion of the wrist. The Disabilities of the Arm, Shoulder and Hand scores for function improved from 80 to 16 at the 2-year follow-up postoperatively, but some impairments of fine motor function persisted. Subtle symptoms of a developing compartment syndrome need to be recognized. Overlooked and untreated, a consecutive Volkmann contracture can turn the extremity nonfunctional. Intensive physical, psychological, and surgical therapy in a specialized center can restore function but requires endurance and perseverance throughout the lengthy recovery.
# Introduction
Supracondylar fractures of the humerus (SFH) represent 6.5% of pediatric fractures and 55 to 80% of elbow fractures in children, predominantly in 5-to 7-year-old children. [bib_ref] A 10-year study of the changes in the pattern and treatment of..., Cheng [/bib_ref] Especially in severely displaced fractures (Gartland type III), direct neuronal and vascular injuries may occur. [bib_ref] Clinical characteristics of severe supracondylar humerus fractures in children, Garg [/bib_ref] [bib_ref] Nerve injuries associated with pediatric supracondylar humeral fractures: a meta-analysis, Babal [/bib_ref] Additionally, increased pressure from swelling and hematoma after soft-tissue trauma can lead to an acute compartment syndrome (ACS) that occurs in 0.1 to 0.3% of children with SFH, while the volar superficial and deep flexor compartments are most commonly affected. [bib_ref] Factors affecting forearm compartment pressures in children with supracondylar fractures of the..., Battaglia [/bib_ref] [bib_ref] Vascular and neural complications in supracondylar fractures of the humerus in children, Lipscomb [/bib_ref] [bib_ref] Acute compartment syndrome of the forearm, Botte [/bib_ref] Since Baumann's vertical suspension and plastering have been replaced by percutaneous Kirschner wire (K-wire) osteosynthesis as the primary choice in the management of SFH, the incidence of the ACS has declined drastically and is, therefore, rarely encountered anymore by the medical staff. [bib_ref] Vascular and neural complications in supracondylar fractures of the humerus in children, Lipscomb [/bib_ref] [bib_ref] Clinical review: Volkmann's ischaemic contracture, Pettitt [/bib_ref] Typical clinical signs of ACS of the forearm in adults are summarized as the six "P's": pain, positive passive stretch test, paresthesia, paralysis, painful tense forearm, and at times, a pulseless limb. [bib_ref] Ischaemic contracture of the hand, Balakrishnan [/bib_ref] In children, however, the primary symptoms are rather summarized as the three "A's": anxiety, agitation, and increased analgesic requirements due to pain out of proportion along with pain on passive stretching.Increasing pain and firmness of a muscle compartment on palpation over time may indicate the presence of a sub-ACS. But even in the absence of pain, a compartment that is rigid on compression with impairment of peripheral blood perfusion can lead to the diagnosis of silent ACS. [bib_ref] Silent compartment syndrome in children: a report of five cases, Lee [/bib_ref] In any case, ACS needs to be recognized and treated immediately by fasciotomy to prevent the development of Volkmann ischemic contracture (VIC). VIC was first described by Richard von Volkmann in 1881 as a loss of function due to tissue damage following ACS. [bib_ref] The ischemic muscular paralysis and trauma, Von Volkmann [/bib_ref] The symptoms of VIC of the volar compartment of the forearm are shown in ►Table 1. 9 Physiotherapy, including passive stretching of contracted muscle fibers and joints and splinting to oppose further contractions, has to start immediately. [bib_ref] Acute compartment syndrome of the forearm, Botte [/bib_ref] Surgical treatment choices include skin release and Z-plasty for mild contractures, the release of secondary nerve compression, muscle slides, and tendon lengthening in moderate contractures, and finally, tendon transfers or free-tissue transfers as well as salvage procedures for severely contracted or neglected extremities. 14 As a consequence of this disorder's heterogeneous clinical presentation, every patient requires a customized therapy strategy to achieve alleviation from handicap to the greatest possible extent. We present the following case to outline our treatment approach and its outcome in a patient with a severe VIC.
## Patient information clinical findings
A 7-year-old boy fell on his right arm while playing and, subsequently, another child stepped on the same arm. A local pediatric trauma surgeon in an adult trauma unit correctly diagnosed the Gartland III SFH and performed an open reduction via a dorsal approach followed by crossed K-wire osteosynthesis. The ulnar nerve was visualized and found unharmed during surgery. A dorsal noncircumferential cast was administered in 90 degree flexion of the elbow at the end of the surgery. Postoperatively, a sensory and motoric deficit in the area covered by the ulnar nerve arose without any signs of further impairment of perfusion, motion, or sensibility. The radial pulse was palpable at all times. The boy complained about severe pain over a prolonged time. In the following weeks, the range of motion of the right hand deteriorated drastically. Nerve conduction velocity was measured 4 weeks after the injury. All three nerves in the forearm were affected with a complete absence of conduction in the ulnar and median nerves. The K-wires and the cast were removed before in-patient rehab 6 weeks after the injury. After failing to improve within the first week of rehab, the child was submitted to our child trauma reference center, where the patient presented with a relieving posture of the entire right arm.
The radial and ulnar pulses were palpable, and the capillary refill showed no delay. Upon assessment of the motoric functions, the shoulder muscles M. deltoideus and M. trapezius were already hypotrophic. The elbow was held in a 20 degree flexed position with a complete lack of active motion. Passively, 90 degree flexion was possible under intensive distraction. In the wrist, the patient could only carry out wiggling motion of a maximal deflection of 20 degree, and the hand was mainly held fixed in a 20 degree flexed position. A minimal active extension of the wrist excluded a complete deficit of radial motor nerve. The fingers could not be moved and displayed flexion in all joints, and passive extension of the fingers required force and again intensive distraction of the child. The sensibility was only absent distally to the metacarpophalangeal joints. Additionally, the thenar presented with dysesthesia. In summary, no isolated or combined nerve lesion explained our clinical findings.
## Medical, family, and psychosocial background
The patient, a boy of lean stature with a gluten-free diet following the diagnosis of celiac disease, has had no prior surgery other than circumcision. The otherwise healthy and well-integrated student in grammar school had many interests including sports (cycling, swimming, and karate) and playing the violin. There was no family history of any medical
## Diagnostic assessment
Our primary clinical findings were inconclusive and did not point to any specific nerve dysfunction. Besides, the patient showed a severe adverse reaction to the right upper limb's touch and motion, indicative of an additional psychological cause. According to the Budapest criteria for the diagnosis of a complex regional pain syndrome, the diagnosis is made when there is continuing pain disproportionate to the inciting event and certain alterations in three out of the four categories sensory, vasomotor, sudomotor, and motor/trophic. [bib_ref] Validation of proposed diagnostic criteria (the "Budapest Criteria") for complex regional pain..., Harden [/bib_ref] Our patient presented with hyperesthesia, temperature asymmetry, and beginning hypotrophy of the affected limb alongside the severe pain so that the criteria were fulfilled. Diagnostics included a magnetic resonance (MR) angiography, an X-ray of the forearm, repeated nerve conduction velocity assessments, and sonography of the nerves (►Fig. 1b, c) and muscles of the forearm. They revealed an injury to the median nerve at the elbow joint and muscle deterioration in both flexor muscle compartments with the ulnar nerve's affection 7 cm distally to the ulnar epicondyle following the suspected compartment syndrome. The brachial, radial, and ulnar arteries did not show any sign of insufficiency. After an intensive course of conservative therapy, the shoulder and upper arm's muscular deficit resolved and the range of motion of the elbow joint improved significantly. Subsequently, the typical Volkmann contracture developed and could be observed 18 weeks postinjury (►Fig. 1a).
## Therapeutic interventions
After the cast and metal removal 6 weeks postinjury, the patient attended an intensive inpatient rehabilitation program for 5 weeks. It included physiotherapy and occupational therapy under analgesics to extend the active and passive range of motion of the affected limb supported by psychological consultations. To prevent worsening of the contracture of the wrist and hand, a volar splint was fitted. Additionally, physical therapy was administered for sensory stimulation and tonus regulation. At the end of rehab, the patient had fully restored function and strength of the affected shoulder and upper arm, extended range of motion of the elbow to an extension/flexion of 0-40-100, and improved pronation to 60 degree. However, the clinical picture of VIC with severely limited extension of the wrist and contracture of the fingers in flexion developed fully. Intensive conservative therapy was continued in an outpatient setting adding further modalities such as electrostimulation, osteopathy, therapeutical swimming, and BEMER physical vessel therapy. [bib_ref] The technological development history and current significance of the "physical BEMER® vascular..., Bohn [/bib_ref] In the meantime, diagnostics were completed.
When conservative therapy failed to improve the clinical picture further, the surgical procedure was planned and performed by an interdisciplinary team of pediatric and hand surgeons. For optimal vision, we operated with a tourniquet and magnifying glasses. We excised severely fibrotic and thickened fascia tangentially via a proximal incision the size of the distal incision (►Fig. 2) in an attempt to improve muscle function, as it has proven to be successful on the lower extremity. In this case, it, unfortunately, did not improve motility. We then created access to the distal volar forearm. Here, the median and ulnar nerves (►Fig. 2a) alongside the radial and ulnar arteries were microsurgically depicted and spared. We tenotomized M. palmaris longus and Mm. flexors digitorum superficialis D2 to 5 in addition to the stepwise elongation of the tendons of Mm flexor carpi ulnaris and flexor carpi radialis as well as Mm. flexor digitorum profundii D2 to 5 (►Fig. 2b). Finally, an elongation of the M. flexor pollicis longus tendon allowed for full extension of all fingers and wrist. The partially necrotic and fibrotic flexor muscles were spared during the surgery to prevent further damage. After surgery, a Kleinert splint was custom-fitted to enable early postoperative training and maintain tendon mobility. Meanwhile, the patient proceeded with occupational therapy.
## Follow-up and outcomes
Immediately after the intervention, the patient demonstrated an excellent range of motion of the fingers on active extension and passive flexion by the Kleinert splint. [bib_ref] Rehabilitation following surgery for flexor tendon injuries of the hand, Peters [/bib_ref] Furthermore, the patient achieved a completely unimpaired range of motion of the elbow and forearm and subsequently improved the right upper limp's flexibility and strength.
In the final clinical assessment and for the time being, the patient felt content with the outcome. Minor residues were no longer noticeable, and the slightly diminished strength in finger flexion did not bother. The range of motion of the hand and wrist was close to normal. The Disabilities of the Arm, Shoulder, and Hand scores throughout the recovery process improved from 80 to 16 in overall function and from 100 to 0 in sports. With the help of his very supportive and encouraging family, who actively took part in the therapy, the patient has now started a career in swimming, where he already competes on regional and national levels.
# Summary and discussion
In this case of a 7-year-old patient with a severe SFH treated by open reduction and K-wire fixation plus dorsal plaster cast, the development of a sub-ACS was missed and therefore left untreated. The subsequent psychological affections further complicated clinical examination and diagnosis upon referral to our pediatric trauma center. Nonetheless, the appropriate initial therapy was commenced in a rehabilitation clinic that comprised physiotherapy with passive stretching of the affected flexor tendons and muscles under analgetic therapy and active training of the affected and unaffected muscles of the arm and shoulder as well as occupational therapy and psychological counseling. In the meantime, further diagnostic measures were taken to elucidate the cause of the symptom complex including MR angiography, repeated electroneurography, and specialized ultrasonography of the limbs' nerves. They revealed damage to the median nerve at the level of the cubital fossa as a consequence of the initial injury and affection of the flexor muscles and ulnar nerve as well as partial radial nerve at the level of the forearm due to ACS prompting the clinical picture of VIC.
A multidisciplinary case conference involving neurosurgeons, hand surgeons, and orthopedic surgeons discussed the diagnostic and further therapeutic steps. These included a surgical intervention after the conservative efforts were exhausted. Surgical options vary according to the severity of the condition. They include skin release and Z-plasty for mild to moderate contractures and the release of secondary nerve compression, muscle slides, tendon lengthening, tendon or Steps of the surgical procedure after neurolysis and stepwise tenolysis and tenotomies. free-tissue (e.g., gracilis muscle) transfers, and salvage procedures for severely contracted or neglected extremities. 14 Unfortunately, an earlier surgical intervention after referral to us and before the completion of the deterioration process would not have been promising to yield a more fortunate result. [bib_ref] Results of neurolysis in established upper limb Volkmann's ischemic contracture, Meena [/bib_ref] In our case a combination of fascia release, neurolysis, tenotomies, and tendon elongations was performed and postoperatively only minor deficits persisted. Sadly, the extension of the wrist, which should be the first goal of therapy, 14 could not be maintained completely throughout time, which is most likely due to the M. palmaris longus tendon's persistent adherence, despite the tenotomy performed. A persisting and impairing fine motor function deficit remained mainly due to the small finger muscles' affection (Mm. lumbricales) by the initial lack of perfusion during the ACS, which cannot be improved further by surgical therapy.
We conclude that VIC resulting from an overlooked ACS after SFH in children is a potentially devastating condition requiring a long course of intensive conservative and surgical treatment. However, despite all efforts restitutio ad integrum may never be achievable as is the case with our patient. In the pediatric population, clinical examination may sometimes be challenging due to adverse behavior and pain. Furthermore, circular plaster casts are regularly administered in the treatment of fractures and present an additional risk for the development of a sub-ACS. They may also limit the access to the limb for clinical assessment and, therefore, delay diagnosis. Health care practitioners must recognize the three A's anxiety, agitation, and increased analgesic requirement as the symptoms of ACS and upon suspicion, compartment pressure measurement should be done immediately followed by fasciotomy to prevent further damage to muscular and nerve structures if necessary. [bib_ref] Ischaemic contracture of the hand, Balakrishnan [/bib_ref] [bib_ref] Shore BJKS LPediatric acute compartment syndrome, Livingston [/bib_ref] Even in the absence of pain and the presence of symptoms such as severe swelling and impaired or inconspicuous neurovascular function, a possible silent ACS needs to be taken into account and treated timely.Despite the remaining limitations mainly consisting of the lack of strength in finger flexion and an impairment of full extension in the wrist, our patient leads a happy life pursuing his hobby swimming, where he feels no restrictions at all and competes on a national level. We are content with the success of our interdisciplinary team effort and wish our patient the best for his future.
## Informed consent
The patient and his mother gave written informed consent for the anonymous usage of their data and pictures.
[fig] Figure 1: (a) Clinical picture of Volkmann ischemic contracture preoperatively and (b and c) nerve ultrasonography of the ulnar nerve at cubital and forearm level. [/fig]
[table] Table 1: Tsuge classification-Volkmann ischemic contracture condition. The patient lived in a household with his parents and a 10-year-old healthy sister. [/table]
|
Reasons for Non-Completion of Health Related Quality of Life Evaluations in Pediatric Acute Myeloid Leukemia: A Report from the Children’s Oncology Group
Background: Health related quality of life (HRQL) assessments during therapy for pediatric cancer are important. The objective of this study was to describe reasons for failure to provide HRQL assessments during a pediatric acute myeloid leukemia (AML) clinical trial.Methods:We focused on HRQL assessments embedded in a multicenter pediatric AML clinical trial. The PedsQL 4.0 Generic Core Scales, PedsQL 3.0 Acute Cancer Module, PedsQL Multidimensional Fatigue Scale, and Pediatric Inventory for Parents were obtained from parent/guardian respondents at a maximum of six time points. Children provided self-report optionally. A central study coordinator contacted sites with delinquent HRQL data. Reasons for failure to submit the HRQL assessments were evaluated by three pediatric oncologists and themes were generated using thematic analysis.Results: There were 906 completed and 1091 potential assessments included in this analysis (83%). The median age of included children was 12.9 years (range 2.0 to 18.9). The five themes for non-completion were: patient too ill; passive or active refusal by respondent; developmental delay; logistical challenges; and poor knowledge of study processes from both the respondent and institutional perspective.Conclusions:We identified reasons for non-completion of HRQL assessments during active therapy. This information will facilitate recommendations to improve study processes and future HRQL study designs to maximize response rates.
# Introduction
Health related quality of life (HRQL) assessments during therapy for pediatric cancer are important. This information is needed to better understand the patient experience and to identify if, when, and how interventions should be considered to improve HRQL [bib_ref] Patientreported outcomes assessment in clinical trials: taking stock, moving forward, Lipscomb [/bib_ref] [bib_ref] Health related quality of life in NCIsponsored cancer treatment trials, O'mara [/bib_ref]. When HRQL is assessed in the context of different treatments, this information may also be used to understand how the regimens differ from the patient and family perspectives and may help with choosing a treatment strategy [bib_ref] Health related quality of life in NCIsponsored cancer treatment trials, O'mara [/bib_ref] [bib_ref] Challenges and recommendations for advancing state-of-the-science of quality of life assessment in..., Buchanan [/bib_ref] [bib_ref] Health-related quality of life and symptom management research sponsored by the National..., Minasian [/bib_ref].
In pediatric cancer, most evaluations of HRQL have been conducted in single center, stand-alone studies. It is possible to embed HRQL evaluations into multi-center co-operative group trials. The advantages of embedding HRQL assessments into a therapeutic trial include efficiency in trial conduct and data management [bib_ref] Challenges and recommendations for advancing state-of-the-science of quality of life assessment in..., Buchanan [/bib_ref]. Furthermore, this approach has the potential for observing interactions among symptoms, calculating quality adjusted survival and cost-effectiveness, and generating new hypotheses [bib_ref] Challenges and recommendations for advancing state-of-the-science of quality of life assessment in..., Buchanan [/bib_ref].
HRQL assessed during intensive treatment is likely to provide informative data. Such intensive treatments may include chemotherapy for acute myeloid leukemia (AML), treatment for relapsed disease, and during hematopoietic stem cell transplantation (HSCT). AAML1031 is a Phase 3 randomized trial being conducted by the Children's Oncology Group (COG) that involves both chemotherapy and HSCT for patients. An ancillary HRQL aim was embedded in the trial for the purpose of describing HRQL associated with chemotherapy versus HSCT in this intensively treated patient population.
In pediatric oncology, response rates when eliciting information about the symptom experience and HRQL can vary significantly with response rates ranging from 58-98% in recent studies [bib_ref] Parental perceptions of health-related quality of life of children with leukemia in..., Tremolada [/bib_ref] [bib_ref] Evaluating the ability to detect change of health-related quality of life in..., Klaassen [/bib_ref] [bib_ref] Health-related quality of life: changes in children undergoing chemotherapy, Banks [/bib_ref] [bib_ref] Assessment of health-related quality of life in children with cancer: consistency and..., Russell [/bib_ref] [bib_ref] Quality of life during active treatment for pediatric acute lymphoblastic leukemia, Sung [/bib_ref] [bib_ref] Effect of dexamethasone on quality of life in children with acute lymphoblastic..., Vries [/bib_ref]. Although response rates vary considerably by study, to date, there are no pediatric studies that have described reasons for noncompletion of HRQL assessments. With our experience with AAML1031, we had a unique opportunity to describe the reasons behind non-completion of HRQL assessments. The objective was to describe reasons for failure to provide HRQL assessments during an ongoing pediatric AML clinical trial.
# Methods
## Details of aaml1031 hrql assessments
AAML1031 (NCT 01371981) is a Phase 3 COG multi-center trial that randomizes patients to receive or not receive bortezomib for de novo AML and determines the safety of sorafenib in patients with high allelic ratio FLT3+/ITD. This study uses a 4-course chemotherapy backbone (Induction I, Induction II, Intensification I, and Intensification II). Patients with low risk disease receive 4 courses of chemotherapy while those with high risk disease receive 3 courses of chemotherapy followed by allogeneic HSCT if an appropriate donor is identified. Thus, all patients receive three courses of chemotherapy and then either Intensification II or HSCT.
This study was approved by the research ethics board at all participating institution (table S1). Participants provided written informed consent to participate in this study. Written informed consent was obtained from parents or guardians on behalf of the minor/child participants in this study.
AAML1031 has an embedded secondary aim to assess HRQL of children and adolescents treated with chemotherapy and HSCT for AML. The study also seeks to describe parental stress for these patients. Consent for the HRQL aim is obtained at the same time as consent for the therapeutic study; eligible patients/families who consent to AAML1031 are offered the ability to participate in the HRQL aim as well. The instruments for HRQL assessment are the PedsQL 4.0 Generic Core Scales [bib_ref] The PedsQL Generic Cores Scales: sensitivity, responsiveness, and impact on clinical decision-making, Varni [/bib_ref] [bib_ref] PedsQL 4.0: reliability and validity of the Pediatric Quality of Life Inventory..., Varni [/bib_ref] , PedsQL 3.0 Acute Cancer Module [bib_ref] The PedsQL in pediatric cancer: reliability and validity of the Pediatric Quality..., Varni [/bib_ref] , and PedsQL Multidimensional Fatigue Scale [bib_ref] The PedsQL in pediatric cancer: reliability and validity of the Pediatric Quality..., Varni [/bib_ref] [bib_ref] The PedsQL Multidimensional Fatigue Scale in pediatric rheumatology: reliability and validity, Varni [/bib_ref] , which measure generic HRQL, cancer-specific HRQL and fatigue respectively. The Pediatric Inventory for Parents is also included and measures parental stress [bib_ref] Childhood illnessrelated parenting stress: the pediatric inventory for parents, Streisand [/bib_ref]. The estimated time to complete all assessments is 23-32 minutes. There are eight assessment time points stipulated by the protocol as follows: 1) within 14 days of Induction I initiation; 2) $ day 21 of Induction II, but prior to start of Intensification I; 3) $ day 21 of Intensification I, but prior to start of Intensification II; 4) one month (67 days) from start of intensification II or HSCT; 5) four months (61 month) from start of Intensification II or HSCT; 6) 12 months (61 month) from date of diagnosis; 7) 24 months (63 months) from date of diagnosis; and 8) 36 months (63 months) from date of diagnosis.
For the HRQL aim, eligible patients are between 2 and 18 years of age with a parent or guardian who can read in English. Parents or guardians provide proxy assessments for all patients, while for patients $5 years of age who can understand English, self-report is also completed if the child provides consent or assent. For all consenting participants, the age-appropriate questionnaires are downloaded from the COG website by institutional clinical research associates (CRAs). Completed forms are reviewed by the CRA for completeness. Responses are entered into the COG database via remote data entry and the original forms are retained by the institutions. Respondents that do not complete a questionnaire (parent or child form) at a particular time point continue to participate in subsequent time points as long as consent to participate is not withdrawn. Respondents continue to submit HRQL assessments until: they complete all questionnaire time points; consent to participate in the HRQL study is withdrawn; or the patient is removed from AAML1031 protocol therapy (for reasons such as relapse, refractory disease or death) in which case they are no longer eligible for the HRQL study.
## Assessment of reasons for failure to submit hrql assessments
In order to improve the response rate for the HRQL study, a central COG study coordinator contacted sites with enrolled patients with delinquent HRQL data. The coordinator encouraged submission of delinquent data and requested reasons for the delinquency. A second email was sent in the case of continued delinquency within the next month. A query was also generated in the case of incomplete HRQL forms.
There were two sources of comments related to non-completion of HRQL data. First, any comments that were communicated to the COG coordinator during the queries related to delinquent data were included. Second, some site CRAs provided comments directly into the COG database about why some time points were not conducted.
# Analysis
Three authors (DJ, RN and LS) independently coded all comments received. Themes were identified using thematic analysis using the methodology described by Braun and Clark [bib_ref] Using thematic analysis in psychology, Braun [/bib_ref]. The same authors also elucidated sub-themes within each theme [bib_ref] Using thematic analysis in psychology, Braun [/bib_ref] [bib_ref] Applying thematic analysis theory to practice: a researcher's experience, Tuckett [/bib_ref]. The study team met repeatedly to redefine themes and sub-themes in an iterative, continuous process. Sample quotes were identified to support themes and sub-themes.
# Results
As of January 25, 2013, there were 253 patients enrolled on AAML1031 who were eligible for the HRQL aspect of the study. Of these 253 patients, 196 consented to this aspect of the study (77.5% participation rate). The median age of these patients was 12.9 years (range 2.0 to 18.9) and 101 (51.5%) were boys. Of these 196 patients, 77 were removed from the HRQL study for the following reasons: withdrew (n = 15), deceased (n = 13), relapsed (n = 13), removed from protocol therapy for other reasons (n = 35), and site did not have the ability to gather the HRQL data (n = 1). Given the activation date of AAML1031, the number of individual HRQL assessments that were completed and the number that should have been completed at each time point from parents and patients combined were as follows: 1 (n = 959/1133); 2 (n = 749/ 923); 3 (n = 624/741); 4 (n = 454/578); 5 (n = 273/304); 6 (n = 64/ 85); 7 (n = 0); and 8 (n = 0). Of these assessments, 3123 were completed from a total of 3764 expected, giving a completion rate of 83%. Of note, no patients had reached time points 7 or 8 at the time of data collection for this analysis.
There were sixty-two comments received that described reasons for non-completion. The reasons for non-completion were grouped into five themes [fig_ref] Figure 1: Reasons for HRQL Non-Completion [/fig_ref]. The five themes were: patient too ill; passive or active refusal by respondent; developmental delay; logistical challenges; and knowledge of study processes. Some comments were categorized into more than one theme. The number of comments in each theme and subtheme, as well as sample comments, is shown in .
## Patient too ill
The theme that was most frequent was that the patient was too ill to complete the assessment. This was further subcategorized into patient too ill to complete self-report forms. The majority of comments described that the patient was in the intensive care unit and often was intubated and ventilated.
''Patient has been extremely ill in the PICU, and although approached… on three occasions, she has not been able to complete it.'' Another subtheme in this category was that the patient was too ill to ask the family to complete the forms. It was felt that it would be too burdensome to the family to complete the HRQL forms.
''.. he was intubated for the majority of the first time point and hospital staff did not feel it was appropriate to give these forms to his family due to their great distress….''
## Passive or active refusal by respondent
Another common theme was that of passive or active refusal by the respondent. Active refusal was subcategorized into 4 subthemes. First was active refusal because the family was overwhelmed.
''Parents indicated that it was just too much for them to do.''
There was also active refusal because the respondent indicated that completion of the HRQL forms was too burdensome. ''We have two patients on this study so far and the frequency of the time points is pretty difficult for the first family.''
In some cases the reason for the active refusal was not specified. Sometimes, a parent would not permit the child complete the HRQL assessment.
## Developmental delay
For children with severe developmental delay, self-assessment was not possible.
## Logistical challenges
Logistical challenges in completing the assessments were frequently described. There were four subthemes within this classification that emerged. First, the forms were lost on several occasions. A second logistical challenge encountered was that the forms were either not given to the respondent or it was uncertain whether they were given to the respondent. Third, the completed forms were unable to be retrieved from the patient and family. The fourth challenge was that the parents were rarely present in the hospital and thus, the institutional CRA was not able to provide the HRQL forms to them.
''Patient is inpatient parent is not here often. Left HRQOL for parent to fill out but did not get returned.''
## Knowledge of study process
Lack of knowledge of the study process was another important reason for non-completion of assessments. First, lack of knowledge was on the part of the respondent.
''The patient did not fill out the form because they thought it was an extra form so they did not complete it. I think they got it confused with the parent fatigue form.'' Second, institutional lack of knowledge of the study process was another reason for non-completion of assessments.
''Site misunderstood the word "proxy." Thought if patient can do it, parent doesn't need to.'' [fig_ref] Table 2: Suggestions for Improving Completion of HRQL Assessments [/fig_ref] illustrates our recommendations for how to address the identified issues in future HRQL studies of similar patients.
# Discussion
We identified reasons for non-completion of HRQL assessments within the current COG de novo AML study. Our observations are relevant for at least two reasons. First, we were able to summarize reasons that HRQL assessments were not completed for an intensively treated group of children and subsequently, can provide guidance to others for how to better design similar studies. Second, we have illustrated how this methodology can efficiently identify such problems.
The most frequent reason for non-completion of HRQL assessments was that the patient was too ill to complete the evaluation or to ask the parent to provide evaluations. A potential strategy to overcome this problem is to increase the time frame during which the respondent can complete the assessment. This approach would provide more opportunities for the respondent to complete the assessment during the window for completion. Second, decreasing the number of assessments and thus minimizing respondent burden for each assessment may be helpful. Third, providing guidance and a formalized ''script'' to CRAs on how to approach parents of critically ill children also may be beneficial. Finally, allowing for CRA administration (where the CRA reads the questions out loud and marks off the appropriate response) may allow these evaluations to be completed.
The second most common reason for non-completion of the HRQL assessments was active or passive refusal by the patient or parent. All of the previously described strategies to address the patient being too ill may also be effective at reducing the active and passive refusal rate.
Electronic patient reported outcome (ePRO) systems use electronic data capture methods to assess topics patients can report about themselves and these may be an extremely effective approach to addressing many reasons for non-completion of HRQL assessments [bib_ref] Recommendations on evidence needed to support measurement equivalence between electronic and paper-based..., Coons [/bib_ref]. ePROs may lead to less administrative burden, high patient acceptance, avoidance of secondary data . Reasons for Non-completion of Health Related Quality of Life Assessments and Sample Quotations.
## Themes and subthemes
No* Sample Quotations
## Patient too ill
a. Patient too ill to complete self-report forms 13 ''Patient has been extremely ill in the PICU, and although approached… on three occasions, she has not been able to complete it.'' b. Patient too ill to ask family to complete forms 3 ''.. he was intubated for the majority of the first time point and hospital staff did not feel it was appropriate to give these forms to his family due to their great distress…as well as the patient was unable to complete the forms himself.'' entry errors, easier implementation of skip patterns, electronic scoring and more accurate and complete data [bib_ref] Recommendations on evidence needed to support measurement equivalence between electronic and paper-based..., Coons [/bib_ref]. Study patients/families can be automatically emailed the questionnaire at the time of each scheduled assessment, ensuring timely presentation of study assessments [bib_ref] Electronic patient-reported outcome systems in oncology clinical practice, Bennett [/bib_ref]. As well, the PRO data is captured in the data file as the patient is responding to the questionnaire and any changes to the protocol can be easily accommodated in the ePROs [bib_ref] Recommendations on evidence needed to support measurement equivalence between electronic and paper-based..., Coons [/bib_ref]. Many logistical challenges can be overcome with ePROs, but another strategy to improve HRQL response rates is to have dedicated research staff. This has been evaluated in a previous analysis of non-therapeutic studies in cooperative group trials where funding for dedicated research staff was found to overcome many logistical challenges [bib_ref] Successful coordination and execution of nontherapeutic studies in a cooperative group setting:..., Carter [/bib_ref]. The benefits of dedicated research staff include optimizing response rates on HRQL studies, ensuring that the HRQL studies are a priority, and improving knowledge of the study processes.
## Refusal by respondent
The strength of our study is the inclusion of comments from a large number of institutions from multiple CRAs. This approach provided more generalizable data. A limitation of our study is the reliance on the institutional CRA to provide comments for reasons of non-completion. Consequently, we may have missed some important reasons for non-completion if some CRAs did not wish to provide their comments. Also, the study was only open for patients between the ages of 2 and 18, and thus, we could not evaluate reasons for non-completion of HRQL assessments specific to infants or older adolescents. Another limitation of our study is that patients with relapsed or refractory disease were removed from AAML1031 protocol therapy and were therefore not eligible to continue with the HRQL study. These patients may have particularly poor compliance with completion of HRQL assessments related to severity of illness and future research should focus on this subgroup.
In conclusion, we have described reasons for non-completion of HRQL assessments on an AML therapeutic trial in progress and provided suggestions to improve future studies. We also demonstrated that the collection of qualitative comments on reasons for non-completion of HRQL assessments is feasible.
## Supporting information
[fig] Figure 1: Reasons for HRQL Non-Completion. Diagram of the reasons for non-completion of health-related quality of life assessment completion. doi:10.1371/journal.pone.0074549.g001 [/fig]
[table] Table S1: List of Institutions with AAML1031 Open at Their Institution. (DOC) [/table]
[table] Table 2: Suggestions for Improving Completion of HRQL Assessments. Patient too ill N Increase time frame for completion of forms N Provide guidance for CRAs on how to approach parents of critically ill children N Decrease number of assessments N Decrease assessment length N Consider CRA administration of forms Passive or active refusal by respondent N Increase time frame for completion of forms N Provide guidance for CRAs on how to approach parents of critically ill children N Decrease number of assessments N Decrease assessment length N Consider CRA administration of forms Developmental delay N Provide specific guidance on completion of self-report for children with development delay Logistical challenges N Use electronic patient-reported outcomes N Provide access for proxy reporting from remote sites Knowledge of study processes N Use electronic patient-reported outcomes N Improve educational materials Abbreviation: CRA -clinical research associate. doi:10.1371/journal.pone.0074549.t002 [/table]
|
Postoperative delirium prediction using machine learning models and preoperative electronic health record data
Background: Accurate, pragmatic risk stratification for postoperative delirium (POD) is necessary to target preventative resources toward high-risk patients. Machine learning (ML) offers a novel approach to leveraging electronic health record (EHR) data for POD prediction. We sought to develop and internally validate a ML-derived POD risk prediction model using preoperative risk features, and to compare its performance to models developed with traditional logistic regression.Methods: This was a retrospective analysis of preoperative EHR data from 24,885 adults undergoing a procedure requiring anesthesia care, recovering in the main post-anesthesia care unit, and staying in the hospital at least overnight between December 2016 and December 2019 at either of two hospitals in a tertiary care health system. One hundred fifteen preoperative risk features including demographics, comorbidities, nursing assessments, surgery type, and other preoperative EHR data were used to predict postoperative delirium (POD), defined as any instance of Nursing Delirium Screening Scale ≥2 or positive Confusion Assessment Method for the Intensive Care Unit within the first 7 postoperative days. Two ML models (Neural Network and XGBoost), two traditional logistic regression models ("clinician-guided" and "ML hybrid"), and a previously described delirium risk stratification tool (AWOL-S) were evaluated using the area under the receiver operating characteristic curve (AUC-ROC), sensitivity, specificity, positive likelihood ratio, and positive predictive value. Model calibration was assessed with a calibration curve. Patients with no POD assessments charted or at least 20% of input variables missing were excluded.Results: POD incidence was 5.3%. The AUC-ROC for Neural Net was 0.841 [95% CI 0. 816-0.863] and for XGBoost was 0.851 [95% CI 0.827-0.874], which was significantly better than the clinician-guided (AUC-ROC 0.763 [0.734-0.793], p < 0.001) and ML hybrid (AUC-ROC 0.824 [0.800-0.849], p < 0.001) regression models and AWOL-S (AUC-ROC 0.762 [95% CI 0.713-0.812], p < 0.001). Neural Net, XGBoost, and ML hybrid models demonstrated excellent calibration, while calibration of the clinician-guided and AWOL-S models was moderate; they tended to overestimate delirium risk in those already at highest risk.Conclusion:Using pragmatically collected EHR data, two ML models predicted POD in a broad perioperative population with high discrimination. Optimal application of the models would provide automated, real-time delirium risk stratification to improve perioperative management of surgical patients at risk for POD.
# Introduction
Postoperative delirium (POD) is a common and serious complication of surgery, and is associated with numerous adverse events including prolonged length of stay, more frequent institutional discharge, higher readmission rates, functional decline, dependency in activities of daily living, and cognitive decline [bib_ref] Effect of delirium and other major complications on outcomes after elective surgery..., Gleason [/bib_ref] [bib_ref] Outcomes of early delirium diagnosis after general anesthesia in the elderly, Neufeld [/bib_ref] [bib_ref] Delirium after spine surgery in older adults: incidence, risk factors, and outcomes, Brown Ch 4th [/bib_ref] [bib_ref] Delirium: an independent predictor of functional decline after cardiac surgery, Rudolph [/bib_ref] [bib_ref] Outcome and quality of life in patients with postoperative delirium during an..., Abelha [/bib_ref] [bib_ref] Cognitive trajectories after postoperative delirium, Saczynski [/bib_ref] [bib_ref] The short-term and long-term relationship between delirium and cognitive trajectory in older..., Inouye [/bib_ref] [bib_ref] Cognitive decline after delirium in patients undergoing cardiac surgery, Brown Ch 4th [/bib_ref]. Many cases of POD can be prevented with multicomponent non-pharmacologic interventions [bib_ref] Effectiveness of multicomponent nonpharmacological delirium interventions: a metaanalysis, Hshieh [/bib_ref] , the Hospital Elder Life Program [bib_ref] Hospital elder life program: systematic review and meta-analysis of effectiveness, Hshieh [/bib_ref] , or perioperative geriatric consultations [bib_ref] Reducing delirium after hip fracture: a randomized trial, Marcantonio [/bib_ref] [bib_ref] The efficacy of peri-operative interventions to decrease postoperative delirium in non-cardiac surgery:..., Moyce [/bib_ref]. Effective perioperative interventions combining delirium risk stratification with focused delirium prevention care practices have been described [bib_ref] An implementation-effectiveness study of a perioperative delirium prevention initiative for older adults, Donovan [/bib_ref] , though further improvement in the discrimination of the delirium risk stratification tool used in this intervention [bib_ref] Derivation, validation, sustained performance, and clinical impact of an electronic medical record-based..., Whitlock [/bib_ref] could allow for better targeting of finite resources to the patients who need them most. Since preventative interventions require significant time and energy from busy clinicians [bib_ref] American Society for Enhanced Recovery and Perioperative Quality Initiative Joint Consensus Statement..., Hughes [/bib_ref] , improving and automating risk stratification procedures is critically important.
Machine learning-derived risk prediction models have been developed to predict delirium in hospitalized patients [bib_ref] Development and validation of an electronic health record-based machine learning model to..., Wong [/bib_ref] , postoperative delirium in focused patient populations [bib_ref] Exploiting machine learning algorithms and methods for the prediction of agitated delirium..., Mufti [/bib_ref] , non-delirium-related intraoperative complications [bib_ref] Effect of a machine learning-derived early warning system for intraoperative hypotension vs..., Wijnberge [/bib_ref] [bib_ref] Explainable machine-learning predictions for the prevention of hypoxaemia during surgery, Lundberg [/bib_ref] , and postoperative mortality [bib_ref] An automated machine learning-based model predicts postoperative mortality using readily-extractable preoperative electronic..., Hill [/bib_ref] , in addition to applications in many other contexts [bib_ref] Artificial intelligence in anesthesiology: current techniques, clinical applications, and limitations, Hashimoto [/bib_ref]. Machine learning (ML) methodology has the potential to improve upon existing POD risk prediction models in many important ways. Whereas existing delirium prediction models tend to rely on well-known delirium risk factors such as age and cognitive impairment [bib_ref] Performance and agreement of risk stratification instruments for postoperative delirium in persons..., Jansen [/bib_ref] [bib_ref] Risk prediction models for postoperative delirium: a systematic review and meta-analysis, Van Meenen [/bib_ref] [bib_ref] Delirium risk stratification in consecutive unselected admissions to acute medicine: validation of..., Pendlebury [/bib_ref] , ML allows for analysis of patterns in large amounts of data pragmatically collected in the electronic health record (EHR) to identify higher-order interactions that would be difficult to identify through traditional data analysis techniques [bib_ref] Gleaning knowledge from data in the intensive care unit, Pinsky [/bib_ref]. In addition, increasingly feasible realtime EHR-based applications of ML-derived predictions in clinical practice [bib_ref] Artificial intelligence in anesthesiology: current techniques, clinical applications, and limitations, Hashimoto [/bib_ref] [bib_ref] Gleaning knowledge from data in the intensive care unit, Pinsky [/bib_ref] have the potential to conserve valuable human resources through automation of risk stratification procedures, since use of existing risk stratification tools have often required too much clinician input to enter clinical workflow [bib_ref] Risk prediction models for postoperative delirium: a systematic review and meta-analysis, Van Meenen [/bib_ref].
We sought to develop and internally validate a MLderived model for the automated prediction of postoperative delirium in a broad surgical patient population using only pragmatically collected EHR-based data elements available prior to the start of surgery. We compare the performance of two ML models (gradient boosting and artificial neural network) against both traditional logistic regression and the POD risk stratification tool currently used at our institution, hypothesizing that use of ML to model POD risk would outperform other methods.
# Methods
This manuscript was prepared in accordance with the TRIPOD guidelines [bib_ref] Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis..., Collins [/bib_ref]. Approval for a retrospective review of the EHR was obtained by the University of California, San Francisco Institutional Review Board (IRB #18-26,109), and the requirement for written informed consent was waived. The study was conducted in accordance with all requirements outlined by the IRB.
## Study population
This study included all encounters in patients ages 18 and over undergoing surgery or a procedure requiring anesthesia care and staying in the hospital at least overnight at either of two adult hospitals in a non-trauma tertiary care health system between December 2016 and December 2019. All adults were included to allow for application of the model broadly in the perioperative setting. Moffitt-Long Hospital is the health system's largest hospital, housing a wide variety of surgical and procedural specialties including high volume neurosurgery and transplant surgery services. Surgical services at Mission Bay Hospital include various surgical subspecialties, primarily focusing on cancer surgery. Procedures requiring anesthesia recovery outside of the main post-anesthesia care unit were excluded, as were patients who were discharged on the same day as their procedure. Patients who had no POD assessments charted and patients missing data for at least 20% of input variables were also excluded.
## Measures predictors
A total of 115 predictor variables derived solely from preoperative characteristics were given as input to each model (Additional file 1, [fig_ref] Table 1: Participant CharacteristicsAbbreviations [/fig_ref]. Variables included were selected by the authors from those used in a MLderived model developed to predict incident delirium in hospitalized medical patients [bib_ref] Development and validation of an electronic health record-based machine learning model to..., Wong [/bib_ref]. Variables selected were relevant to surgical patients and consistently available in the EHR prior to surgery. Only preoperative variables were included to allow delirium prediction to occur at the start of surgery, so that anesthesia providers and surgeons would be able to immediately adjust the intraoperative and postoperative management strategy. AWOL-S is an EHR-based risk stratification tool in which predefined locally-derived regression coefficients are applied to each of five terms [Age, ability to spell WORLD backward, Orientation to place, American Society of Anesthesiologists Classification (iLlness severity), and procedure-specific Surgical risk] to calculate a predicted risk of POD for an individual patient [bib_ref] Derivation, validation, sustained performance, and clinical impact of an electronic medical record-based..., Whitlock [/bib_ref]. Each of these individual terms was included as a predictor variable for ML models.
## Outcomes
The POD outcome was defined as any instance of Nursing Delirium Screening Scale (NuDESC) [bib_ref] Fast, systematic, and continuous delirium assessment in hospitalized patients: the nursing delirium..., Gaudreau [/bib_ref] score ≥ 2 or positive Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) [bib_ref] Evaluation of delirium in critically ill patients: validation of the Confusion Assessment..., Ely [/bib_ref] recorded during the first seven postoperative days. Patients were assessed by their bedside nurse at least once every 12-h shift using NuDESC on acute care wards or CAM-ICU in the ICU. Bedside nurses were trained in delirium assessment as part of a hospital-wide delirium care program [bib_ref] Outcomes Following Implementation of a Hospital-Wide, Multicomponent Delirium Care Pathway, Lahue [/bib_ref].
## Data collection and preprocessing
All patient data were gathered from the EHR (Epic, Verona, WI) using a unique surgical encounter code.
Missing numeric values were rare and were therefore substituted with the population mean [bib_ref] Missing data and multiple imputation in clinical epidemiological research, Pedersen [/bib_ref] , and missing categorical values were labeled as 'unknown. ' The entire dataset was randomly split into a training dataset (80%) and a test dataset (20%) for model development. All numeric variables in the training dataset were rescaled such that absolute values were contained between 0 and 1, so that numeric variables with higher absolute values would not be weighed inappropriately higher than those with lower absolute values. All categorical variables were converted into indicator variables with values of 0 (no/ absent) or 1 (yes/present). 20% of the training dataset (i.e., 16% of the overall dataset) was reserved for a validation dataset used to tune hyperparameters. The test dataset was left untouched throughout all model development. After each respective model was fully developed, the numeric and categorical variables in the test dataset were normalized based on the same rules determined by the training dataset. The end result was the following: training dataset (15,926 patients), validation dataset (3982 patients), and test dataset (4977 patients).
# Statistical analysis
Descriptive statistics and ML model development were performed using Rand Python, respectively. For numeric variables, differences in mean and standard deviation were tested with the t-test if parametric, and differences in median and interquartile range were tested with the Kruskal-Wallis test if non-parametric. For categorical variables, chi-square was used for comparisons, unless expected cell frequencies were less than 5, in which case Fisher's exact test was used.
## Model development
Each machine learning model approaches classification problems in a unique way. Therefore, individual models have relative strengths and weaknesses in making predictions. Because POD pathophysiology is complex with incompletely understood interactions between risk factors, we selected two ML models to allow us to identify the model that most accurately predicts POD using our rich dataset.
## Gradient boosting
We used the eXtreme Gradient Boosting (XGBoost) algorithm to train a decision tree-based model on the training dataset. XGBoost was chosen for its robustness to overfitting and interpretability of results. The validation dataset was used to fine-tune specific hyperparameters (learning rate = 0.01, maximum tree depth = 4, minimum child weight = 2, number of estimators = 1000, scaled positive weight = 1) with 10-fold cross-validation. Using the feature importance summary plot function of the SHapley Additive exPlanations (SHAP) package, we visualized the 20 most influential prediction variables chosen by XGBoost.
## Neural network
We used TensorFlowto build a neural network consisting of three sequential layers: a densely connected hidden layer, a dropout layer, and a sigmoid-based output layer. Neural network was chosen because of its ability to find complex interactions between variables. The neural network was trained on the training dataset in batch sizes of 1500 cases over 10 epochs. The validation dataset was used to determine proper weights to account for the unbalanced distribution of outcomes.
## Multivariable logistic regression
Traditional logistic regression is one of the most widely used and accepted modeling methods in medicine, and performance of logistic regression has been competitive to that of machine learning in many settings [bib_ref] A systematic review shows no performance benefit of machine learning over logistic..., Christodoulou [/bib_ref] [bib_ref] Logistic regression was as good as machine learning for predicting major chronic..., Nusinovici [/bib_ref]. Using fewer risk features may allow for simpler, more interpretable, point-based models that are easier to implement into clinical practice. Thus, two multivariable logistic regression models were created for comparison to ML-derived models: one based on an existing POD prediction model ("clinician-guided"), and one based on variables selected by ML ("ML hybrid"). We based the clinician-guided regression model on a 20-variable model predicting POD in older patients using data from the multi-institution American College of Surgeons National Surgery Quality Improvement Project (ACS NSQIP) database [bib_ref] Postoperative delirium as a target for surgical quality improvement, Berian [/bib_ref]. Seventeen of the 20 variables were available in our EHR; three (work relative value units, wound class, and surrogate consent) were not. In some cases, a surrogate variable with the same general clinical implication was substituted for the exact variable reported by Berian, et al., when data for the exact variable did not exist in our EHR (e.g., inability to spell WORLD backward was used as a marker for preoperative cognitive impairment).
For the ML hybrid model, 18 out of the 20 predictor variables derived from the feature importance summary of an iteration of the XGBoost algorithm (using the following hyper-parameters in XGBoost model development: learning rate = 0.05, maximum tree depth = 7, minimum child weight = 7, number of estimators = 150, scaled positive weight = 10) were included. Four variables which had considerable overlap were combined into two variables without changing the clinical implication [i.e., inpatient (risk) and outpatient (protective) patient status, unable (risk) and able (protective) to spell 'WORLD' backwards].
## Comparison to awol-s
The ML-derived models were also compared to AWOL-S, the POD risk stratification tool used in our institution. Risk stratification with AWOL-S is performed preoperatively for all adult surgical patients; a calculated probability of delirium of 5% or greater is considered "high risk. " For those terms not already in the EHR (i.e., WORLD backward and orientation to place), assessments are performed and documented by preoperative nurses. Sensitivity, specificity, positive likelihood ratio, and positive predictive value of AWOL-S were calculated at a predicted POD risk (i.e., threshold or cutoff value) of 5%.
## Model evaluation
Model performance was evaluated based on the area under the receiver operating characteristic curve (AUC-ROC), and confidence intervals (CI) were derived from 10-fold cross validation (CV) and DeLong's method (DL) [bib_ref] Comparing the areas under two or more correlated receiver operating characteristic curves:..., Delong [/bib_ref] [bib_ref] Fast implementation of DeLong's algorithm for comparing the areas under correlated receiver..., Sun [/bib_ref]. For each ML model, an optimal decision threshold, defined as the threshold at which the sum of sensitivity and (1-specificity) is greatest, was determined on the validation dataset for subsequent calculation of model sensitivity, specificity, positive likelihood ratio, positive predictive value, and negative predictive value. Because AWOL-S was previously validated on a separate dataset, cross validation was not applicable, and performance was evaluated on AUC-ROC with confidence intervals derived by DeLong's method. A calibration plot was generated for each model.
# Results
Twenty-nine thousand four surgical encounters were evaluated. After exclusion of 4119 encounters (1965 patients with no delirium score ever recorded and 2154 patients with > 20% missing variables), 24,885 surgical encounters were included in the analysis [fig_ref] Figure 1: Inclusion flow diagram [/fig_ref]. 77,125/81,515 (94.6%) of total delirium assessments were performed using NuDESC, while the remainder were performed using CAM-ICU. The overall incidence of delirium was 5.3%. 325/4390 (7.4%) screens performed using CAM-ICU were positive, while 1373/77,125 (1.8%) screens performed using NuDesc were positive. Patients who developed delirium were older, more likely male, and had more comorbidities and a higher American Society of Anesthesiologists Classification [fig_ref] Table 1: Participant CharacteristicsAbbreviations [/fig_ref]. Patients developing delirium were also more likely to have undergone inpatient surgery, emergency surgery, and/or neurological surgery.
In the test dataset, the AUC-ROC was 0.840 (95% CI 0.825-0.855 by CV) and 0.841 (95% CI 0.816-0.863) by DL for Neural Network [fig_ref] Table 2: Model CharacteristicsAbbreviations [/fig_ref] , [fig_ref] Figure 2: Model AUC-ROC curves and calibration plots [/fig_ref]. AUC-ROC was 0.852 (95% CI 0.839-0.865) by CV and 0.851 (95% CI 0.827-0.874) by DL for XGBoost [fig_ref] Table 2: Model CharacteristicsAbbreviations [/fig_ref] , [fig_ref] Figure 2: Model AUC-ROC curves and calibration plots [/fig_ref]. Under the optimal threshold, Neural Network achieved a mean sensitivity of 72.9% (95% CI 69.1-76.7%), mean specificity of 77.5% (95% CI 76.2-78.7%), and mean positive likelihood ratio of 3.25 (95% CI 3.03-3.47). XGBoost achieved a mean sensitivity of 80.6% (95% CI 77.1-84.1%), mean specificity of 73.7% (95% CI 72.4-74.9%), and mean positive likelihood ratio of 3.08 (95% CI 2.87-3.29). Additional model characteristics and thresholds used for calculation of reported performance metrics are shown in [fig_ref] Table 2: Model CharacteristicsAbbreviations [/fig_ref]. The 20 variables (out of the total pool of 115 variables) with the highest impact on XGBoost outcome prediction are pictured in . and C demonstrate the XGBoost algorithm's decision path for two individual patients. A calibration plot for each of the models is pictured in [fig_ref] Figure 2: Model AUC-ROC curves and calibration plots [/fig_ref]. Neural Network, XGBoost, and ML hybrid models demonstrated excellent calibration, while calibration of the clinicianguided and AWOL-S models was moderate; they tended to overestimate delirium risk in those already at highest risk. We performed two sensitivity analyses: one including patients age 65 and older, and a second to further examine the impact of including neurosurgery patients in the broad population model on delirium prediction. In both patient subgroups, AUC-ROC of the XGBoost model did not change significantly (65+ AUC-ROC [fig_ref] Figure 2: Model AUC-ROC curves and calibration plots [/fig_ref]. Coefficients and odds ratios from the two regressions are available in Additional file 1, Tables S3 and S4. AWOL-S had an AUC-ROC of 0.762 (95% CI 0.713-0.812 by DL), mean sensitivity of 78.2% (95% CI 66.0-89.3%), mean specificity of 60.0% (95% CI 57.0-63.0%), and mean positive likelihood ratio of 1.95 (95% CI 1.62-2.28 [fig_ref] Table 2: Model CharacteristicsAbbreviations [/fig_ref] , [fig_ref] Figure 2: Model AUC-ROC curves and calibration plots [/fig_ref].
# Discussion
We developed and internally validated two ML-derived risk prediction models which used preoperative data available in the EHR prior to the start of surgery to predict incident postoperative delirium in a broad surgical patient population. ML models offer better performance than traditional clinician-based regression models, both in this population and by comparison to published literature [bib_ref] Prediction of acute kidney injury after liver transplantation: machine learning approaches vs...., Lee [/bib_ref] [bib_ref] Multicenter comparison of machine learning methods and conventional regression for predicting clinical..., Churpek [/bib_ref] [bib_ref] Machine learning and neurosurgical outcome prediction: a systematic review, Senders [/bib_ref] , implying that they could be used to more efficiently direct resources to patients at high risk compared with existing models [bib_ref] Derivation, validation, sustained performance, and clinical impact of an electronic medical record-based..., Whitlock [/bib_ref] and with a clinicianguided logistic regression model derived on the same data. Further, a hybrid approach, using ML to select variables which were then input into a multivariable logistic regression, performed better than the purely clinicianguided approach.
Both ML-derived models achieved high AUC-ROCs (0.841 for Neural Net and 0.851 for XGBoost), similar to other published ML-derived risk prediction models [bib_ref] Development and validation of an electronic health record-based machine learning model to..., Wong [/bib_ref] [bib_ref] Development and validation of machine learning models to identify high-risk surgical patients..., Corey [/bib_ref] , and better than many risk prediction models specific for postoperative delirium [bib_ref] Performance and agreement of risk stratification instruments for postoperative delirium in persons..., Jansen [/bib_ref] [bib_ref] Risk prediction models for postoperative delirium: a systematic review and meta-analysis, Van Meenen [/bib_ref]. As opposed to most previously reported POD risk prediction models that focus on one particular surgical population [bib_ref] Do risk prediction models for postoperative delirium consider patients' preoperative medication use?, Kassie [/bib_ref] , this model attempts to predict delirium in a broad surgical population over a wide age range. The inclusion of such a broad patient population was intentional, to make this a pragmatic tool for implementation in the perioperative setting. When compared to a simplified regression approach in an overlapping nonspecific perioperative population from our institution, ML models offer substantial improvement in discrimination [bib_ref] Derivation, validation, sustained performance, and clinical impact of an electronic medical record-based..., Whitlock [/bib_ref]. Performance did not change significantly when models were rerun in patients over the age of 65 only, or when neurosurgical patients were excluded, suggesting the models are robust to different specifications of the underlying derivation population. Providing optimal care to patients at risk for delirium is a resource-intensive process which includes implementing measures such as multicomponent non-pharmacologic nursing care bundles and consultations from busy clinicians including pharmacists and rehabilitation professionals, whose time is a limited resource. A schematic depicting how this screening tool would be used in our healthcare system is pictured in , which also highlights the extensive resources automatically triggered by a high-risk delirium screen in our institution's existing delirium prevention pathway [bib_ref] An implementation-effectiveness study of a perioperative delirium prevention initiative for older adults, Donovan [/bib_ref] [bib_ref] Postoperative delirium: why, what, and how to confront it at your institution, Curtis [/bib_ref]. The care interventions were modeled after available guidelines for the prevention of postoperative delirium [bib_ref] American Society for Enhanced Recovery and Perioperative Quality Initiative Joint Consensus Statement..., Hughes [/bib_ref] [bib_ref] European Society of Anaesthesiology evidence-based and consensus-based guideline on postoperative delirium, Aldecoa [/bib_ref] [bib_ref] Optimal perioperative management of the geriatric patient: a best practices guideline from..., Mohanty [/bib_ref] [bib_ref] Best practices for postoperative brain health: recommendations from the fifth international perioperative..., Berger [/bib_ref]. Recommendations common to nearly all these guidelines include performing preoperative cognitive screens and avoidance of high-risk medications. Postoperative components of our pathway, including use of multicomponent bundles, treatment of underlying causes, early mobility, and medication review, are recommended by those guidelines addressing care in the postoperative period [bib_ref] American Society for Enhanced Recovery and Perioperative Quality Initiative Joint Consensus Statement..., Hughes [/bib_ref] [bib_ref] European Society of Anaesthesiology evidence-based and consensus-based guideline on postoperative delirium, Aldecoa [/bib_ref] [bib_ref] Optimal perioperative management of the geriatric patient: a best practices guideline from..., Mohanty [/bib_ref]. Identification of the patients at highest risk of developing delirium is the critical step to patient entry into the pathway. Thus, by improving the performance of delirium screening models, we could not only better target these resources to benefit the highest risk patients, but also potentially improve healthcare value. Next steps needed to operationalize the model would include prospective and external validation of the model followed by integration of the model into the EHR for real-time use; such real-time applications of ML models have been previously described [bib_ref] Effect of a machine learning-derived early warning system for intraoperative hypotension vs..., Wijnberge [/bib_ref] , demonstrating potential feasibility of this approach.
Reporting model discrimination as the sole evaluation metric of a model's performance is a commonly cited weakness of ML studies [bib_ref] Opportunities and challenges in developing risk prediction models with electronic health records..., Goldstein [/bib_ref] , since other measures of performance which take prevalence into account may be more indicative of clinical applicability and importance. Both of our ML models provide similar discrimination, but we also report other model characteristics (i.e., sensitivity, positive likelihood ratio, and positive predictive value) to address the potential for clinical applicability of our model. XGBoost and Neural Network both have high positive likelihood ratios, which would allow resourceintensive interventions to be directed to patients at high risk for developing delirium. We selected thresholds to optimize sensitivity and positive likelihood ratio, rather than positive predictive value, since positive predictive value is more subject to disease prevalence, and the overall prevalence of delirium is small in our population. The low prevalence of delirium compromised the positive predictive value somewhat. Further, we took steps to maximize interpretability of the XGBoost model by using the open-source SHAP package to visualize the global predictors that were deemed important by the algorithm as well as its individual, patient-centered decision approach to delirium prediction. These types of visualizations are helpful to increase clinician confidence in ML-derived predictions if they are to be used to augment clinical decision making [bib_ref] Machine learning for clinical decision support in infectious diseases: a narrative review..., Peiffer-Smadja [/bib_ref]. The 20 most influential features identified by the XGBoost model generally align well with well-known risk factors for delirium, as discussed further below, lending credibility to the model's prediction.
Not surprisingly, both ML models perform better than a traditional logistic regression analysis using delirium risk predictors selected from the 20-predictor model derived from multivariable analysis of the multi-institutional ACS-NSQIP dataset [bib_ref] Postoperative delirium as a target for surgical quality improvement, Berian [/bib_ref]. The ML hybrid model using the predictors derived from the XGBoost feature importance summary also outperforms the expert clinician regression model. This result suggests that ML-derived models are capable of uncovering higher-order interactions between variables that are difficult to identify using traditional regression approaches. Most of the predictors recovered by XGBoost (e.g., age, surgery type, cognitive impairment, comorbidities) are consistent with known predictors of POD [bib_ref] Postoperative delirium as a target for surgical quality improvement, Berian [/bib_ref] [bib_ref] Preoperative risk assessment for delirium after noncardiac surgery: a systematic review, Dasgupta [/bib_ref] [bib_ref] Postoperative delirium in the elderly: risk factors and outcomes, Robinson [/bib_ref]. Protective factors against delirium not frequently described but important to the model include having private insurance and a self-reported history of alcohol use, and a self-reported history of recent falls. It is likely that such features reflect the existence of associations which are not otherwise accounted for by the model. For example, those individuals with private insurance may have a higher socioeconomic status and/or education level. Patients capable of self-reporting falls may have better cognitive function when compared with those who are unable to do so. While alcohol consumption is a well-known risk factor for delirium [bib_ref] Delirium in elderly people, Inouye [/bib_ref] , self-reported alcohol consumption has been associated with better functional outcomes including lower frailty in females [bib_ref] Association between lifestyle behaviors and frailty in Atlantic Canadian males and females, Declercq [/bib_ref] and lower likelihood of mobility limitation or arm function limitation independent of muscle strength in older men [bib_ref] Alcohol consumption and functional limitations in older men: does muscle strength mediate..., Wang [/bib_ref]. Other authors have suggested that the protective effect of moderate alcohol use is explained by social or lifestyle factors [bib_ref] Alcohol consumption and functional limitations in older men: does muscle strength mediate..., Wang [/bib_ref] ; these may not be captured by other terms in the ML model. However, the exact mechanisms underlying these associations are unknown, and will need to be confirmed by further studies.
There are important limitations to consider when interpreting our findings. First, NuDESC ≥2 was used as the definition of delirium for patients on acute care wards. NuDESC is a delirium screening (not diagnostic) tool that was originally developed in a medical patient population [bib_ref] Fast, systematic, and continuous delirium assessment in hospitalized patients: the nursing delirium..., Gaudreau [/bib_ref]. It has been reported to have a wide sensitivity range to detect delirium in older surgical patients on the wards and in the post-anesthesia care unit, but specificity in surgical patients is reported to be 80% or greater [bib_ref] Evaluation of two delirium screening tools for detecting postoperative delirium in the..., Neufeld [/bib_ref] [bib_ref] A comparison of three scores to screen for delirium on the surgical..., Radtke [/bib_ref]. The high specificity of NuDESC in surgical patients suggests that the more severe, most clinically relevant cases, are detected. Despite reasonable concern for both false negative (i.e., undercounting the delirium outcome due to low sensitivity) and false positive (i.e., patients screening positive with NuDESC would not have delirium when formally assessed) results, it was necessary to take a pragmatic approach using the delirium screening procedures in place in our institution, due to the large training dataset needed to conduct this machine learning analysis. Comparison of our study to a recently published study by Racine, et al. [bib_ref] Machine learning to develop and internally validate a predictive model for post-operative..., Racine [/bib_ref] illustrates this point. This study evaluated the performance of five ML algorithms to predict POD in a much smaller group of surgical patients (560 older adults from an existing dataset). ML algorithms had a reported AUC-ROC 0.53-0.71, which was not superior to the logistic regression model reported in the study, despite using an in-person examination by experienced interviewers with the Confusion Assessment Method [bib_ref] Clarifying confusion: the confusion assessment method. A new method for detection of..., Inouye [/bib_ref] and medical chart review to detect delirium. This comparison highlights the importance of an adequate training dataset size [bib_ref] Predicting sample size required for classification performance, Figueroa [/bib_ref] , among other things, to conduct high-quality machine learning analyses. Even with the large sample size in our study, discrimination and other aspects of model performance such as positive predictive value and calibration would likely be further improved with even more data and a larger number of events per predictor because our outcomes occur rarely [bib_ref] Modern modelling techniques are data hungry: a simulation study for predicting dichotomous..., Van Der Ploeg [/bib_ref].
Additional limitations include the lower POD rate (5.3%) in our population, which likely reflects inclusion of a younger population and all types of surgeries including those known to be associated with lower risk of delirium (i.e., gynecologic, urologic, plastics), in addition to robust delirium prevention procedures in our hospital system. When the analysis is excluded to patient age 65 and over the POD rate increases to 7.7%, which remains lower than commonly reported POD rates, likely still reflecting inclusion of patients having lower-risk surgery and staying in the hospital only overnight in our study population. Patients with no POD assessments were excluded from the analysis; it is possible that missingness is nonrandom for these patients. The fact that this was a single center study may introduce bias and limit generalizability. We conducted a sensitivity analysis to determine whether the high prevalence of neurosurgical patients in our population may have influenced the model, as this population tends to be at high risk of delirium despite younger age and fewer medical comorbidities. We found that the discrimination of the machine learning models was not significantly affected by exclusion of the neurosurgical population, suggesting that the model is able to conclude on delirium risk based on other risk features. In addition, there were limitations to gathering certain types of data from the EHR at our institution. Laboratory data is a good example; many of our patients either have no preoperative laboratory data available, or we are unable to easily extract this data from the EHR because results are housed in a scanned text report from an outside facility. There may also be potentially important delirium risk predictors (e.g., frailty indices) that are not included in our model because they are not routinely part of the preoperative examination at our institution. Expansion of the terms included in the ML model and external validation on a multicenter dataset would help to address these shortcomings.
# Conclusion
We developed and internally validated two ML-derived models that predict POD in a broad perioperative population using pragmatically collected EHR data. The XGBoost model offers the ability to understand the most impactful predictors and the process by which the algorithm arrives at a prediction for an individual patient. A ML-hybrid approach outperformed a clinician-guided logistic regression, suggesting that ML has the potential to uncover POD predictors that were previously overlooked by clinicians. As real-time clinical implementation of ML models becomes increasingly feasible, POD prediction using ML --allowing targeted resource direction toward patients at the highest risk --may be an important focus for improving perioperative care.
Abbreviations POD: Postoperative delirium; ML: Machine learning; EHR: Electronic health record; XGBoost: EXtreme Gradient Boosting; AWOL-S: Delirium risk stratification tool using Age, ability to spell 'World' backward, Orientation to place, iLlness severity score, and Surgical risk; AUC-ROC: Area under the Receiver Operating Characteristic Curve; NuDESC: Nursing Delirium Screening Scale; CAM-ICU: Confusion Assessment Method for the Intensive Care Unit; ACS-NSQIP: American College of Surgeons National Surgical Quality Improvement Program; CI: Confidence interval; CV: 10-fold cross validation; DL: DeLong's method.
## Supplementary information
The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s12871-021-01543-y.
## Additional file 1 : supplementary
[fig] Figure 1: Inclusion flow diagram [/fig]
[fig] Figure 2: Model AUC-ROC curves and calibration plots. A Receiver Operating Characteristic curves for five POD prediction models. B Calibration plots for five POD prediction models. XGBoost (orange), Neural network (blue), Clinician-guided regression (green), ML hybrid regression (red), AWOL-S (purple)Fig. 3 Visualization of decisions made by the XGBoost algorithm. A Top 20 most influential variables used by XGBoost. Interpretation: Each dot represents a variable for an individual patient instance. Variables pictured to the right side of the y-axis influenced the model to predict delirium, whereas variables to the left of the y-axis influenced the model against prediction of delirium. Red signifies a higher absolute value (numeric variables) or yes/present (categorical variables), and blue signifies a lower absolute value (numeric variables) or no/absent (categorical variables). For example: Higher age (red color) influenced the model to predict delirium (right of y-axis), whereas lower age (blue color) influenced the model toward prediction of no delirium. Decision path for a true negative (B) and a true positive (C) delirium prediction by XGBoost for two individual patients. Interpretation: The algorithm begins at the center of the x-axis with a baseline value. The model considers each variable along the y-axis one at a time (values shown in parenthesis), to influence the model toward making a positive (vertical line moves toward the right) or negative (vertical line moves toward the left) delirium risk prediction. For example: In panel B, variables which significantly influenced the model toward a negative delirium prediction include outpatient surgery, ASA class 1, low fall risk, not neurosurgery, and short case length. In panel C, variables which significantly influenced the model toward a positive delirium prediction include not oriented to place, older age, ASA class 4, unable to rate pain using numeric assessment scale, high pressure ulcer risk, unable to spell 'WORLD' backwards, and high fall risk. Abbreviations: ASA, American Society of Anesthesiologists; ICD-10, International Classification of Diseases, 10th revision; ICD-10 F00-F99, mental and behavioral disorders; kg, kilograms; ICD-10 Z00-Z99, factors influencing health status and contact with health services; ICD-10 G00-G99, diseases of the nervous system; ERAS, enhanced recovery after surgery; ICD-10 C00-D48, neoplasms; ICD-10 I00-I99, diseases of the circulatory system; ICD-10 N00-N99, disease of the genitourinary systemFig. 4 Role of automated delirium screen in our institution's postoperative delirium prevention care pathway. Figure legend: A high-risk delirium screen triggers a set of care modifications in the preoperative, intraoperative, and postoperative phases of care. Particularly in the postoperative phase, these modifications require time and input from busy clinicians [/fig]
[fig] Additional file 2: Sensitivity Analyses. 1. Sensitivity Analysis in Patients Age 65 and Over. Supplementary Figure S1.1.: Inclusion Flow Diagram. Supplementary Table S1.1: Baseline Demographics. Supplementary Figure S1.2: Receiver Operating Characteristic (ROC) Curve For 5 Models. Supplementary Table S1.2: Comparison of Model Characteristics. Supplementary Figure S1.3: Feature Importance Summary of XGBoost Model.Supplementary [/fig]
[table] Table 1: Participant CharacteristicsAbbreviations: SD standard deviation, ASA American Society of Anesthesiologists a Nine surgical services with the highest patient volume (out of 19 total services) are listed b Three language categories with the largest number of patients (out of 8 total categories) are listed [/table]
[table] Table 2: Model CharacteristicsAbbreviations: CI confidence interval, CV cross validation, DL DeLong's method a AWOL-S is pre-validated, therefore cross validation was not performed to derive confidence intervals [/table]
[table] Table S1: List of variables included in the machine learning models. Supplementary Table S2: Comparison of the most important variables chosen by XGBoost and Neural Network. Supplementary Table S3: Multivariable logistic regression using variables selected by expert clinicians. Supplementary Table S4: Multivariable logistic regression using variables chosen by the XGBoost algorithm. [/table]
|
Gastrointestinal Tolerance and Glycemic Response of Isomaltooligosaccharides in Healthy Adults
Ingredients delivering functional and nutritional benefits are of interest to food manufacturers. Isomaltooligosaccharides (IMOs) which serve as alternate sweeteners fit into this category. IMOs are a mixture of α-(1 → 6) and α-(1 → 4)-linked glucose oligomers, synthesized by an enzymatic reaction from starch (corn, tapioca). The aim of this study was to evaluate the fermentability and glycemic response of IMO in a healthy population. Two randomized, double-blind, placebo-controlled, cross-over human studies were conducted. In the first study (n = 26), participants' breath hydrogen over 24 h, gastrointestinal tolerance, and glycemic and insulinemic response to BIOLIGO TM IL5040 isomaltooligosaccharide were measured. In another study (n = 10), participants' two-hour post-prandial glycemic response to BIOLIGO TM IL5040 isomaltooligosaccharide and BIOLIGO TM IL7010 isomaltooligosaccharide was measured compared to dextrose (control). The IMOs differed in the composition of mono and di-saccharide sugars. IMO syrup dose was matched for 50 g of total carbohydrates and was consumed by mixing in water (237 mL/8 oz.). Mean composite gastrointestinal score was not significantly different (p = 0.322) between the control (1.42) and IMO (1.38). Lack of difference in glycemic response (p = 0.662), with no impact on breath hydrogen (24 h; p = 0.319) and intestinal tolerance, demonstrates that IMO is digestible and can be used to replace sugars in product formulations.
# Introduction
Global forecasts on obesity and diabetes prevalence are among the top issues of concern, and they have led to increased scrutiny of sugar-sweetened products [bib_ref] Obesity and severe obesity forecasts through 2030, Finkelstein [/bib_ref]. The World Health Organization (WHO) recommends limiting sugar intake to 10% of total energy intake, equivalent to 12 teaspoons or 50 g/day . This has been adopted by various nations, with public policies restricting added sugars through taxes on sugar beverages and warning labels. More recently, the US Food and Drug Administration (FDA) mandated the inclusion of "added sugars" on the nutrition facts labeling. Although there is no consistent evidence that added sugars cause weight gain leading to obesity in children and adults, consumer surveys point to the heightened market need for sugar alternatives in full-calorie products [bib_ref] Are restrictive guidelines for added sugars science based?, Erickson [/bib_ref] [bib_ref] What is the appropriate upper limit for added sugars consumption?, Rippe [/bib_ref].
One such alternative is sugar alcohols (polyols). Polyols provide sweet taste while yielding a low glycemic index, are non-cariogenic, and thus can be used as sugar replacers [bib_ref] Health potential of polyols as sugar replacers, with emphasis on low glycaemic..., Livesey [/bib_ref]. However, polyols use at higher levels is limited due to their effects on gastrointestinal intolerance in healthy subjects and patients with irritable bowel syndrome (IBS) [bib_ref] A digestive tolerance study of maltitol after occasional and regular consumption in..., Ruskone-Fourmestraux [/bib_ref] [bib_ref] A systematic review of the effects of polyols on gastrointestinal health and..., Lenhart [/bib_ref]. Maltitol, at higher levels, caused intestinal discomfort by inducing osmatic pressure, but when combined with short-chain fructooligosaccharide, a prebiotic fiber, has been shown to attenuate the intestinal symptoms in healthy adults [bib_ref] Digestive tolerance and postprandial glycaemic and insulinaemic responses after consumption of dairy..., Respondek [/bib_ref]. More commonly used alternate sweeteners in Asia are fructooligosaccharides (FOS) and isomaltooligosaccharides (IMO). Oligosaccharides are appealing as alternate sweeteners due to their improved functionality and nutritional benefits [bib_ref] Will isomalto-oligosaccharides, a well-established functional food in Asia, break through the European..., Goffin [/bib_ref]. Fructooligosaccharides are unique in a way providing both nutritional (fiber) and functional (sweetness) benefits when formulated in food and beverages. However, the use of FOS has few limitations with regard to its effect on digestive intolerance at higher doses and lack of stability in highly acidic products. The physicochemical properties of isomaltooligosaccharides make them highly functional as sweeteners, due to their higher stability in food and beverage formulations, compared to FOS [bib_ref] Physical and Physiological Properties of isomaltooligosaccharides and frucooligosaccharides, Kim [/bib_ref].
Isomaltooligosaccharides are mixtures of α-(1 → 6)-linked glucose oligomers with degrees of polymerization from 2-10, and their carbohydrate composition includes isomaltose, panose, isomaltotriose, isomaltotetraose, isomaltopentaose and so forth. IMOs are functional sweeteners and commonly derived from the enzymatic processing of starch (transglycosylation of hydrolyzed starch from corn and tapioca being the most commonly used bases) on a commercial scale. Its sweetness depends on the composition, specifically on the amount of lower molecular weight components, such as glucose (about 70-75% as sweet as sucrose) and maltose (about 30-35% as sweet as sucrose). IMOs can also be produced via bacterial fermentation of sucrose in the presence of a maltose acceptor by a glucosyltransferase enzyme, such as dextransucrase [bib_ref] A survey of commercially available isomaltooligosaccharide-based food ingredients, Madsen [/bib_ref]. In addition to α-(1 → 6)-linked glucose oligomers, commercial IMO products also contain some level of α-(1 → 4)-linked glucose oligomers, including maltose, maltotriose, etc. [bib_ref] Continuous production of Isomaltooligosaccharides from maltose syrup by immobilized cells of permeabilized..., Yun [/bib_ref]. Rarely, it is possible to find minor amounts of α-(1 → 2) and α-(1 → 3)-linked kojibiose and nigerose in the products as well. IMOs naturally exist in honey, and fermented foods, such as soy sauce, miso, and sake.
Although IMOs are promoted as prebiotic fiber in Asia, there is conflicting evidence on their digestibility with high caloric value as shown in rat and human studies [bib_ref] Digestibility characteristics of isomalto-oligosaccharides in comparison with several saccharides using rat jejunum..., Kaneko [/bib_ref] [bib_ref] Metabolism of 13C-isomaltooligosaccharides in healthy men, Kohmoto [/bib_ref]. Madsen et al. [bib_ref] A survey of commercially available isomaltooligosaccharide-based food ingredients, Madsen [/bib_ref] have surveyed a number of commercially available IMO products in the US, and their results indicated that the digestibility and potential glycemic impact of these ingredients were inconsistent with product labels, including soluble fiber content and glycemic response. Our aim was to assess gastrointestinal tolerance and glycemic response of two IMOs, BIOLIGO TM IL5040 and BIOLIGO TM IL7010, in healthy individuals. In the first study, breath hydrogen response, gastrointestinal tolerance, glycemic and insulinemic response to BIOLIGO TM IL5040 IMO was tested. In the second study, two-hour glycemic response to BIOLIGO TM IL5040 and BIOLIGO TM IL7010 IMOs was evaluated.
# Materials and methods
Two IMO syrups manufactured by Ingredion Incorporated (Bridgewater, NJ, USA) were used in these studies. The first product BIOLIGO™ IL5040 contains 50% IMO, and 40% mono-and disaccharides, consisting of mainly dextrose and maltose, and a small amount of isomaltose. The second product BIOLIGO™ IL7010 is a reduced mono-and disaccharide version of BIOLIGO™ IL5040, and contains 70% IMO, and less than 10% mono-and disaccharides. IMO content for these products is defined as the IMO components from DP2 to DP7 (isomaltose, panose, isomaltotriose, isomaltotetraose, isomaltopentaose, isomaltohexaose, isomaltoheptaose).
The studies were conducted in accordance with the ethical principles outlined in the Declaration of Helsinki Inclusion criteria: Healthy men or women aged 18 to 75 years, body mass index (BMI) 18.50-29.99 kg/m 2 , normally active and judged to be in good health on the basis of their medical histories, were enrolled in the study.
Exclusion criteria: Subjects were excluded if they had fasting capillary glucose ≥100 mg/dL at screening; major trauma or a surgical event within 3 months of screening; history of drug or alcohol abuse; have body weight change ≥4.5 kg in the 2 months prior to screening; uncontrolled hypertension; use of antibiotics; symptoms of an active infection; intolerance to any ingredients in the study products; extreme dietary habits; cannot abstain from consuming probiotics; alcohol; smoking, and who are unwilling to comply with the experimental procedures.
## Study 2
Inclusion criteria: Subjects were males or non-pregnant females aged 18-75 years and in good health.
Exclusion criteria: Subjects less than 18 years old or older than 75 years, with a known history of AIDS, hepatitis, diabetes or a heart condition, and unwillingness or inability to comply with the experimental procedures, and to follow GI Labs safety guidelines.
## Study design and test products
Study 1: The study was a randomized, double-blinded, placebo-controlled, cross-over design, with 26 healthy adults (age 18-75 years, body mass index (BMI) 18.50-29.99 kg/m 2 ). Eligible participants were tested on separate days and were assigned randomly to either BIOLIGO TM IL5040 IMO (test, 68.46 g), or dextrose (CERELOSE ® Dextrose, Ingredion Incorporated, Bridgewater, NJ, USA) (control; 54.77 g) mixed in 237 mL (8 oz) of water. Both the test and control were matched for 50 g total carbohydrates. The interval between two testing periods was one week. Of 31 randomized subjects, five did not meet inclusion/exclusion criteria and 26 participants completed the study.
Study 2: In a double-blind, randomized crossover design, eligible participants (n = 10) were studied on three separate days, over a period of 2 to 3 weeks with an interval of no less than one day between tests. A total of 10 subjects were randomly assigned to either BIOLIGO TM IL5040 IMO (66.3 g); BIOLIGO TM IL7010 IMO (66.0 g), or dextrose (Clintose ® Dextrose, ADM, Chicago, IL, USA) (control; 54.9 g) mixed in 250 mL of water. The test ingredients and control were matched for 50 g of total carbohydrates.
The doses of the IMOs were based on the batch certificate of analysis, and slight differences in BIOLIGO TM IL5040 IMO doses between the two studies are due to differences in water content.
## Study visit procedures
Study 1: On the test day, all participants arrived at the clinic after an overnight fasting. The initial breath hydrogen, fasting blood glucose and insulin were measured and participants were provided either IMO or dextrose (control) mixed in water. Blood samples were obtained for serum glucose and insulin measurements via an indwelling venous catheter or venipuncture at t = −15, 15, 30, 45, 60, 90, 120, 150, 180, 210, and 240, where t = 0 min was the start of study product consumption. Carbohydrate malabsorption and fermentation were measured by assessing each subject's end-alveolar (breath) hydrogen concentrations on test days, following the t = −15 (pre-dose), 60, 120, 180, and 240 min blood collections. Additionally, a breath sample was collected in the clinic at t = 24 h, and subjects collected breath samples at home at t = 8 and 12 h. During both test days, a low-fiber, very low-dairy standardized lunch was administered in the clinic on the study days, and a low-fiber, very low-dairy dinner, and evening snack were dispensed to be consumed at home that evening.
A gastrointestinal tolerability (GI) questionnaire [bib_ref] Fibermalt is well tolerated in healthy men and women at intakes up..., Maki [/bib_ref] was administered immediately after the test visits, to assess the presence and severity of selected GI symptoms including nausea, GI rumblings, abdominal pain, bloating, flatulence, and diarrhea over the past 24-h period. GI symptoms were scored as follows: 0 = none, 1 = no more than usual, 2 = somewhat more than usual, and 3 = much more than usual. A composite score was also calculated as the sum of the six individual GI symptom ratings, for a total possible score of 0-12.
Study 2: On each test occasion, after subjects were weighed, two fasting blood samples were obtained by finger-prick at 5-min intervals. Following consumption of either IMO or dextrose (control) in water, blood samples were collected at 15, 30, 45, 60, 90, and 120 min. Study 2: Blood glucose analysis was done using a YSI (Yellow Spring Instruments, Yellow Springs, OH, USA) analyzer, and took place within five days of collection. The YSI uses a wet method for glucose analysis based on the reaction of glucose in the sample with immobilized glucose oxidase. The typical analytical CV for fasting glucose is <2%.
# Biochemical analysis
## Data analysis and statistics (sample size, data analysis and statistical analysis)
Study 1: A power analysis indicated that sample size of 26 subjects would be required to detect a 0.58 standard deviation difference between treatments for continuous outcome variables with 80% power, alpha = 0.05, 2-sided. A total of 31 subjects were randomized to allow for subject attrition.
Study 2: Using the t-distribution, and assuming an average CV of within-individual variation of incremental area under the curve (iAUC) values of 25%, n = 10 subjects has 80% power to detect a 33% difference in incremental AUC with 2 tailed p < 0.05. Paired t-tests were conducted on blood glucose, insulin, and values at individual and incremental area-under-curveusing GraphPad Prism 7 (v7.03, GraphPad Software, Inc., La Jolla, CA, USA). p values ≤ 0.05 were deemed statistically significant, and data are presented as mean ± SEM.
# Results
Subject demographics for both studies are shown in Tables 1 and 2. All participants were healthy in both studies.
Composite gastrointestinal tolerance in response to IMO and dextrose (control) is shown in [fig_ref] Study 1: Glucose was measured using an enzymatic colorimetric method-GOD/PAP Method [/fig_ref]. There was no significant difference (p = 0.322) between the control (1.42) and IMO (1.38) in the mean composite score on the GI tolerability questionnaire. Similar findings were observed for individual gastrointestinal frequency of scores of 2 or greater (somewhat more than usual and much more than usual) on the components of the GI tolerability questionnaire as shown in [fig_ref] Table 3: Frequency of scores ≥2 a on individual gastrointestinal [/fig_ref]. Composite gastrointestinal tolerance in response to IMO and dextrose (control) is shown in [fig_ref] Study 1: Glucose was measured using an enzymatic colorimetric method-GOD/PAP Method [/fig_ref]. There was no significant difference (p = 0.322) between the control (1.42) and IMO (1.38) in the mean composite score on the GI tolerability questionnaire. Similar findings were observed for individual gastrointestinal frequency of scores of 2 or greater (somewhat more than usual and much more than usual) on the components of the GI tolerability questionnaire as shown in [fig_ref] Table 3: Frequency of scores ≥2 a on individual gastrointestinal [/fig_ref]. 0.23 a Scoring system was 0 = none, 1 = no more than usual, 2 = somewhat more than usual, and 3 = much more than usual. b p-values derived from repeated measures analysis using the GEE method with subjects included as a random effect. IMO: isomaltooligosaccharides.
GI symptoms (nausea, bloating, GI rumblings, flatulence, abdominal pain, diarrhea) were scored as follows: 0 = none, 1 = no more than usual, 2 = somewhat more than usual and 3 = much more than usual. The composite score was the sum of the six individual GI symptom ratings. Data are mean ± SEM.
The changes in breath hydrogen concentration in response to dextrose and BIOLIGO TM IL5040 IMO, over 24 h are shown in [fig_ref] Figure 2: Breath hydrogen [/fig_ref]. There was no significant difference between IMO and the control at all time points. Post-prandial glycemic [fig_ref] Figure 3: Post-prandial glycemic response in healthy adults [/fig_ref] and insulinemic response [fig_ref] Figure 3: Post-prandial glycemic response in healthy adults [/fig_ref] to dextrose and BIOLIGO TM IL5040 IMO, showed no significant differences over 4 h. GI symptoms (nausea, bloating, GI rumblings, flatulence, abdominal pain, diarrhea) were scored as follows: 0 = none, 1 = no more than usual, 2 = somewhat more than usual and 3 = much more than usual. The composite score was the sum of the six individual GI symptom ratings. Data are mean ± SEM.
The changes in breath hydrogen concentration in response to dextrose and BIOLIGO TM IL5040 IMO, over 24 h are shown in [fig_ref] Figure 2: Breath hydrogen [/fig_ref]. There was no significant difference between IMO and the control at all time points. Post-prandial glycemic [fig_ref] Figure 3: Post-prandial glycemic response in healthy adults [/fig_ref] and insulinemic response [fig_ref] Figure 3: Post-prandial glycemic response in healthy adults [/fig_ref] to dextrose and BIOLIGO TM IL5040 IMO, showed no significant differences over 4 h. Glucose iAUC in the first two and four hours was ~11% and ~15% lower, respectively, for IMO compared to the control [fig_ref] Table 4: Incremental area-under-curve [/fig_ref]. These differences were not large enough to reach statistical significance, but the difference in glucose iAUC over 4 h, neared significance (p = 0.058). Glucose iAUC from 2-4 h was significantly lower for IMO compared to the control (p = 0.008). There were no significant differences between treatments, for any of the other glucose and insulin iAUC. In study 2, the glucose iAUC 0-2 h means (± SEM) were not significantly different from one another (BIOLIGO IL7010: 201.5 ± 25.5; BIOLIGO IL5040: 181.3 ± 23.4; Dextrose: 191.3 ± 22.2: p = 0.662). Following consumption of two IMOs, there was no significant difference in capillary blood glucose concentrations compared to dextrose, in healthy adults [fig_ref] Figure 4: Post-prandial glycemic response to BIOLIGO TM IL5040 and BIOLIGO TM IL7010 IMO [/fig_ref]. The compositional differences between two IMOs [fig_ref] Table 5: Compositional differences of tested IMOs [/fig_ref] did not impact their effects on glycemic response. Glucose iAUC in the first two and four hours was ~11% and ~15% lower, respectively, for IMO compared to the control [fig_ref] Table 4: Incremental area-under-curve [/fig_ref]. These differences were not large enough to reach statistical significance, but the difference in glucose iAUC over 4 h, neared significance (p = 0.058). Glucose iAUC from 2-4 h was significantly lower for IMO compared to the control (p = 0.008). There were no significant differences between treatments, for any of the other glucose and insulin iAUC. In study 2, the glucose iAUC 0-2 h means (± SEM) were not significantly different from one another (BIOLIGO IL7010: 201.5 ± 25.5; BIOLIGO IL5040: 181.3 ± 23.4; Dextrose: 191.3 ± 22.2: p = 0.662). Following consumption of two IMOs, there was no significant difference in capillary blood glucose concentrations compared to dextrose, in healthy adults [fig_ref] Figure 4: Post-prandial glycemic response to BIOLIGO TM IL5040 and BIOLIGO TM IL7010 IMO [/fig_ref]. The compositional differences between two IMOs [fig_ref] Table 5: Compositional differences of tested IMOs [/fig_ref] did not impact their effects on glycemic response. Glucose iAUC in the first two and four hours was~11% and~15% lower, respectively, for IMO compared to the control [fig_ref] Table 4: Incremental area-under-curve [/fig_ref]. These differences were not large enough to reach statistical significance, but the difference in glucose iAUC over 4 h, neared significance (p = 0.058). Glucose iAUC from 2-4 h was significantly lower for IMO compared to the control (p = 0.008). There were no significant differences between treatments, for any of the other glucose and insulin iAUC. In study 2, the glucose iAUC 0-2 h means (± SEM) were not significantly different from one another (BIOLIGO™ IL7010: 201.5 ± 25.5; BIOLIGO™ IL5040: 181.3 ± 23.4; Dextrose: 191.3 ± 22.2: p = 0.662). Following consumption of two IMOs, there was no significant difference in capillary blood glucose concentrations compared to dextrose, in healthy adults [fig_ref] Figure 4: Post-prandial glycemic response to BIOLIGO TM IL5040 and BIOLIGO TM IL7010 IMO [/fig_ref]. The compositional differences between two IMOs [fig_ref] Table 5: Compositional differences of tested IMOs [/fig_ref] did not impact their effects on glycemic response. Table5. Compositional differences of tested IMOs. DP: Degree of polymerization. BIOLIGO TM IL5040 and BIOLIGO TM IL7010 are syrups with an average of 24% water content, as-is. * IMO content, in this paper, is defined as the sum of IMO components from DP2 to DP7 (isomaltose, panose, isomaltotriose, isomaltotetraose, isomaltopentaose, isomaltohexaose, isomaltoheptaose).
## Imo dp 1-2 (% dry basis) dp 3-8 (% dry
# Discussion
Sweeteners provide taste and rheological attributes (texture, flavor, preservative, and color) in food and beverages [bib_ref] The role of artificial and natural sweeteners in reducing the consumption of..., Mooradian [/bib_ref]. Although lowering post-prandial blood glucose response is considered as a physiological benefit, sugar-reduced products are less accepted by consumers. Thus, food manufacturers utilize various sweeteners based on their origin (natural or synthetic), technological function (taste and fillers), texture (powders and syrups), and nutritional value (caloric and non-caloric). Low digestible carbohydrates such as polyols or sugar alcohols are the most commonly used sugar substitutes, due to their low-caloric value, low glycemic response, and non-cariogenicity. Despite their health benefits, sugar alcohols may have transient gastrointestinal effects at excessive intakes.
Emerging nutritive sweeteners include rare sugars and oligosaccharides, and are appealing due to their natural source [bib_ref] The role of artificial and natural sweeteners in reducing the consumption of..., Mooradian [/bib_ref]. Oligosaccharides such as fructooligosaccharides (FOSs), and isomaltooligosaccharides (IMOs) are used as either partial or full-sugar replacers in food formulations [bib_ref] Effect of isomaltooligosaccharide syrup on quality characteristics of sponge cake, Lee [/bib_ref]. The majority of the studies that tested IMOs in Asian populations had conflicting evidence on their digestibility and fermentability. Glycemic response is an indicator of carbohydrate digestibility. Fully digestible carbohydrates such as dextrose produce a rapid rise and fall in blood glucose. Insulin is released in response to initial blood glucose rise and causes it to fall. Non-digestible carbohydrates containing glucose show negligible glycemic response, while partially absorbed polyols do not cause increased blood glucose levels. Previous studies, with different commercial IMOs, are inconsistent and reported mixed results (bifidogenic and no impact on blood glucose breath hydrogen) suggesting IMO to be partly digestible and partly fermentable [bib_ref] Functional food properties of non-digestible oligosaccharides: A consensus report from the ENDO..., Van Loo [/bib_ref]. This is the first study to evaluate gastrointestinal tolerance and glycemic response of IMO in a healthy population. In this short-term study, BIOLIGO IL5040 IMO at 68.5 g/day on an as-is basis (25 g/day pure IMO on a dry basis where IMO content is defined as in Section 2) was well-tolerated as demonstrated by having no impact on the composite gastrointestinal symptom score, and frequency of individual intestinal symptoms. Bouhnik et al. [bib_ref] The capacity of nondigestible carbohydrates to stimulate fecal bifidobacteria in healthy humans:..., Bouhnik [/bib_ref] tested IMO, among other oligosaccharides, at 10 g/day for one week in healthy adults and reported no significant changes in four intestinal symptoms (excess flatus, bloating, borborygmi, and abdominal pain). We measured breath hydrogen to assess fermentability of IMO over 24 h. Although breath hydrogen increased numerically between 6-12 h, there was no significant difference between IMO and dextrose over 24
# Discussion
Sweeteners provide taste and rheological attributes (texture, flavor, preservative, and color) in food and beverages [bib_ref] The role of artificial and natural sweeteners in reducing the consumption of..., Mooradian [/bib_ref]. Although lowering post-prandial blood glucose response is considered as a physiological benefit, sugar-reduced products are less accepted by consumers. Thus, food manufacturers utilize various sweeteners based on their origin (natural or synthetic), technological function (taste and fillers), texture (powders and syrups), and nutritional value (caloric and non-caloric). Low digestible carbohydrates such as polyols or sugar alcohols are the most commonly used sugar substitutes, due to their low-caloric value, low glycemic response, and non-cariogenicity. Despite their health benefits, sugar alcohols may have transient gastrointestinal effects at excessive intakes.
Emerging nutritive sweeteners include rare sugars and oligosaccharides, and are appealing due to their natural source [bib_ref] The role of artificial and natural sweeteners in reducing the consumption of..., Mooradian [/bib_ref]. Oligosaccharides such as fructooligosaccharides (FOSs), and isomaltooligosaccharides (IMOs) are used as either partial or full-sugar replacers in food formulations [bib_ref] Effect of isomaltooligosaccharide syrup on quality characteristics of sponge cake, Lee [/bib_ref]. The majority of the studies that tested IMOs in Asian populations had conflicting evidence on their digestibility and fermentability. Glycemic response is an indicator of carbohydrate digestibility. Fully digestible carbohydrates such as dextrose produce a rapid rise and fall in blood glucose. Insulin is released in response to initial blood glucose rise and causes it to fall. Non-digestible carbohydrates containing glucose show negligible glycemic response, while partially absorbed polyols do not cause increased blood glucose levels. Previous studies, with different commercial IMOs, are inconsistent and reported mixed results (bifidogenic and no impact on blood glucose breath hydrogen) suggesting IMO to be partly digestible and partly fermentable [bib_ref] Functional food properties of non-digestible oligosaccharides: A consensus report from the ENDO..., Van Loo [/bib_ref]. This is the first study to evaluate gastrointestinal tolerance and glycemic response of IMO in a healthy population. In this short-term study, BIOLIGO™ IL5040 IMO at 68.5 g/day on an as-is basis (25 g/day pure IMO on a dry basis where IMO content is defined as in Section 2) was well-tolerated as demonstrated by having no impact on the composite gastrointestinal symptom score, and frequency of individual intestinal symptoms. Bouhnik et al. [bib_ref] The capacity of nondigestible carbohydrates to stimulate fecal bifidobacteria in healthy humans:..., Bouhnik [/bib_ref] tested IMO, among other oligosaccharides, at 10 g/day for one week in healthy adults and reported no significant changes in four intestinal symptoms (excess flatus, bloating, borborygmi, and abdominal pain). We measured breath hydrogen to assess fermentability of IMO over 24 h. Although breath hydrogen increased numerically between 6-12 h, there was no significant difference between IMO and dextrose over 24 h. All participants were provided with a standardized (low fiber/low dairy) lunch, snack and dinner before and during the test days, to avoid the dietary impact on breath hydrogen. Earlier studies [bib_ref] Comparison of digestibility and breath hydrogen gas excretion of fructo-oligosaccharide, galactosyl-sucrose, and..., Oku [/bib_ref] examined breath hydrogen response to IMO up to 25 g/day for 7 h in healthy adults and reported no effect on breath hydrogen, reflecting insufficient evidence of fermentation.
In another study [bib_ref] Metabolism of 13C-isomaltooligosaccharides in healthy men, Kohmoto [/bib_ref] , increase in blood glucose indicated IMO to be highly glycemic. In both clinical studies presented in this paper, BIOLIGO TM IL5040 and IL7010 IMO dose-matched for 50 g total carbohydrates showed similar glycemic response compared to dextrose. Though iAUC for venous blood glucose was significantly lower than dextrose beyond 2 h, venous insulin response to BIOLIGO TM IL5040 showed no significant change compared to the control. The tested IMOs differed compositionally in Study 2, with BIOLIGO TM IL7010 having a higher average DP than BIOLIGO TM IL5040. However, this difference in the degree of polymerization between the two IMOs did not affect their digestibility and was shown to be both fully digestible and hence caloric.
The majority of the IMO studies assessed Bifidobacteria in Asian populations [bib_ref] Dose-response test of isomaltooligosaccharides for increasing fecal bifidobacteria, Kohmoto [/bib_ref] [bib_ref] Effects of isomaltooligosaccharides with different degrees of polymerization on human fecal bifidobactcria, Kaneko [/bib_ref] , while no Bifidogenic effect was observed in European men and women [bib_ref] The capacity of nondigestible carbohydrates to stimulate fecal bifidobacteria in healthy humans:..., Bouhnik [/bib_ref]. The studies that showed changes in Bifidobacteria, were not well-designed, with no proper control groups. Positive effects of IMO on stool frequency and stool weight were reported in constipated populations [bib_ref] Effects of isomalto-oligosaccharides on bowel functions and indicators of nutritional status in..., Chen [/bib_ref] [bib_ref] Use of isomalto-oligosaccharide in the treatment of lipid profiles and constipation in..., Wang [/bib_ref] [bib_ref] Long-term supplementation of isomalto-oligosaccharides improved colonic microflora profile, bowel function, and blood..., Yen [/bib_ref]. Although the present Study 1 did not assess changes in fecal microbiota, the low fermentation of the IMO in study 1 suggests that it would not provide a bifidogenic effect. Further research is necessary to confirm changes in fecal microbiota and determine if this product affects bowel habits.
According to the UK Food Standard Agency, commercial IMO is a novel ingredient and glycemic or digestible carbohydrate. The evidence suggests that IMO is highly digestible with a small residual portion reaching the colon (estimated at 10%) and affecting the microbiota. Certain populations, such as Asians may experience beneficial changes in the microbiota and changes in laxation for constipated individuals, with a sufficient dose, but no evidence exists to confirm these effects in non-Asian populations. Variability in IMO compositions from different manufacturers may be one of the reasons for the conflicting evidence on digestibility and fermentability.
Although we have demonstrated that the tested IMOs are digestible with high gastrointestinal tolerance, these two clinical studies have a few limitations. The current studies were conducted in a Caucasian population. Additional research in an Asian population may provide insights into population-specific differences in digestibility and fermentability. Secondly, these were acute studies, and microbial changes in stool samples were not measured. Long-term evaluation would be needed to determine whether these ingredients have a prebiotic effect or impact laxation patterns. However, due to the high digestibility and low fermentability reported in the present studies, a prebiotic effect is unlikely.
# Conclusions
BIOLIGO TM IMOs are well-tolerated as demonstrated by the lack of adverse gastrointestinal symptoms and they have no effect on breath hydrogen (an indicator of fermentability). Additionally, these IMOs are caloric sweeteners based on the glycemic and insulinemic response in healthy adults. Further studies are needed to determine the postprandial effects of larger doses of IMO on blood glucose, gastrointestinal tolerance and gut microbiota over longer durations.
[fig] Study 1: Glucose was measured using an enzymatic colorimetric method-GOD/PAP Method (Randox Laboratories Ltd., Kearneysville, WV, USA) utilizing glucose oxidase and peroxidase to degrade into Phenol, and 4-Aminoantipyrine, measured using Trinder indicator reaction at 505 nm. The increase in absorbance correlates with the glucose concentration of the sample with an analytical coefficient of variation (CV) of <2%. Insulin was measured with an immunoturbidimetry assay (Kamiya Biomedical Company, Seattle, WA, USA). A radioimmunoassay method for measuring insulin (HI-14K, Millipore Corporation, Billerica, MA, USA), was conducted at the University of California, Davis (Davis, CA, USA) in a subset of hemolyzed, and non-hemolyzed samples. A regression equation was developed to convert radioimmunoassay values to immunoturbidimetric values. Breath samples of end-alveolar air were collected into 10 mL glass vacuum tubes using an EasySampler device (Quintron Instruments, Milwaukee, WI, USA). The concentrations of hydrogen in breath samples were analyzed by gas chromatography with a resolution of 1 ppm, and accuracy of ±2-3 ppm and a linear range: 2-150 ppm for hydrogen (Microlyzer Gas Analyzer, model SC; Quintron Instruments, Milwaukee, WI, USA). [/fig]
[fig] Figure 1: Composite gastrointestinal tolerance scores (Study 1). [/fig]
[fig] Figure 2: Breath hydrogen (ppm) measured over 24 h (Study 1); Data are mean ± SEM. [/fig]
[fig] Figure 3: Post-prandial glycemic response in healthy adults (a) venous glucose (b) venous insulin concentration (Study 1). Data are mean ± SEM. [/fig]
[fig] Figure 4: Post-prandial glycemic response to BIOLIGO TM IL5040 and BIOLIGO TM IL7010 IMO (Study 2). Data are mean ± SEM. [/fig]
[table] Table 1: Subject demographics in study 1.Table 2. Subject demographics in study 2. [/table]
[table] Table 2: Subject demographics in study 2. [/table]
[table] Table 3: Frequency of scores ≥2 a on individual gastrointestinal (GI) symptoms (Study 1). [/table]
[table] Table 4: Incremental area-under-curve (iAUC) for glucose and insulin (Study 1). [/table]
[table] Table 5: Compositional differences of tested IMOs. [/table]
|
Exploring the nanomechanical concepts of development through recent updates in magnetically guided system
This article outlines an analytical analysis of unsteady mixed bioconvection buoyancy-driven nanofluid thermodynamics and gyrotactic microorganisms motion in the stagnation domain of the impulsively rotating sphere with convective boundary conditions. To make the equations physically realistic, zero mass transfer boundary conditions have been used. The Brownian motion and thermophoresis effects are incorporated in the nanofluid model. Magnetic dipole effect has been implemented. A system of partial differential equations is used to represent thermodynamics and gyrotactic microorganisms motion, which is then transformed into dimensionless ordinary differential equations. The solution methodology is involved by homotopy analysis method. The results obtained are based on the effect of dimensionless parameters on the velocity, temperature, nanoparticles concentration and density of the motile microorganisms profiles. The primary velocity increases as the mixed convection and viscoelastic parameters are increased while it decreases as the buoyancy ratio, ferro-hydrodynamic interaction and rotation parameters are increased. The secondary velocity decreases as viscoelastic parameter increases while it increases as the rotation parameter increases. Temperature is reduced as the Prandtl number and thermophoresis parameter are increased. The nanoparticles concentration is increased as the Brownian motion parameter increases. The motile density of gyrotactic microorganisms increases as the bioconvection Rayleigh number, rotation parameter and thermal Biot number are increased.In the presence of an external magnetic field, ferrofluids (a portmanteau of liquid and ferromagnetic particles) are magnetized liquid. These fluids are colloidal liquids made up of nanoscale ferrimagnetic or ferromagnetic particles that are suspended inside a fluid transport mechanism (usually a water or organic solvent). Brownian motion causes particle suspension, but these particles do not settle under normal conditions. Furthermore, each ferromagnetic particle is coated with a surfactant to prevent clumping. Magnetic attraction of nanoscale ferromagnetic particles is weak when the van der Waals force of the surfactant is strong enough to prevent agglomeration or magnetic clumping. Ferromagnetic fluids have a wide range of applications like friction-reducing agent, angular momentum change agent, heat transfer agent, and other applications including electronic devices, analytical instruments, and medicine 1-3 . Due to these numerous applications, many scientists and researchers OPEN have accelerated the study of ferrofluid. Andersson and Vanes 4 investigated the effect of a magnetic dipole on ferrofluid. Zeeshan et al. 5 reported the mixed convection flow and heat transfer in ferromagnetic fluid over a stretching sheet with partial slip effects. Hayat et al.6investigated the effects of thermal radiation and magnetic dipole on the flow of ferromagnetic Williamson liquid past a stretched surface. Some MHD studies can be read from the references 7-17 .Rotating flows and heat transfer levels over stationary spinning bodies with a revolution in forced flow are important in a wide range of engineering applications like missile re-entry, projectile motion, fiber coating, and rotating machine design. Many researchers have already studied its axis of revolution parallel to the free flow velocity to solve problems of heat transfer and fluid flow on a rotating sphere. Here are some studies on this topic 18-28 . However, most recently, Mahdy 29 investigated the concurrent effects of MHD and varying wall temperatures on the transient mixed Casson nanofluid flow at the stagnation point of the rotating sphere. Mahdy and Hossam 30 reported the time-mixed convective nanofluid flow from the stagnaion point area of the fast rotating sphere of Newtonian's heating with microorganisms. Some other rotating flows investigations exist in the references 31-33 .In perspective of the pervasive engineering applications and industry, the study of non-Newtonian fluid flows has indeed achieved an exemplary commitment. Materials exhibiting a non-linear relationship between strain rate and stress, with variable coefficient viscosity, and belongings described in the Navier-Stokes equation, are recognized as non-Newtonian fluids. Many industrial fluid processes, like molten polymers, paper production, paints, cosmetics, clay mixtures, oil recovery, food dispensation and movement of biological fluids, exhibit non-Newtonian characteristics. Several models of non-Newtonian material have been developed. However, in general, the flow equations are often more non-linear compared to Newtonian fluid equations. For these reasons, the research area of non-Newtonian fluids brings some exciting and interesting difficulties for engineers, mathematicians, and computer scientists alike. Ibrahim 34 provided the numerical analysis of time-dependent flow of viscous fluid due to a stretchable rotating disk with heat and mass transfer. Ibrahim et al.35studied the numerical simulation and sensitivity analysis of a non-Newtonian fluid flow inside a square chamber exposed to a magnetic field using the FDLBM approach. Interesting recent studies on non-Newtonian fluids have been conducted by various researchers[36][37][38][39][40][41]. The second-grade fluid model is the kindest subtype of viscoelastic fluid that can be expected for analytical solution as in[42][43][44][45].A larger percentage of nuclear and thermal-hydraulic systems requires the heat transmissions that attempt to incorporate fluid flow. A wide range of fluids and operating conditions have been verified to enhance the heat transport process. The interaction of such fluids with the existing system can help to reduce capital costs, improve working efficiency, and the design of the system. Furthermore, the involvement of cooling in several technological processes, such as engines, laptops, computers, electrical strips, is crucial to maintain the effective thermal performance of these products. Nanofluid is the attribute of a nanometer-sized particle suspended in liquid. Nanofluid has been shown to improve the thermal conductivity of the base fluid under the influence of such suspendend particles, and this concept originates from the work of Choi and Eastman 46 , which is then clarified by several researchers. El-Shorbagy et al.47investigated numerically the mixed convection in nanofluid flow in a trapezoidal channel with different aspect ratios in the presence of porous medium. Ali et al. 48 examined the navigation effect of tungsten oxide nano-powder on ethylene glycol surface tension by artificial neural network and response surface methodology. Chu et al. 49 studied the rheological behavior of MWCNT-TiO2/5W40 hybrid nanofluid based on experiments and RSM/ANN modeling. This leads to a number of researchers to see the huge efficient flow of nanofluids under different aspects 50,51,53-57 .Bioconvection is a new type of industrial and biological fluid mechanics in which the phenomenon of it arises from macroscopic convective fluid movements due to the inclusion of upwardly swimming motile microorganisms (denser than the medium). As a result, the agglomerate of the loaded self-propelled microorganisms on the liquid upper surface occurs, where it contributes to density stratification, is unstable and results in the image of the bioconvection dust clouds. Water must be a standard fluid so that these self-propelled microorganisms are active in the base fluid 58-66 .Inspired mostly by the above-mentioned references to bioconvection nanofluid flows, the purpose of this paper is to examine the behavior of transient mixed stagnation point boundary layer flow of second-grade fluid across the rashly revolving sphere consisting of nanoparticles and motile gyrotactic microorganisms under magnetic dipole effects. The impetuous movement of nanofluid, microorganisms and the impulsive rotation of the sphere lead to unsteadiness. Similarity transformation approach is used to determine the non-linear ordinary differential equations and solution through the homotoy analysis method HAM 67-70 .
## Problem formulation
The two-dimensional time-state, laminar incompressible, mixed convection boundary layer flow of viscous electrically conducting second grade nanofluid with swimming gyrotactic microorganisms in the stagnation domain of the rotating sphere and thermal convective boundary condition is scrutinized. Recognizing the Buongiorno nanofluid model, comprising of Brownian motion and thermophoresis effects is adopted in the present investigations. It is assumed that uniform dispersion is achieved by the lack of aggregation and accretion of nanoparticles. The sphere is revolving with a constant angular velocity . At time t = 0, the sphere is at rest and the surface temperature, concentration and motile microorganisms are T ∞ , C ∞ and N ∞ in an ambient fluid, respectively. The flow of motile microorganisms remains constant along sphere surface and zero mass flux condition is applied at the surface. The sphere surface is warmed due to convection from a warm nanofluid at the temperature T ∞ which provides a heat transfer factor h f , is to be strengthen or weaken to the value T ∞ , where T w > T ∞ leads to aiding flow and T w < T ∞ leads to reversing flow, respectively. Apart from nanofluid properties, which are chosen to be constant, density variation is based on Boussinesq approximation. Both Joule heating and viscous Magnetic dipole. The characteristics of the magnetic field have an effect on the flow of ferrofluid due to the magnetic dipole 2,4,6 . Magnetic dipole effects are recognized by the magnetic scalar potential shown as in Eq. [bib_ref] Commercial applications of ferrofluids, Raj [/bib_ref] where γ stands for the magnetic field strength at the source. Taking H x and H y as the components of magnetic field as shown in Eqs. [bib_ref] Effect of magnetic dipole on viscous ferro-fluid past a stretching surface with..., Zeeshan [/bib_ref] and (3),
Since the magnetic body strength is usually proportional to H x and H y gradient, c is the distance of the line currents from the leading edge, it is therefore given as in (4) Equation (5) can approximate the linear shape of magnetization M by temperature T as
The value of K 1 is identified as a ferromagnetic coefficient. [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] shows the physical diagram of the problem about magnetic dipole. The problem equations for second grade nanofluid with dipole effect are given as [bib_ref] Unsteady laminar MHD flow and heat transfer in the stagnation region of..., Takhar [/bib_ref] [bib_ref] Unsteady MHD rotating flow over a rotating sphere near the equator, Chamkha [/bib_ref] [bib_ref] Self-similar solution of the unsteady mixed convection flow in the stagnation point..., Anilkumar [/bib_ref] [bib_ref] Simultaneous impacts of MHD and variable wall temperature on transient mixed Casson..., Mahdy [/bib_ref] [bib_ref] Microorganisms time-mixed convection nanofluid flow by the stagnation domain of an impulsively..., Mahdy [/bib_ref] [bib_ref] Flow and heat transfer to modified second grade fluid over a non-linear..., Khan [/bib_ref] [bib_ref] Variable fluid properties of a second-grade fluid using two different temperaturedependent viscosity..., Salahuddin [/bib_ref] [bib_ref] Heat and mass transfer in hydromagnetic second-grade fluid past a porous inclined..., Bilal [/bib_ref] (1) www.nature.com/scientificreports/ The boundary conditions are as in [bib_ref] Simultaneous impacts of MHD and variable wall temperature on transient mixed Casson..., Mahdy [/bib_ref] [bib_ref] Microorganisms time-mixed convection nanofluid flow by the stagnation domain of an impulsively..., Mahdy [/bib_ref] where α 1 (>0) is the material parameter and u w is the stretching velocity.
[formula] � = γ 2π x x 2 + (y + c) 2 , (2) H x = − ∂� ∂x = γ 2π x 2 − (y + c) 2 [x 2 + (y + c) 2 ] 2 ,(3)H y = − ∂� ∂y = γ 2π 2x(y + c) [x 2 + (y + c) 2 ] 2 . (4) H = H 2 x + H 2 y . (5) M = K 1 (T − T ∞ ). (6) ∂(xu) ∂x + ∂(xv) ∂y = 0,(7)∂u ∂t + u ∂u ∂x + w ∂u ∂y − v 2 x = U ∂U ∂x + ν f ∂ 2 u ∂y 2 + α 1 ρ f ∞ u ∂ 3 u ∂x∂y 2 + v ∂ 3 u ∂y 3 + ∂u ∂x ∂ 2 u ∂y 2 − ∂u ∂y ∂ 2 u ∂x∂y + ∂w ∂y ∂ 2 w ∂x∂y + ∂w ∂x ∂ 2 w ∂y 2 + µ o M ρ f ∞ ∂H ∂x + (1 − C ∞ )βρ f ∞ (T − T ∞ ) − (ρ p − ρ f ∞ )(C − C ∞ ) − (ρ m − ρ f ∞ )γ * (N − N ∞ ) gx Rρ f ∞ , [/formula]
Using the following transformations as in [bib_ref] Simultaneous impacts of MHD and variable wall temperature on transient mixed Casson..., Mahdy [/bib_ref] [bib_ref] Microorganisms time-mixed convection nanofluid flow by the stagnation domain of an impulsively..., Mahdy [/bib_ref]
[formula] ζ = 2c 1 ν f η 1 2 y , η = 1 − e −c 1 t , u(x, y, t) = c 1 xf ′ (η, ζ ) , v(x, y, t) = �g(η, ζ ) , w(x, y, t) = − 2c 1 ν f ηf ′ (η, ζ ) , θ(ζ ) = T−T ∞ T w −T ∞ , φ(ζ ) = C−C ∞ C w −C ∞ , χ(ζ ) = N−N ∞ N w −N ∞ , [/formula]
the above governing equations provide the following non-dimensional form with non-dimensional boundary conditions given as Here c 1 is the constant such that c 1 > 0, η designates the dimensionless time parameter, ζ indicates the converted variable, f ′ and g stand for the velocity components in the x− and y− directions, T means the dimensionless temperature, φ is the dimensionless nanoparticles concentration, N is density of motile microorganisms, γ 1 denotes the mixed convection parameter, Nr points out the buoyancy ratio, β is the ferrohydrodynamic interaction parameter, Pr is the Prandtl number, Sc is the traditional Schmidt number, Sb is the bioconvection Schmidt number, 1 is the rotation parameter, is the heat dissipation parameter, ε is the curie temperature, d is the dimensionless distance, Nt is the thermophoresis, Nb is the Brownian motion parameter, N δ denotes the microorganisms concentration difference parameter, Rb gives bioconvection Rayleigh number, Pe is the bioconvection Peclet number, Re is the Reynolds number, A 1 is the viscoelastic parameter, A 2 is the dimensionless parameter, Bi is the thermal Biot number and ′ indicates the derivatives with respect to ζ . These parameters are given in mathematical expressions by www.nature.com/scientificreports/
[formula] (8) ∂v ∂t + u ∂v ∂x + w ∂v ∂y − uv x = µ f ∞ ρ f ∞ ∂ 2 v ∂y 2 + α 1 ρ f ∞ u ∂ 3 w ∂x∂y 2 + v ∂ 3 w ∂y 3 ,(9)(c p ρ) f ∞ ∂T ∂t + u ∂T ∂x + w ∂T ∂y = k f ∂ 2 T ∂y 2 + τ D B ∂C ∂y ∂T ∂y + D T T ∞ ∂T ∂y 2 + u ∂H ∂x + v ∂H ∂y µ o T ∂M ∂T ,(10)∂C ∂t + u ∂C ∂x + w ∂C ∂y = D B ∂ 2 C ∂y 2 + D T T ∞ ∂ 2 T ∂y 2 ,(11)∂N ∂t + u ∂N ∂x + w ∂N ∂y + bW c C ∞ ∂N ∂y ∂C ∂y = D n ∂ 2 N ∂y 2 . (12) t < 0 : u(x, y, t) = 0, v(x, y, t) = 0, w(x, y, t) = 0, T(x, y, t) = T ∞ , C(x, y, t) = C ∞ , N(x, y, t) = N ∞ ,(13)t ≥ 0 : u(x, 0, t) = U, v(x, 0, t) = �x, w(x, 0, t) = 0, −k f ∂T ∂y = h f (T f − T), D B ∂C ∂y + D T T ∞ ∂T ∂y = 0, N(x, 0, t) = N w , u(x, ∞, t) → U, v(x, ∞, t) → 0, T(x, ∞, t) → T ∞ , C(x, ∞, t) → C ∞ , N(x, ∞, t) → N ∞ ,(14)f ′′′ + 1 2 ζ(1 − η)f ′′ − η 1 + 1 g 2 − ff ′ 2 + A 1 2f ′ f ′′′ − f ′′ 2 + A 2 √ η f ′′′′ g − 2βθ (ζ + d) 4 + γ 1 η[θ − Nrφ − Rbχ] = 0, (15) g ′′ + 1 4 ζ(1 − η)g + 1 2 η(fg − f ′ g) − 2A 1 f ′′′ g = 0,(16)θ ′′ + Pr 1 4 ζ(1 − η)θ ′ + ηf θ ′ + Nbφ ′ θ ′ + Ntθ ′ 2 + ηβ (θ − ε) 2�g (ζ + d) 3 + ηPrβ (θ − ε) 2f ′ (ζ + d) 4 + 4�g (ζ + d) 5 = 0, (17) φ ′′ + Sc Nt Nb θ ′′ + 1 4 ζ(1 − η)φ + ηf φ ′ = 0,(18)χ ′′ − Pe χ ′ φ ′ + (χ + N δ )φ ′′ + Sb 4 ζ(1 − η)χ ′ + Sbηf χ ′ = 0,(19)f (η, 0) = f ′ (η, 0) = 0, g(η, 0) − 1 = 0, θ ′ = B i (θ − 1), Nbφ ′ (η, 0) + Ntθ(η, 0) = 0, χ(η, 0) = 1, f ′ (η, ∞) → 1, g(η, ∞) → 0, θ(η, ∞) → 0, φ(η, ∞) → 0, χ(η, ∞) → 0.1 = c 1 2 , A 1 = c 1 α 1 µ f∞ , A 2 = α 1 � µ f∞ 2c 1 ν f , β = γ µ 0 K 1 (T ∞ −T w )ρ f∞ 2πµ 2 f∞ , d = 2c 1 a 2 ν f , γ 1 = β T g(1−C ∞ )(T w −T ∞ ) c 2 1 R , Nr = (ρ p −ρ f∞ )C ∞ ρ f∞ β T (1−C ∞ )(T w −T ∞ ) , Rb = (ρ m −ρ f∞ )γ * (N w −N ∞ ) ρ f∞ β T (1−C ∞ )(T w −T ∞ ) , Pr = ν f (ρc p ) f∞ k f , Nb = τ D B C ∞ ν f , Nt = D T (T w −T ∞ ) T ∞ ν f , = c 1 µ 2 f∞ ρ f∞ (T w −T ∞ )k T , ε = T ∞ T ∞ −T w , Sc = ν f D B , Sb = ν f D B , Pe = bW c D n and Bi = h f k f ην f 2c 1 1 2 . [/formula]
According to the dimensionless variables, the significant designed physical quantities termed as the skin friction factor C f , Nusselt number Nu, and the density of the motile microorganisms number Nn are given in the form where shear stress, surface heat and motile surface microorganisms fluxes are mathematically expressed as Using the similarity transformations and Eq. (21) through Eq., the simplified forms are
where Re x = cx 2 ν f is the Reynolds number.
## Solution by homotopy analysis method
For nonlinear systems of partial or ordinary differential equations, Homotopy Analysis Method (HAM) is recognized as an important alternative to the conventional numerical methods. Liao [bib_ref] An explicit, totally analytic approximate solution for Blasius' viscous flow problems, Liao [/bib_ref] proposed the Homotopy Analysis Method, which uses the basic concepts of homotopy in topology to develop an alternative and general analytical-numerical method for nonlinear problems. The validity of HAM is independent of whether or not the considered equation contains small parameter(s). As a result, HAM overcomes the limitations of perturbation methods.
Taking the initial guesses and the linear operators as Equation (23) The non-linear operators are given by
[formula] (20) C f = τ w ρ f ∞ U 2 , Nu = xq r k f (T w − T ∞ ) , Nn = xq n D n (χ w − χ ∞ ) ,(21)τ w = µ f ∞ ∂u ∂y | y=0 , q r = −k f ∂u ∂y | y=0 , q n = −D n ∂u ∂y | y=0 .(22)Re 1 2 η 1 2 C f = 2 √ 2f ′′ (η, 0), Re − 1 2 η 1 2 Nu = − √ 2θ ′ (η, 0), Re − 1 2 η 1 2 Nn = − √ 2N ′ (η, 0),(23)f 0 (ζ ) = Aζ + (1 − A)(1 − e −ζ ), g 0 (ζ ) = e −ζ , θ 0 (ζ ) = Bi 1 + Bi e −ζ , φ 0 (ζ ) = e −ζ , χ 0 (ζ ) = e −ζ . [/formula]
(24)
[formula] L f (E 1 + E 2 e ζ + E 3 e −ζ ) = 0, L g (E 4 e ζ + E 5 e −ζ ) = 0, L θ (E 6 e ζ + E 7 e −ζ ) = 0, L φ (E 8 e ζ + E 9 e −ζ ) = 0, L χ (E 10 e ζ + E 11 e −ζ ) = 0,(25)(1 − q)L f [f (ζ , q) − f 0 (ζ )] = qh f N f [f (ζ , q), g(ζ , q), θ(ζ , q), φ(ζ , q), χ(ζ , q)], (1 − q)L g [g(ζ , q) − g 0 (ζ )] = qh g N g [g(ζ , q), f (ζ , q)], (1 − q)L θ [θ(ζ , q) − θ 0 (ζ )] = qh θ N θ [θ(ζ , q), f (ζ , q), g(ζ , q), φ(ζ , q)], (1 − q)L φ [φ(ζ , q) − φ 0 (ζ )] = qh φ N φ [φ(ζ , q), θ(ζ , q), f (ζ , q)], (1 − q)L χ [χ(ζ , q) − χ 0 (ζ )] = qh χ N χ [χ(ζ , q), φ(ζ , q), f (ζ , q)], (26) f (0, q) = 0, f ′ (0, q) = 1, f ′ (∞, q) = A, g ′ (0, q) = 1, g(∞; q) = 0, θ ′ (0, q) = −Bi(1 − θ(0, q)), θ(∞, q) = 0, φ ′ (0, q) = 1, φ(∞, q) = 0 χ ′ (0, q) = 1, χ(∞; q) = 0. Scientific Reports |(N f [f (ζ , q), g(ζ , q), θ(ζ , q), φ(ζ , q), χ(ζ , q)] = ∂ 3 f (ζ , q) ∂ζ 3 + 1 2 ζ(1 − η) ∂ 2 f (ζ , q) ∂ζ 2 − η 1 + 1 g 2 (ζ , q) − f ∂f (ζ , q) ∂ζ 2 + A 1 2 ∂f (ζ , q) ∂ζ ∂ 3 f (ζ , q) ∂ζ 3 − ∂ 2 f (ζ , q) ∂ζ 2 2 + A 2 √ η ∂ 4 f (ζ , q) ∂ζ 4 g(ζ , q) − 2βθ(ζ , q) (ζ + d) 4 + γ 1 (ζ , q)η θ(ζ , q) − Nrφ(ζ , q) − Rbχ(ζ , q) , N g [f (ζ , q), g(ζ , q)] = ∂ 2 g(ζ , q) ∂ζ 2 + 1 4 ζ(1 − η)g(ζ , q) + 1 2 η(f (ζ , q)g(ζ , q) − ∂f (ζ , q) ∂ζ g(ζ , q)) − 2A 1 ∂ 3 f (ζ , q) ∂ζ 3 g(ζ , q),(28)N θ [θ(ζ , q), φ(ζ , q), f (ζ , q), g(ζ , q)] = ∂ 2 θ(ζ , q) ∂ζ 2 + Pr 1 4 ζ(1 − η) ∂θ(ζ , q) ∂ζ + ηf θ(ζ , q) ∂θ(ζ , q) ∂ζ + Nb ∂θ(ζ , q) ∂ζ ∂φ(ζ , q) ∂ζ + Nt ∂θ(ζ , q) ∂ζ 2 + ηβ (θ − ε) 2�g(ζ , q) (ζ + d) 3 + ηPrβ (θ(ζ , q) − ε) 2 (ζ + d) 4 ∂f (ζ , q) ∂ζ + 4�g(ζ , q) (ζ + d) 5 ,(29)N φ [φ(ζ , q), θ(ζ , q), f (ζ , q)] = ∂ 2 φ(ζ , q) ∂ζ 2 (30) N φ [φ(ζ , q), θ(ζ , q), f (ζ , q)] = ∂ 2 φ(ζ , q) ∂ζ 2 N χ [χ(ζ , q), φ(ζ , q), f (ζ , q)] = ∂ 2 χ(ζ , q) ∂ζ 2 − Pe ∂χ(ζ , q) ∂ζ ∂φ(ζ , q) ∂ζ + (χ(ζ , q) + N δ ) ∂ 2 φ(ζ , q) ∂ζ 2 + Sb 4 ζ(1 − η) ∂χ(ζ , q) ∂ζ + Sbηf (ζ , q) ∂χ(ζ , q) ∂ζ ,(31)L f [f m (ζ , q) − ξ m f m−1 (ζ )] = h f R f ,m (ζ ),(32)L g [g m (ζ , q) − ξ m g m−1 (ζ )] = h g R g,m (ζ ),(33)L θ [θ m (ζ , q) − ξ m θ m−1 (ζ )] = h θ R θ ,m (ζ ),(34)L φ [φ m (ζ , q) − ξ m φ m−1 (ζ )] = h φ R φ,m (ζ ),(35)L χ [χ m (ζ , q) − ξ m χ m−1 (ζ )] = h χ R χ ,m (ζ ),(36)f m (0) = f ′ m (0) = f ′ m (∞) = 0, g ′ m (0) = g m (0) = g m (∞) = 0, θ ′ m (0) − B i1 θ m (0) = θ m (∞) = 0, φ ′ m (0) = φ m (0) = φ m (∞) = 0, χ ′ m (0) = χ m (0) = χ m (∞) = 0,(37) [/formula]
# Results and discussion
Solution authentication has an important role in the evaluations of the problems. Therefore, the present solution is compared with the published work. Order of approximation of the present work in is given which presents the close agreement with the published paper results [bib_ref] Development of dynamic model and analytical analysis for the diffusion of different..., Usman [/bib_ref]. . Comparison of the current work.
[formula] (38) R m f (ζ ) = f ′′′ m−1 + 1 2 ζ(1 − η)f ′′ m−1 − η 1 + 1 g 2 m−1 − m−1 r=0 r k=0 f m−1−r f ′ r−k f ′ k + A 1 2 m−1 r=0 f ′ m−1−r f ′′′ r − m−1 r=0 f ′′ m−1−r f ′′ r + A 2 √ η m−1 r=0 g m−1−r f ′′′′ r − 2βθ m−1 (ζ + d) 4 + γ 1 η[θ m−1 − Nrφ m−1 − Rbχ m−1 ] = 0,(39)R m g (ζ ) = g ′′ m−1 + 1 4 ζ(1 − η)g m−1 + 1 2 η m−1 r=0 f m−1−r g r − m−1 r=0 f ′ m−1−r g r − 2A 1 m−1 r=0 f ′′′ m−1−r g r ,(40)R m θ (ζ ) = θ ′′ m−1 + Pr 1 4 ζ(1 − η)θ ′ m−1 + η m−1 r=0 f m−1−r θ ′ r + Nb m−1 r=0 φ ′ m−1−r θ ′ r + Nt m−1 r=0 θ ′ m−1−r θ ′ r + ηβ 2� m−1 r=0 g m−1−r θ r (ζ + d) 3 − ηβ 2ε�g m−1 (ζ + d) 3 + ηPrβ 2 m−1 r=0 f ′ m−1−r θ r (ζ + d) 4 + 4� m−1 r=0 g m−1−r θ r (ζ + d) 5 − ηPrβ ε 2f ′ m−1 (ζ + d) 4 + 4�g m−1 (ζ + d) 5 ,(41)R m φ (ζ ) = φ ′′ m−1 + Sc Nt Nb θ ′′ m−1 + 1 4 ζ(1 − η)φ m−1 + η m−1 r=0 f m−1−r φ ′ r ,(43)R m χ (ζ ) = χ ′′ m−1 + Pe m−1 r=0 φ ′ m−1−r χ ′ r + m−1 r=0 φ ′′ m−1−r χ r + N δ φ ′′ m−1 + Sb 4 ζ(1 − η)χ ′ m−1 + Sbη m−1 r=0 f m−1−r χ ′ r . (44) f m (ζ ) = f * m (ζ ) + E 1 + E 2 e ζ + E 3 e −ζ ,(45)g m (ζ ) = g * m (ζ ) + E 4 e ζ + E 5 e −ζ ,(46)θ m (ζ ) = θ * m (ζ ) + E 6 e ζ + E 7 e −ζ ,(47)φ m (ζ ) = φ * m (ζ ) + E 8 e ζ + E 9 e −ζ ,(48)χ m (ζ ) = χ * m (ζ ) + E 10 e ζ + E 11 e −ζ , [/formula]
Order of approximation f ′′ (0) 28 f ′′ (0) (Present) g ′ (0) 28 g ′ (0) (Present) www.nature.com/scientificreports/ Velocity profiles. [fig_ref] Figure 2: f ′ [/fig_ref] show the impact of the rotation parameter on the velocity profiles. For elevating values of the rotation parameter 1 , the velocity f ′ is weakened and the velocity g is enhanced by higher values of rotation parameter 1 . The physical interpretation for this feature is attributed to the reduction of momentum and thermal boundary layers, which lead to an increase in the gradients of nanofluid velocity. [fig_ref] Figure 4: f ′ [/fig_ref] shows a declining trend in velocity f ′ for the higher estimation of the ferromagnetic parameter β .
Physically higher values develop more resistance to fluid flow. At the end, it reduces the velocity profile. [fig_ref] Figure 5: f ′ [/fig_ref] shows a growing trend in velocity f ′ for the larger values of γ 1 . This can be perceived as buoyancy aspects understanding, the convection cooling effects are enhanced by a strong acceleration of the flow. [fig_ref] Figure 6: f ′ [/fig_ref] shows that the dimensionless velocity decreases with an increase in the buoyancy ratio parameter Nr leading to an increase in the negative buoyant force caused by the presence of nanoparticles. The effect of increasing the bioconvection Rayleigh number Rb is that the convection power caused by bioconvection is enhanced against the convection of the buoyancy force. As a result, it could be noted that the flow velocity decreases with increasing the bioconvection Rayleigh number Rb values as shown in [fig_ref] Figure 7: f ′ [/fig_ref]. [fig_ref] Figure 8: f ′ [/fig_ref] is plotted here to measure the velocity variance against the viscoelastic parameter A 1 . The increasing velocity trend is aligned with the rising viscoelastic parameter A 1 . This behavior is rationalized by the mathematical representation of A 1 = α 1 c 1 /µ f ∞ that, by increasing www.nature.com/scientificreports/ the magnitude of A 1 , the viscosity decreases due to the velocity of the fluid. Mounts and the chaotic behavior of the uplifters are seen. It is efficient to note that the present problem reduces to the Newtonian case when A 1 = 0 . From the boundary layer point of view, the thickness of the fluid increases with an increase in the viscoelastic parameter A 1 . The opposite trend of velocity g for the viscoelastic parameter A 1 is observed as shown in [fig_ref] Figure 9: g [/fig_ref].
Temperature profile. [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] is used to measure the effect of the ferromagnetic parameter β on temperature. The temperature here rises with the greater values of ferromagnetic parameter β . The effect of the heat dissipation parameter on temperature is shown in [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref]. The temperature here increases with an increase in the the heat dissipation parameter values. Physically thermal conductivity of the fluid decays with the higher values of heat dissipation parameter . The temperature characteristics with the curie temperature parameter ε are shown in [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref]. The temperature decreases in the estimation of the curie temperature parameter ε . Nanofluid thermal conductivity increases with the increase in the curie temperature ε . In effect, the added heat is removed and the temperature rises from the surface to the nanofluid. [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] shows variations in temperature for the [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] shows the response of the thermal Biot number Bi to the temperature profile. The temperature is indicated to rise as a result of the increase in the thermal Biot number values. Thermal Biot number Bi depends on the coefficient of heat transfer or is directly proportional to the coefficient of heat transfer. [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] shows that the non-dimensional temperature and thermal boundary layer thickness decrease as the thermophoresis parameter Nt increases. The extra energy produced by the interaction of nanoparticles with the fluid due to the thermophoresis effect reduces the temperature. As a result, the thickness of the thermal boundary layer is reduced by higher values of the thermophoresis parameter Nt. As shown in [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] , the dimensionless temperature and thermal boundary layer thickness are increased with an increase in Brownian motion parameter Nb. The additional heat generated by the interaction between nanoparticles and the fluid due to Brownian motion increases the temperature. As a result, the thermal boundary layer thickness enhances with the positive values of the Brownian motion parameter Nb.
Nanoparticles concentration profile. The Schmidt number Sc is attributed to mass diffusion and therefore increases the mass diffusivity leading to a lower concentration of nanoparticles due to less mass diffusion transport, as shown in [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref]. [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] shows that the concentration of nanoparticles and hence the thickness of concentration boundary layer increase with the increase of the Brownian motion parameter Nb. [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] Motile microoganisms concentration profile. [fig_ref] Figure 2: f ′ [/fig_ref] shows the effect of the rotation parameter 1 on the motile microorganism density for the elevated values which increases the density of the motile microorganisms. Due to the strong relation of parameter in the governing equation 18 for the microorganisms, the dimensionless density of the motile microorganisms is highly influenced by the bioconvection Schmidt number Sb. It can be seen from [fig_ref] Figure 1: Schematic diagram of the problem [/fig_ref] that the rising values of Sb reduces the dimensionless density of the motile microorganisms profile. This is due to the bioconvection Schmidt number Sb which aids in the weakening of microorganisms concentration layer thickness, as indicated. The effect of the bioconvection Rayleigh number Rb on the density of motile microorganisms fluctuations is shown in [fig_ref] Figure 2: f ′ [/fig_ref]. It is clear that the bioconvection Rayleigh number Rb is enhancing the motile microorganisms. Of course, the Biot number Bi strengthens the convective
# Conclusion
Analysis of the themodynamics with the dipole effect and microorganisms shows that flow and heat transfer are enhanced with rotation parameter while for the magnetic dipole effect they have the opposite trend. Prandtl number has the cooling effect and the heat transfer increases with convective conditions. Nanoparticles concentration is improved with the Brownian motion parameter while the motile microorganisms motion is also enhanced with the rotation parameter, bioconvection Rayleigh number and convective condition. www.nature.com/scientificreports/
## Data availability
Availability exists for whole of the data. www.nature.com/scientificreports/
[fig] Figure 1: Schematic diagram of the problem. [/fig]
[fig] Figure 2: f ′ (ζ ) in terms of 1 . [/fig]
[fig] Figure 3: g(ζ ) in terms of 1 . [/fig]
[fig] Figure 4: f ′ (ζ ) in terms of β. [/fig]
[fig] Figure 5: f ′ (ζ ) in terms of γ 1 . of Prandtl number Pr. For the higher values of the Prandtl number Pr, the temperature of the nanofluid decreases. [/fig]
[fig] Figure 6: f ′ (ζ ) in terms of Nr. [/fig]
[fig] Figure 7: f ′ (ζ ) in terms of Rb. [/fig]
[fig] Figure 8: f ′ (ζ ) in terms of A 1 . [/fig]
[fig] Figure 9: g(ζ ) in terms of A 1 . [/fig]
|
Early Referral to Nephrological Care and the Uptake of Peritoneal Dialysis. An Analysis of German Claims Data
Citation: Schellartz, I.; Mettang, S.; Shukri, A.; Scholten, N.; Pfaff, H.; Mettang, T. Early Referral to Nephrological Care and the Uptake of
# Introduction
Patients with chronic kidney disease (CKD) suffer from a series of clinical problems. Their bodies are not sufficiently detoxified. When patients with CKD reach a certain point in the progression of their disease (end-stage renal disease, ESRD), they face a choice among different renal replacement therapies (RRTs). The two most commonly used types of dialysis are hemodialysis (HD) and peritoneal dialysis (PD).
With HD, toxins in the blood are filtered by a machine. It is usually conducted in ambulatory dialysis centers three times a week, each session taking four hours. PD can be applied by the patient themselves at home or in any clean environment. In order to enable the detoxication, a dialysate is filled inside the peritoneum and must be replaced regularly. This procedure must usually be conducted four times a day and takes about 30 minutes. Between the dialysate replacements, the dialysate stays in the body and enables detoxication. As the treatment can be arranged flexibly in terms of time and place, PD can enable a more autonomous life. Both dialysis treatments are symptomatic but life-sustaining. HD and PD are considered medically equivalent in terms of survival.
The progression of CKD and the time in which the need for dialysis becomes apparent can vary among patients. ESRD is usually reached with a glomerular filtration rate (GFR) < 15. The lower the GFR, the lower the level of kidney function is. The aim is to provide patient education, information about different RRT options, and dietary counseling, as well as psychological and social care. Therefore, an early referral (ER) of patients to a nephrological specialist at least one year before the expected dialysis starts, and with a GFR ≤ 30 mL/min, is recommended. Early care by a nephrological specialist can positively influence the progression of CKD. ER is also associated with better clinical outcomes such as a lower risk for mortality, better quality of life, and a lower hospitalization rate. CKD patients who were referred late (late referral, LR) were more likely to suffer from worse psychosocial adjustment. LR to a nephrological specialist is often seen as a barrier to PD. The literature shows that CKD patients with ER to nephrological care are more likely to conduct PD. A pooled analysis of a systematic Cochrane review reported patients with ER were twice as likely to choose PD. In these studies, ER was operationalized as care by a nephrological specialist between one month and one year before the patient's first dialysis.
Germany historically shows a low proportion of patients with ESRD dialyzing via PD of 6.7%. The PD ratio generally depends on patient-and physician-related factors as well as the socioeconomic context of the healthcare system. Among others, this includes whether PD or the transport to in-center HD are covered by health insurance providers. The compulsory statutory health insurance (SHI) in Germany reimburses all dialysis options, covers the transport to in-center HD, and providers also support a higher utilization of PD. Hence, framework conditions seem to be in place, and researchers have been trying to grasp, for years, why these reasons are strong drivers, especially in Germany.
Therefore, the aim of the current study is to investigate the association between ER and PD uptake with a large, unselected population of those insured by SHI in Germany. The underlying research question was: Does the ER of patients with CKD to a nephrological specialist increase the uptake of PD? To our knowledge, this article presents the first analysis on this topic using a large claims database reflecting the patient flow with all physician encounters and treatments in the healthcare sector. Thus, our study adds deeper insight into the referral situation in Germany from a system perspective, as well as its impact on PD uptake.
# Materials and methods
Ninety percent of the German population is insured with the compulsory SHI . Each SHI fund is obliged to contract every person. Therefore, their claims data offers an unselected view on the uptake of services of a broad collective of that population. German SHI claims data have also been used for investigations in nephrologyand when dealing with patients with other chronic conditions. We analyzed anonymous claims data of two of the largest SHIs in Germany: DAK-Gesundheit (Deutsche Angestellten Krankenkasse) and SBK (Siemens Betriebskrankenkasse). These funds cover 5.6 and 1.1 million insured members.
The database for this study consists of all medical encounters of 34,200 patients with ESRD from 2012 to 2016, i.e., those who had at least one documentation of International Classification of Diseases (ICD) code N18.5 between 2012 and 2016. The investigation was part of the MAU-PD study (Multidimensional analysis of causes for the low prevalence of ambulatory peritoneal dialysis in Germany). The study was funded by the Federal Joint Committee's innovation fund (Funding No 01VSF16036). It was ethically approved by the ethical committee of the University Hospital of Cologne, which considered all aspects of the Declaration of Helsinki.
## Structure of german claims data
In Germany, the reimbursement of outpatient medical services covered by the SHI is based on a codebook of different medical services (Einheitlicher Bewertungsmaßstab, EBM). In this codebook, each EBM code corresponds to a certain medical service. For inpatient admissions, each OPS code (Operationen-und Prozedurenschlüssel) corresponds to a certain medical procedure, e.g., the conduction of a dialysis treatment. The transfer of EBM and OPS codes to the SHI is relevant for the reimbursement of a service. All medically necessary services are covered by the SHI and are therefore contained in the claims data with OPS and EBM codes. These reimbursement codes make it possible to use claims data to analyze the care provided to patients and its effects.
## Identification of the first dialysis and the different dialysis treatments
We selected the OPS and EBM codes relevant for the reimbursement of dialysis (Supplementary. In our case control study, HD and PD patients were retrospectively identified depending on whether the code of their first dialysis was an in-center HD or home-based PD. With additional reimbursement codes for special dialysis forms, we excluded patients conducting special types of dialysis: intermittent peritoneal dialysis (IPD) and home hemodialysis (HHD). IPD is conducted in special cases and also at the dialysis center. HHD patients usually start with in-center HD and later switch to HHD. We cannot directly distinguish between the pre-dialysis care and the time between the first HD and the implementation of HHD as a direct preparation time for HHD. Therefore, reimbursement codes in SHI claims data are not adequate to assess the impact of ER on the uptake of HHD. Hence, the difference between in-center HD and home-based PD is of interest for the claims data analysis of referral.
## Identification of the first encounter with a nephrologist
The identification of outpatient specialist care by a nephrologist before the first dialysis is based on the documentation of the continuous care of a CKD patient with a GFR below 40 mL/min. Only nephrologists are authorized to bill the continuous care of a CKD patient. Further, the nephrologist is obliged to educate the patient about dialysis and/or transplantation. In our analysis, the first documentation was determined as the first encounter of a CKD patient with a nephrologist. This scheduled the time of referral and the beginning of the care by a nephrological specialist. Patients who did not have any documentation in their data since 2012 were treated as having no encounter with a nephrologist before their first dialysis. There is a different reimbursement code for the continuous care of patients who have had a kidney transplant. Hence, our approach excluded patients who have had a kidney transplant.
## Personal characteristics
Year of birth and sex were also provided in the claims data. We calculated the patient's age at their first dialysis with their year of birth and the date of the first dialysis. The non-age-adjusted Charlson Comorbidity Index (CCI) was calculated for the time of each patient's first dialysis. The CCI weighs 17 comorbid conditions, e.g., diabetes or congestive heart failure, to assess the risk for mortality. Higher scores indicate a higher mortality risk.
# Analysis
We included incident HD and PD patients in 2015 and 2016 in our study population. Our case-control study retrospectively focuses on the referral to the nephrologist and, therefore, on the time before the first dialysis. This gave us the possibility to analyze the medical encounters since 2012. The time between a patient's first dialysis and their first encounter with a nephrologist was calculated.
In the literature, ER was operationalized as care by a nephrological specialist between one month and one year before a patient's first dialysis. Hence, patients who had their first encounter with a nephrologist within six months (≤180 days) before their first dialysis were classified as LR, and those whose first encounter exceeded six months were classified as ER. Using the clinical binary classification allowed us to correctly assign those patients who already received nephrological care before 2012 to the ER group.
The aim is to investigate the impact of the category of referral on the choice between HD and PD. Chi-square tests and Wilcoxon-Mann-Whitney tests were applied to investigate significant differences between the category of referral (ER or LR), type of dialysis (HD or PD), sex, age, setting (outpatient or inpatient), and CCI, respectively. Patients starting their dialysis in an inpatient setting often suffer from a worse medical condition. Adjusted odds ratios (AORs) for the chance for PD (outcome: HD or PD) were calculated in a multivariate logistic regression model. To adjust for possible confounders, the independent variables besides the category of referral (ER or LR) were sex, age at first dialysis, setting (outpatient or inpatient), and the CCI. The significance level for all tests was 0.05. The 95% confidence intervals (CI) and R 2 were reported. Data analysis was conducted using Stata 16 (StataCorp., College Station, TX, USA).
# Results
After excluding patients who did not meet inclusion criteria, the study population consisted of 4727 patients. In 2015, 2326 patients started with dialysis treatment, and in 2016, 2401 patients began treatment. Twelve percent were contracted to SBK and 88% to DAK-Gesundheit. Forty-two percent were female. The mean age at first dialysis was 71 years (median 75). The youngest patient was 19, and the oldest was 97. Twenty-nine percent of the dialysis initiations were performed in an outpatient setting. In 96.3% of the cases, patients started with HD, and the remaining 3.7% began with PD. Mean CCI was 8.0. Forty-three percent had their first encounter with a nephrologist more than six months before their first dialysis and were thus categorized as patients with ER. A total of 2673 patients (57%) have been referred within six months before their first dialysis (LR).
Single test results of differences between HD and PD as well as ER and LR are displayed in Tables 1 and 2. The proportion of outpatient dialysis initiations was higher among PD patients than HD patients (51% vs. 28%, p < 0.001). A Wilcoxon-Mann-Whitney test showed PD patients to be significantly younger at their first dialysis than HD patients (60 vs. 72 years, p < 0.001). They also had fewer comorbidities (5.5 vs. 8.0, p < 0.001). The ratio of PD patients was significantly higher among those with ER than those with LR (5.8% vs. 2.0%, p < 0.001). A univariate logistic regression shows that patients referred early were 31% more likely to start their dialysis therapy within an outpatient setting (OR = 0.69, p < 0.001, 95% CI 0.27-0.35). Patients having their first dialysis within an inpatient setting were significantly older (mean 72 vs. 69 years, p < 0.001). The mean CCI of patients having a first inpatient dialysis was higher (8.4 vs. 6.9, p < 0.001). Single test results of differences between HD and PD as well as ER and LR are displayed in Tables 1 and 2. The proportion of outpatient dialysis initiations was higher among PD patients than HD patients (51% vs. 28%, p < 0.001). A Wilcoxon-Mann-Whitney test showed PD patients to be significantly younger at their first dialysis than HD patients (60 vs. 72 years, p < 0.001). They also had fewer comorbidities (5.5 vs. 8.0, p < 0.001). The ratio of PD patients was significantly higher among those with ER than those with LR (5.8% vs. 2.0%, p < 0.001). A univariate logistic regression shows that patients referred early were 31% more likely to start their dialysis therapy within an outpatient setting (OR = 0.69, p < 0.001, 95% CI 0.27-0.35). Patients having their first dialysis within an inpatient setting were significantly older (mean 72 vs. 69 years, p < 0.001). The mean CCI of patients having a first inpatient dialysis was higher (8.4 vs. 6.9, p < 0.001). As HD and PD patients significantly varied in sex, age, setting (inpatient or outpatient), and CCI, these factors were added as confounding variables in a multivariate logistic regression model that examines the influence of ER on PD uptake. Results are displayed in. The AOR in this model adjusted for sex, age, setting, and CCI highlights that patients with ER had a threefold increased chance for PD (AOR = 3.05; 95% CI 2.16-4.32; p < 0.001). Sex was not significant in this model. One additional year of life meant a 4% lower chance for PD (AOR = 0.96; 95% CI 0.95-0.97; p < 0.001). An AOR of 0.71 in inpatient admission meant a 29% lower chance for PD (AOR = 0.71; 95% CI 0.51-0.99; p = 0.044). Each point higher on the CCI resulted in a 15% lower chance for PD (AOR = 0.85; 95% CI 0.44-1.58; p < 0.001). Pseudo R 2 was 0.135.
# Discussion
We investigated the impact of ER to nephrological treatment on PD uptake with a claims data analysis. Forty-three percent of the patients were referred to the nephrologist more than six months before their first dialysis. These patients were categorized as ER. The adjusted multivariate logistic regression model highlights that ER significantly increases PD uptake. Adjusted for potential confounders, patients with ER had a threefold increased chance for a PD. The confounder's age, setting (inpatient or outpatient), and CCI also maintained their impact on the chance for PD. In the multivariate regression model, PD uptake was associated with ER, outpatient setting, younger age, and lower CCI.
The operationalization of ER in the literature varied between one month and one year. The different definitions discussed in a reviewconsidered the time for medical preparation as a reason for the narrow definition of 3-4 months. However, they also highlighted that guidelines' referral recommendations and other wider definitions are oriented toward the prevention of early CKD progression. As this investigation focused on the choice between HD and PD, we assumed that a period of more than six months offers the possibility for an informed decision. Our results confirmed recent studies: patients with ER to care by a nephrological specialist have a higher chance for PD. Effect sizes vary among these studies. A pooled analysis reported twice a chance. Our adjusted model from a system perspective revealed ER patients have a three-times increased chance for a PD.
The base for ER operationalization was the documentation of the continuous care of CKD patients with a GFR < 40. The opportunity for CKD patients in this state of progression to receive care by a nephrological specialist even allows for an earlier referral than recommended (GFR < 30). The results of our study indicate that the framework condition given by the German healthcare system contributes to ER. Forty-three percent of patients with ER and its effect on PD uptake highlighted the relevance of long-term nephrologist care for patients with CKD. Information about different RRT options is fundamental for pre-dialysis care by nephrologists and is thus a mandatory and crucial part of the long-term nephrologist care defined by this code. Six months of continuous nephrological care seems sufficient to educate patients about different RRTs. Nephrologists can provide information about different treatment options and involve relatives. The education gives both patients and nephrologists time to discuss different RRTs and weigh the advantages and disadvantages. Patients may have the opportunity to meet a patient who is already on PD. A patient's situation at home can be prepared. For a technically planned dialysis start, at least a couple of weeks are required for shunt or catheter implantation. PD also needs a certain "break-in period" between the implantation and the first dialysis. As the goal is to enable a shared decision and not to increase the PD ratio itself, ER is valuable to informed decision-making for an RRT.
The referral to care by a nephrological specialist in an early stage of CKD is not only meaningful with regard to the start of the dialysis treatment. It is also a relevant issue because ER can slow down the progression of CKD, and it offers a way to prevent ESRD and the start of a dialysis treatment. As this is difficult to assess with claims data, we focused on investigating the association between ER and PD uptake.
Despite the possibility of nephrological treatment and a close care network in Germany, there is still a high proportion of patients with LR (57%). Baer and colleagues summarized reasons for LR. Besides medical reasons, they also stress possible personal factors of patients, such as age and comorbidities. General practitioners could misinterpret the benefits of an RRT for the elderly, which can end in LR. Information campaigns for general practitioners about the comorbidities of chronic diseases and CKD can help to decrease the high proportion of LR.
The ambulatory and stationary sectors are strongly separated in Germany. An inpatient dialysis start may be an indicator of an urgent start due to a patient's acute medical condition during an inpatient admission. In our sample, the proportion of patients starting their dialysis in an inpatient setting is higher than those with LR (82% vs. 57%, respectively). There are data in the literature suggesting that HD and PD are both feasible and safe alternatives for urgent dialysis. Nevertheless, an urgent start of dialysis is often seen as a barrier to PD. The significantly lower PD ratio within inpatient dialysis initiations in our study also seems to indicate that PD is still not generally adopted as an equivalent alternative for an urgent dialysis start.
Our results corroborate the findings of previous studies about the effect of ER on PD uptake based on a large sample from a system perspective. The adjusted model highlights that ER has a major impact on PD uptake. Adjusted odds ratios indicate that ER is more important than a patient's underlying comorbidities, age, and setting. Despite the close care structure in Germany, there is still a high proportion of patients with LR. Further investigations should pay attention to reasons for the high proportion of LR to reduce this barrier to informed decisions.
# Strengths and limitations
This study added a system perspective on the ER of patients with CKD by analyzing the claims data of two large German SHIs. About 90% of the German population is contracted with SHIs [25], with 2401 patients starting dialysis only in 2016. DAK-Gesundheit and SBK cover 6.6 million people. The large database allowed us to ensure medical encounters and treatment groups with valid criteria. The PD proportion in our sample was lower than the one stated by the official national report for 2016 (3.7% vs. 6.7%, respectively). This may be due to the fact that our calculation was based on the number of dialysis patients, whereas the national report counted the number of dialysis treatments. Nevertheless, the large database still enabled us to examine a large sample of 4727 incident dialysis patients in 2015 and 2016.
Due to the compulsory SHI, German SHI claims data provide a representative sample. It is well defined how diseases and treatments are to be coded. Therefore, we expect our results can be confirmed by other SHI funds data and that they have high external validity.
We are aware that our research comes with limitations concerning SHI claims data analysis in general. Such data only consists of procedures relevant to reimbursement. Parts of treatment that are not reimbursed were not included. Consistent with practitioners' routine coding practice, we used a valid code for the counseling of pre-dialysis CKD patients. This code cannot reflect the extent to which patients have actually been informed. However, even with the selected measurements in primary data, the intensity and time spent on educational counseling over a longer period of time are difficult to measure. The reimbursement code can reflect pre-dialysis nephrological treatment, and this is what the referral category is meant to measure. The precise reimbursement codes made it possible to investigate referrals to nephrologists and pre-dialysis treatments from a system perspective.
The multivariate model's variables predicted 13.5% of PD uptake. As mentioned before, there are several factors on patient-, physician-, and system-level relevant for the choice between HD and PD. The Pseudo R 2 in our model confirms that ER is one of them, but not the only factor impacting on PD uptake.
# Conclusions
The considerable number of patients with ER led us to conclude that framework conditions in the German healthcare system support the early care of patients with CKD by nephrologists. The results of the first claims data analysis with a broad collective of those covered by SHIs confirmed the existing evidence: ER increases PD uptake in Germany. ER gives both nephrologists and patients time for patient education about different RRTs and contributes to informed decisions about the dialysis treatment.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/ijerph18168359/s1,. Dialysis-associated EBM and OPS codes and their operationalization for the analysis.
Author Contributions: I.S.: conceptualization, data curation, formal analysis, investigation, methodology, project administration, software, supervision, validation, visualization, and writing-original draft preparation; S.M.: conceptualization, methodology, and writing-review and editing; A.S.: formal analysis, methodology, software, validation, and writing-review and editing; N.S.: conceptualization, funding acquisition, investigation, project administration, resources, and writing-review and editing; H.P.: funding acquisition, resources, supervision, and writing-review and editing; T.M.: conceptualization, formal analysis, funding acquisition, methodology, supervision, and writingreview and editing. All authors have read and agreed to the published version of the manuscript. Informed Consent Statement: Patient consent was waived due to the anonymous transfer of claims data.
# Data availability statement:
Restrictions apply to the availability of these data. Data were obtained from DAK-Gesundheit and SBK and are available with the permission of DAK-Gesundheit and SBK.
|
Diffuse lesion and necrosis tied to poorer prognosis of interdigitating dendritic cell sarcoma: cases report and a pooled analysis
Interdigitating dendritic cell sarcoma is a neoplastic proliferation of interdigitating dendritic cells and no therapeutic consensus exists. This study aimed to investigate the prognostic impacts of tumor lesion, cellular atypia, mitosis and necrosis on the interdigitating dendritic cell sarcoma. Case reports and pooled analyses were designed to explore the relationships. One case was a 40-years old man with localized lesion, moderate to notable cellular atypia, 30 mitoses per 10 high-power fields and no necrosis and the progression-free survival was longer than 20 months. The other case was a 62-years old woman with diffuse lesion, notable cellular atypia, less than one mitosis per 10 high-power fields and diffuse necrosis and the progression-free survival was shorter than 1 month. Cellular atypia and mitosis had not any relationship with survival. Compared with localized lesion, diffuse lesion presented a 2.92fold risk of progression (HR = 2.92, 95% CI 1.01, 8.51) and an 8.79-fold risk of death (HR = 8.79, 95% CI 1.86, 41.64). Diffuse necrosis presented a 4.39-fold higher progression risk (HR = 5.39, 95% CI 1.78, 16.29) and a 5.37-fold higher death risk (HR = 6.37, 95% CI 1.46, 27.86) than focal or no necrosis. Diffuse lesion and diffuse necrosis were indicators of poorer prognosis and the clinical application should be warranted in further studies.
smashed white-grayish lymph nodes and the section showed solid, white-grayish, focal black, rough and diffused necrosis [fig_ref] Figure 1: The section of case 1 was solid, white-grayish, fine and smooth, no... [/fig_ref].
Histopathology and immunochemistry. Case 1: Histologic sections showed the lymph node replaced by spindle cells of moderate/severe cellular atypia. Neoplastic cells were disposed in fascicles and a focal storiform pattern. The nucleoli were irregular and presented bizarre, binucleated, or multinucleated forms, with 30 mitoses/10 high-power fields (HPF) [fig_ref] Figure 2: Histologic sections of case 1 displayed the spindle tumor cells disposing in... [/fig_ref]. There was no necrosis. The tumor cells showed positive of CD68 [fig_ref] Figure 2: Histologic sections of case 1 displayed the spindle tumor cells disposing in... [/fig_ref] , S100 [fig_ref] Figure 2: Histologic sections of case 1 displayed the spindle tumor cells disposing in... [/fig_ref] , Vimentin [fig_ref] Figure 2: Histologic sections of case 1 displayed the spindle tumor cells disposing in... [/fig_ref] and Ki-67 index 80% [fig_ref] Figure 2: Histologic sections of case 1 displayed the spindle tumor cells disposing in... [/fig_ref]. CKpan, cam5.2, EMA, CD21, CD23, CD35, SMA, HMB45, Melan A, CD1a, GFAP, ALK, LCA, CD117, CD34 and Myoglobin were negative in the cells.
Case 2: Section showed the lymph nodes destroyed and replaced by epithelium-like cells of severe cellular atypia. The tumor cells included spindle and oval cells, had a diffuse distribution and showed abundant cytoplasm, prominent nucleoli and rare mitoses (less than 5/50 HPF) [fig_ref] Figure 3: Histologic sections of case 2 displayed the sheets of atypia of epithelium-like... [/fig_ref]. The necrosis was diffuse. The tumor cells expressed CD68 [fig_ref] Figure 3: Histologic sections of case 2 displayed the sheets of atypia of epithelium-like... [/fig_ref] , S100 [fig_ref] Figure 3: Histologic sections of case 2 displayed the sheets of atypia of epithelium-like... [/fig_ref] , CKpan [fig_ref] Figure 3: Histologic sections of case 2 displayed the sheets of atypia of epithelium-like... [/fig_ref] and Ki-67 index 30% [fig_ref] Figure 3: Histologic sections of case 2 displayed the sheets of atypia of epithelium-like... [/fig_ref]. Cam5.2, EMA, CD21, CD23, CD35, SMA, HMB45, Melan A, CD1a, GFAP, ALK, LCA, CD117, CD34 and Myoglobin were negative.
ISH and PCR. Detection of Epstein-Barrvirus-encoded small RNA was negative for two cases by in situ hybridization. T and B cell polymerase chain reaction (PCR) clonality studies were not rearrangement. Braf with V600E mutation was not detected.
Electron microscopic analysis. The tumor cells had mutual interlaced long finger bumps and the borderline was unclear in the two cases. There were small amount of rough endoplasmic reticulum in cytoplasm. Nucleus had dents. There was no a typical bridge [fig_ref] Figure 4: Electron microscopic sections presented the mutual interlaced long finger bumps and small... [/fig_ref].
Treatment regimens and outcomes. Case 1 received three cycles of the CHOPE regimen, consisting of 1 g cyclophosphamide (day 1), 70 mg pirarubicin (day 1), 4 mg vindesine (day 1), 15 mg dexamethasone (days 1-5) and 100 mg etoposide (days 2-4), and six cycles of the DHAP regimen, consisting of 20 mg dexamethasone (days 1-4),100 mg cisplatin (day 1) and 500 mg cytarabine (Q12H days 2). Currently, the patient was followed upto 20 months, and no disease progression or distant metastasis occurred. The case 2 patient had two cycles of paclitaxel and nedaplatin regimen, but the disease was progressed.
Pooled analysis. Median age of included cases was 51.0-year old, with interquartile range of 30.0 [fig_ref] Table 1: General characteristics of included cases in the pooled analysis [/fig_ref]. 47.2% cases were females, 26.4% cases had a localized lesion, 39.6% cases had mild/moderate cellular atypia, 46.5% cases had 5 mitoses per 10HPF at least and 31.7% cases had diffuse necrosis [fig_ref] Table 1: General characteristics of included cases in the pooled analysis [/fig_ref]. Median age was similar between histopathological factors of tumor lesion, cellular atypia and necrosis (p > 0.05) [fig_ref] Figure 5: Median age has not any significant differences between tumor stage [/fig_ref] ,B,D). However, median age was 49 years old among patients with nuclear mitosis less than 5/10HPF, 12 years younger than patients with nuclear mitosis equal or higher than 5/10HPF (p = 0.011) [fig_ref] Figure 5: Median age has not any significant differences between tumor stage [/fig_ref].
Patients with diffused lesion had median PFS of 3 months, in contrast, patients with localized lesion had not reached median PFS [fig_ref] Figure 6: Patients with diffused stage had median PFS of 3 months, in contrast,... [/fig_ref]. The difference was significant. Cellular atypia and mitosis were not associated with PFS [fig_ref] Figure 6: Patients with diffused stage had median PFS of 3 months, in contrast,... [/fig_ref]. Median PFS of diffuse necrosis was 3 months but patients with no necrosis or focal necrosis had not observed median PFS, the difference being significant [fig_ref] Figure 6: Patients with diffused stage had median PFS of 3 months, in contrast,... [/fig_ref]. In multivariate analysis, diffuse lesion presented a 2.92-fold risk of progression (HR = 2.92, 95% CI 1.01, 8.51) and an 8.79-fold risk of death (HR = 8.79, 95% CI 1.86, 41.64) .
Diffuse tumor lesion was associated with a shorter OS significantly [fig_ref] Figure 7: Tumor stage was associated with a shorter OS significantly [/fig_ref]. Patients with localized lesion did not reach median OS, but patients with diffuse lesion had a 9-month OS. Cellular atypia and mitosis were not related with OS [fig_ref] Figure 7: Tumor stage was associated with a shorter OS significantly [/fig_ref]. Patients with no or focal necrosis did not get median OS, compared with patients with diffused necrosis having a 9-month median OS (p < 0.001) [fig_ref] Figure 7: Tumor stage was associated with a shorter OS significantly [/fig_ref]. Diffuse necrosis was associated with a 5.39-fold high progression risk (HR = 5.39, 95% CI 1.78, 16.29) and a 6.37-fold high death risk (HR = 6.37, 95% CI 1.46, 27.86) .
# Materials and methods
This study was approved by Beijing Shijitan Hospital Institutional Review Board (IRB). Written informed consent was obtained from all patients. No identifying information or image was published. All of the methods were in accordance with the particular regulations.
Study design. This study design consisted of a case report and a pooled analysis.
Pooled analysis. This study chose the databases of PubMed and CHINAINFO. PubMed was accessed through NCBI and CHINAINFO was accessed through WANFANG DATA.
Search term was "interdigitating dendritic cell sarcoma". The publication date ranged from Jan. 1, 1980 to Mar. 1, 2016. The language covered English and Chinese. Included article type was original research and the excluded types were reviews, meeting abstract, commentaries or re-analysis of data. All of the include IDCS cases should had pathlogical diagnosis in the publication. The endpoint outcome was progression-free survival (PFS), and overall survival (OS). The extracted data were age, gender, tumor lesion, cellular atypia, mitosis, necrosis, PFS and OS.
126 articles were obtained in PubMed and 53 articles were obtained in CHINAINFO. We excluded 42 articles in PubMed and 11 articles in CHINAINFO without IDCS cases. We reviewed all of the rest articles and found 34 cases in PubMed and 17 cases in CHINAINFO with histopathology data and outcome data.
Our two IDCS cases came from Beijing Shijitan hospital. The confirmed diagnosis covered by histological review of morphology in slides stained with hematoxylin and eosin, detection of a wide immunohistochemistry panel and application of electron microscope. The two cases were included in the pooled analysis.
Case report. Case 1: a 40-years old man visited the outpatient clinic presenting a two-week history of a painless mass in his left inguinal. The patient had foot nail removal for paronychia caused by trauma five years ago. No skin moles and tumor was found at lower limbs. The patient had no family cancer history and was in good health before. Ultrasound showed one low echo node in the right neck, the size being 0.9 × 0.2 cm; some low echo nodes in the left neck, the biggest size being 0.5 × 0.3 cm; some low echo nodes in the right inguinal, the biggest size being 0.8 × 0.3 cm; some low echo nodes in the left inguinal, the biggest size being 1.3 × 1.0 cm. All of the nodes had clearly border and regular contour. Pelvic CT [fig_ref] Figure 8: Pelvic CT scanning in case 1 displayed the lymph nodes aside the... [/fig_ref] showed the lymph nodes aside the left external iliac blood vessels enlarged and the biggest one was 1.8 × 1.4 cm, but there was no enlarged lymph node in the right external iliac blood vessels and retroperitoneal. There was no abnormality in the liver and spleen. PETCT [fig_ref] Figure 8: Pelvic CT scanning in case 1 displayed the lymph nodes aside the... [/fig_ref] displayed a high bone salt metabolism in right coracoid and left sacroiliac joint, indicating a benign lesion. Blood and bone marrow aspiration were normal.
Case 2: a 62-years old woman presented cough and hard breath for 3 months. Superficial lymph nodes were not touched and blood test was normal. Chest CT [fig_ref] Figure 8: Pelvic CT scanning in case 1 displayed the lymph nodes aside the... [/fig_ref] scan showed: multiple lymph nodes enlarged in right hilar and mediastinal, multiple nodules with calcification in the upper lobe of right lung and some patchy enhancement spot in the middle of the upper lobe, left lower lobe and secondary adjacent bronchiectasis. Abdomen CT [fig_ref] Figure 8: Pelvic CT scanning in case 1 displayed the lymph nodes aside the... [/fig_ref] scan showed a clear borderline mass in the right kidney, with the size of 4.4 × 4.1 × 5.8 cm, and a slightly larger spleen. Mediastinal lymph node was biopsied and pathological diagnosis was tending to be metastasis of poorly differentiated carcinoma with a large number of necrosis. The patient had two cycles chemotherapy of paclitaxel and nedaplatin but the disease progressed. The patient had mediastinal lymph node biopsy again.
Definition. Diffuse lesions means the lesion involved other organs outside the primary site. Focal necrosis is small, usually no more than 0.3 mm, and the margin is relatively regular, cannot be seen in the gross appearance; diffuse necrosis is large, usually more than 0.5 mm, and has irregular margin, often be seen in the gross appearance, it is also been called "map-like" necrosis.
Immunochemistry. All immunohistochemical stains were performed on the Ventana Benchmark automated staining system using 4 mm paraffin tissue sections. The primary antibodies used in this study including CD68, S100,CKpan, cam5. follicular DCs and IDC. IDC reside in t-zones in the peripheral lymphoid tissues, such as the paracortex and deep cortex of lymph nodes. Dendritic neoplasms are rare hematological malignancies with less than 1% occurrence in the lymph nodes or soft tissues. IDCS can occur in lymph node, nasopharynx, small intestine, mesentery, spleen, testis, skin, tonsil, bladder, eyelid, uterine cervix and so on [bib_ref] Interdigitating dendritic cell sarcoma presenting simultaneously with acute myelomonocytic leukemia: report of..., Jiang [/bib_ref] [bib_ref] Small intestine perforation due to metastatic uterine cervix interdigitating dendritic cell sarcoma:..., Mahamid [/bib_ref] [bib_ref] Interdigitating dendritic cell sarcoma of urinary bladder mimicking large intravesical calculus, Rupar [/bib_ref] [bib_ref] Primary intratesticular spindle cell tumors: interdigitating dendritic cell tumor and inflammatory myofibroblastic..., Nistal [/bib_ref] [bib_ref] Interdigitating dendritic cell sarcoma of the eyelid with a rapidly fatal course, Boldin [/bib_ref]. Usually, IDCS had an indistinct border, astoriform pattern of spindled to ovoid cell and abundant slightly eosinephillic cytoplasm. Small to large multinucleate cells appeared with distinct nucleoli. IDCS had a moderate cytologic atypia and a low mitotic rate. Necrosis was less seen 4 . In our overview, we found that IDCS had epithelioid cells and mononucleated or multinucleated tumor giant cells, a high mitotic rate, diffuse necrosis and granuloma [bib_ref] Interdigitating dendritic reticulum cell sarcoma: cytologic, histologic and immunocytochemical features, Jayaram [/bib_ref] [bib_ref] Granuloma-like interdigitating dendritic cell sarcoma: report of a case, Mao [/bib_ref] [bib_ref] Interdigitating dendritic cell sarcoma of lymph node mimicking granuloma: a case report..., Ye [/bib_ref]. Patients usually present with painless lymph node enlargement or extranodal masses. Systemically atypical symptoms include fever, weight loss, fatigue and night sweats. IDCS were often misdiagnosed as carcinoma, melanoma, sarcoma, and so on [bib_ref] Dendritic cell and histiocytic neoplasms: biology, diagnosis, and treatment, Dalia [/bib_ref] [bib_ref] What is your diagnosis? Metastatic sarcoma with interdigitating dendritic cells in lymph..., Macak [/bib_ref] [bib_ref] Interdigitating dendritic cell sarcoma and follicular dendritic cell sarcoma: histopathological findings for..., Ohtake [/bib_ref]. There was diversity in histopathological morphology. Immunohistochemically, there is no specific marker for IDCS. Therefore, the diagnosis of IDCS was difficult and achieved by excluding other diseases. IDCS cells were consistently positive for S100 protein and vimentin, weakly positive for fascin, CD68, lysozyme and CD45. The tumor cells were negative for CD21, CD23, CD35 and CD1a, CKpan, cam5.2, EMA, HMB45, Melan A, CD3, CD20, Desmin, SMA, and so on. However, in case 2, the tumor cells expressed CKpan focally, which needed further investigation. Ultrastructurally, the cells demonstrated complex features of interdigitating cells. Birbeck granules and melanosomes were not seen from other studies 4, 18 . Patients have variations in prognosis (some patients received operation and chemotherapy and radiotherapy, but PFS was no more than 1 month, some patient had no treatment after surgery but PFS lasted for years) [bib_ref] Interdigitating dendritic cell sarcoma of the eyelid with a rapidly fatal course, Boldin [/bib_ref]. The extreme rarity of IDCS impeded researches about the prognostic factors for the diverse prognosis. Pas 21 et al. reported necrosis and mitosis were unconfirmed prognosis factors. Therefore, we proposed a pooled analysis for IDCS for the histopathological prognosticators. Our analysis showed median PFS of diffuse necrosis was 3 months but patients with no necrosis or focal necrosis had not observed median PFS, with a significant difference. Caner Saygin 22 et al. studied 25 cases of IDCS and reported that necrosis was not an indicator of poor prognosis. It was not conflict with this study, because we divided the necrosis into focal necrosis and diffuse necrosis and the diffuse necrosis indicated the poorer prognosis.
# Discussion
Our analysis indicated high mitotic count (≥5/10 HPF) and the nuclear atypia were not associated with the poor prognosis the results were in accord with previous report 21 . We also found median age has not any significant differences between tumor lesion, cellular atypia and necrosis. However, the patients with nuclear mitosis <5/10HPF were younger than the patients with nuclear mitosis ≥5/10HPF. The significance is needed to be confirmed further. Patients with diffused lesion had a shorter PFS. These results indicated diffuse lesion had a worse prognosis and were identical with the reported in literature [bib_ref] Interdigitating dendritic reticulum cell sarcoma: cytologic, histologic and immunocytochemical features, Jayaram [/bib_ref] [bib_ref] Interdigitating dendritic cell sarcoma: case report with review of the literature, Zhou [/bib_ref].
However, there are some limitations of our analysis. Our data stemmed from the published literature and we could not obtain complete clinical, pathological and follow-up data for some patients. It is hard to compare IDCS patients with uniform treatments. Therefore, it was necessary to implement a consolidated and well-designed prospective design in further study.
In conclusion, we found that the diffuse lesion and diffuse necrosis of IDCS was associated with a worse biological behavior and prognosis, and the cellular atypia and mitosis had no distinct relevance with the biological behavior and prognosis of IDCS.
[fig] Figure 1: The section of case 1 was solid, white-grayish, fine and smooth, no bleeding and necrosis (A). Case 2 presented smashed white-grayish lymph nodes and the section was solid, white-grayish, focal black, rough and patchy necrosis (B). [/fig]
[fig] Figure 2: Histologic sections of case 1 displayed the spindle tumor cells disposing in fascicles with a focal storiform pattern (A). Neoplastic cells had irregular nuclei, moderately to prominent nucleoli, and notable mitoses (B). The tumor cells positive for CD68 (C), S100 (D), Vimentin (E) and Ki-67 index 80% (F). [/fig]
[fig] Figure 3: Histologic sections of case 2 displayed the sheets of atypia of epithelium-like cells and patchy necrosis (red arrow) (A). The tumor cells were spindle to oval, had abundant cytoplasm and prominentnucleoli (B). The tumor cells expressed CD68 (C), S100 (D), CKpan focal positive (E) and Ki-67 index 30% (F). [/fig]
[fig] Figure 4: Electron microscopic sections presented the mutual interlaced long finger bumps and small amount of rough endoplasmic reticulum in cytoplasm. [/fig]
[fig] Figure 5: Median age has not any significant differences between tumor stage (A), nuclear abnormalities (B) and necrosis (D) (p > 0.05). Median age was 49 years old among patients with nuclear mitosis <5/10HPF, 12 years younger significantly than patients with nuclear mitosis ≥5/10HPF (C) (p = 0.011). [/fig]
[fig] Figure 6: Patients with diffused stage had median PFS of 3 months, in contrast, patients with localized stage had not reached median PFS (A). Nuclear abnormalities and mitosis were not associated with PFS (B,C). Median PFS of diffuse necrosis was 3 months but patients with no necrosis or localized necrosis had not observed median PFS, with a significant difference (D). [/fig]
[fig] Figure 7: Tumor stage was associated with a shorter OS significantly (A). Nuclear abnormalities and mitosis were not related with OS (B,C). Patients with no or localized necrosis did not get median OS, compared with patients with diffused necrosis having a 9-month median OS (D) (p < 0.001). [/fig]
[fig] Figure 8: Pelvic CT scanning in case 1 displayed the lymph nodes aside the left external iliac blood vessels enlarged, with the biggest size of 1.8 × 1.4 cm (A). PETCT in case 1 showed a high bone salt metabolism in right coracoid and left sacroiliac joint (B). Chest CT scanning in case 2 displayed multiple lymph nodes enlarged in right hilar and mediastinal (C). Abdomen CT scanning in case 2 displayed a clear borderline mass in the right kidney (red arrow) and a slightly larger spleen (yellow arrow) (D). [/fig]
[table] Table 1: General characteristics of included cases in the pooled analysis.Electron microscopic analysis. The tissue was postfixed in buffered osmium tetroxide, dehydrated in ethanol and embedded in Spurr's resin. Ultrathin sections were prepared, stained with uranyl acetate-lead citrate and examined using a transmission electron microscope.ISH and PCR.In-situ hybridization was conducted through Epstein-Barr virus (EBV)-encoded small RNA assay (EBVit; catalogue number ISH-5021; OriGene Technologies, Inc., Rockville, MD, USA). Polymerase chain °C for 10 min. BRAF V600E mutation analysis was performed to detect a T > A transversion of nucleotide 1799 of the BRAF oncogene. Tissue from the corresponding unstained slide was lysed, and the genomic DNA was purified from the paraffin block. Real-time PCR using a single primer set was used to amplify the region of the BRAF gene, which contained the mutation site, and two different fluorogenically labeled probes were used to detect the wild-type and V600E mutant. analysis. All analyses were conducted by SPSS 17.0 (IBM. Inc). The difference of age between tumor stage, nuclear abnormalities and nuclear mitosis was estimated by Mann Whitney test. Age difference between necrosis was analyzed by Kruskal-Wallis test. Kaplan-Meier survival curves were estimated for PFS and OS. Log-rank test was used to indicate the difference of survival curves. Age, gender, tumor stage and necrosis were analyzed together in COX Hazard Proportional Regression Model to present hazard ration (HR) and 95% confidence interval (95% CI). All tests were two-sided and the significant level was 0.05. [/table]
|
A case report of an unusual mycotic pseudoaneurysm of the ascending aorta
Background: Mycotic pseudoaneurysms of the ascending aorta are a rare and devastating complication of previous cardiac surgery.Case presentation:We present an unusual case of a fungal mycotic pseudoaneurysm secondary to an aortic suture line successfully repaired under deep hypothermic circulatory arrest.Conclusions:Patients with mycotic pseudoaneurysms of the aorta require a multidisciplinary team approach to prevent devastating complications that may occur in these complex surgical cases.
# Background
Mycotic pseudoaneurysms of the ascending aorta are rare, particularly those related to fungal organisms such as Aspergillus. We present an unusual case of a mycotic pseudoaneurysm secondary to an unroofing procedure for anomalous aortic origin of a coronary artery (AAOCA) and discuss the potential treatment options. [fig_ref] Figure 1: Pre-operative imaging Fig [/fig_ref].
## Case presentation
A 48 year old gentleman initially presented with chest pain and exertional dyspnoea and was found to have an aberrant right coronary artery (RCA) arising from the left coronary sinus. He underwent an unroofing procedure to re-site the origin of the right coronary artery in the correct sinus. His recovery was uncomplicated with no signs of infection.
Three months post discharge, the patient presented to ophthalmology with an acutely red painful left eye. The patient was admitted for antivirals and steroids for possible herpes simplex virus (HSV) infection, but despite this treatment he had lost his vision in the left eye by July 2021.
In August of 2021, the patient represented with rapidly deteriorating vision in his right eye and was transferred to a tertiary ophthalmology unit. Corneal scrapings demonstrated Aspergillus Fumigatus and Saccharomyces Cerevisiae and a CT thorax abdomen and pelvis confirmed disseminated fungal infection and fungal endophthalmitis. A 74 mm × 76 mm saccular pseudoaneurysm was found arising from the mid anterior ascending aorta, with a wide 24 × 21 mm diameter orifice and no neck. The pseudoaneurysm was in close apposition to the manubrium and proximal sternal body. The features were in keeping with a large fungating mycotic pseudoaneurysm arising from the previous aortic suture line and aspergillus infection was strongly suspected. The patient was commenced on anti-mycotic therapy and transferred to a quaternary aortovascular centre for re-sternotomy and repair of the ascending aortic pseudoaneurysm.
The patient was placed on cardiopulmonary bypass via the right axillary artery and right femoral vein. During cooling a left anterior thoracotomy was performed in the 5th intercostal space and a left ventricular vent sited via the apex of the heart. Once a core temperature of 24 degrees was reached, the circulation was arrested prior Open Access to resternotomy. On opening the chest, the aneurysm sac had a fungating appearance and breached. A retrograde cardioplegia cannula was placed through the defect in the ascending aorta and the balloon inflated to occlude the opening. The innominate artery was dissected out and snared allowing antegrade cerebral perfusion (ACP) to be delivered via the arterial cannula in the right axillary artery. The distal ascending aorta was then dissected out and a cross-clamp applied allowing the circulation to be restarted after 30 min of deep hypothermic circulatory arrest (DHCA). The infected pseudoaneurysm sac was dissected out and the defect in the ascending aorta closed with a pericardial patch. The patient was admitted to ITU for post-operative monitoring where he made an uncomplicated recovery. He was repatriated back to his local hospital for ongoing ophthalmology care.
# Discussion
Mycotic pseudoaneurysms are often caused by Staphyloccous, Streptococcus or Salmonella rather than fungal infections, and present with nonspecific systemic sepsis like symptoms [bib_ref] Aspergillus pseudoaneurysm post aortic valve replacement, Perdomo [/bib_ref]. Aspergillus infection is rarely seen in an immunocompetent patient such as in the current case. Aortic pseudoaneurysms can arise along anastomotic suture lines or cannulation sites following previous cardiac surgery [bib_ref] Aspergillus pseudoaneurysm post aortic valve replacement, Perdomo [/bib_ref]. Treatment options for pseudoaneurysms in the ascending aorta abutting the sternum are limited and depend on pre-empting the probability of entering the aneurysm on resternotomy. The theatre team must be briefed to expect this outcome and strategies put in place to manage it. The strategy utilised in the current case was to open the sternum under DHCA. The left ventricular apical vent was sited to prevent ventricular distension when the heart fibrillates during cooling [bib_ref] Repair of ascending aorta pseudoaneurysm eroding through the sternum, Hasan [/bib_ref]. We prefer cannulation of the right axillary artery over the right femoral artery due to the reduced risk of organ mal-perfusion or retrograde embolism and the ability to deliver ACP upon snaring the innominate artery [bib_ref] Moderate versus deep hypothermic circulatory arrest for elective aortic transverse hemiarch reconstruction, Vallabhajosyula [/bib_ref].
DHCA with antegrade or retrograde cerebral perfusion is a strategy used to optimise cerebral protection and provides a bloodless field during complex aortic operations [bib_ref] Varying evidence on deep hypothermic circulatory arrest, Gupta [/bib_ref]. There is no consensus on duration and optimal temperature for DHCA with the decision left to the operating surgeon. Moderate hypothermia (> 25C) with ACP is an alternative strategy for aortic surgery and it is proposed to have similar outcomes as DHCA [bib_ref] Moderate versus deep hypothermic circulatory arrest for elective aortic transverse hemiarch reconstruction, Vallabhajosyula [/bib_ref]. This strategy was deemed unsuitable for the current case as it was not possible to quickly instigate ACP following commencement of DHCA due to the need to complete the sternotomy and dissect out the adhered pericardial structures. The role of the anaesthesiologist and perfusionist are critical in minimising the risk of cerebral ischaemia during this period of DHCA without cerebral perfusion. An alternative approach to this scenario of a pseudoaneurysm abutting the sternum would have been to perform trans-femoral aortic endoclamping. This approach takes careful planning, is not appropriate for urgent or emergency cases, requires significant preoperative monitoring, and would not be advisable in patients with pseudoaneurysm near the brachiocephalic trunk [bib_ref] Endovascular aortic clamping for pseudoaneurysms of the aortic root with aortic regurgitation, Maselli [/bib_ref]. It also risks precipitating a rupture of the pseudoaneurysm by applying a radial force on the aortic wall. As such it was not considered in this case despite the potential benefits of not requiring DHCA, reducing cardiopulmonary bypass time and the risk of hypoxic brain injury.
# Conclusion
In patients who present with thromboembolic events or are systemically unwell after bypass surgery, we urge clinicians to have low threshold for suspicion of pseudoaneurysms secondary to bacterial or fungal infections. Patients with mycotic pseudoaneurysms of the aorta abutting the sternum require a team approach to control for significant blood loss and minimise the risk of cerebral ischaemia that may occur on opening the pseudoaneurysm sac. Case specific planning is essential and ultimately made a potentially complex case, straight forward with a positive outcome.
[fig] Figure 1: Pre-operative imaging Fig. 2 A Opening aneurysm sac. B Retrograde cannula inserted to occlude defect to allow aorta to be dissected and cross clamp to be applied. C Patch repair under cross clamp. D Patch repair completed and cross clamp off [/fig]
|
Laryngeal Tube as airway rescue device from prehospital to tracheostomy: case report
# Introduction
Airway management is a priority in the care of any critically ill or injured patient. The insertion of a cuffed tracheal tube is essential to obtain an early and effective control of the airway. However, the attempted insertion of a tracheal tube under direct laryngoscopy is associated to a number of practical problems in pre-hospital trauma care. An extraglottic airway may be the answer in those patients where this simple and common procedure becomes complex and unobtainable.
# Methods
Our case report is on a 54 year old woman victim of a multi vehicle collision brought to a level one Trauma Center emergency department by the Emergency Medical Service. Initial evaluation revealed a Glasgow Coma Scale score of 8 (eyes 1; verbal 2; motor 5) and a fixed-midriatic right pupil which suggested a severe traumatic head injury. The patient didn't show any evident predictable sign for difficult intubation. After oxygen administration and cervical spine immobilization a rapid sequence induction was carried out and intubation failed after three attempts. Subsequently a laryngeal tube (LT) was successfully placed and connected to a transport ventilator. Transfer to the hospital took 20 minutes with SpO2 level of 99% and end tidal carbon dioxide not above 5 kPa.
# Results
The patient was properly ventilated with the LT during all CT scan investigations. Due to the impossibility of endotracheal intubation the patient underwent surgical tracheostomy as suggested by the ENT surgeon consultant to the trauma leader.
# Conclusion
This case suggests that LT could be an important alternative device for airway management in trauma patients when tracheal intubation is not possible either in pre-hospital or in-hospital setting. LT could also be a precious tool to achieve good ventilation and oxygenation from the field to the operating theatre.
|
Use of recombinant activated factor VII for bleeding following resection of mediastinal angiosarcoma
Sarcomatous lesions of the mediastinum usually present as aggressive and multicentre masses often attached to adjoining structures including heart and lungs. A forty one year male diagnosed with sarcomatous lesion in mediastinum presented for biopsy through midsternotomy later confirmed as angiosarcoma on histopathology. Patient bled excessively after surgery and required reopening of the chest. However, bleeding could not be controlled with reopening, blood products and packing of the mediastinal cavity. Bleeding could only be controlled by using recombinant activated factor VII as rescue therapy without any adverse effects.Cite this article as: Arora D, Mehta Y, Trehan N. Use of recombinant activated factor VII for bleeding following resection of mediastinal angiosarcoma.
# Introduction
Angiosarcoma (AS) is an uncommon malignant neoplasm characterized by rapidly proliferating, extensively infiltrating anaplastic cells derived from blood vessels and lining irregular blood-filled spaces. AS of the mediastinum is a rare entity often undiagnosed preoperatively. Resection of the tumor often leads to extensive and refractory bleeding in perioperative period. Therefore, perioperative management of these tumours warrant special treatment. We present a case of AS of mediastinum that was managed with recombinant acti-vated factor VII (rFVIIa) during postoperative period.
## Case report
A 41 years old male presented with chief complaint of gradual onset breathlessness on exertion New York Heart Association (NYHA) II and pain in right lower chest. There was no history of haemoptysis, orthopnea or paroxysmal nocturnal dyspnea. Pain in chest was dull in nature, non-radiating and relieved by taking medication. There was no history of hypertension, diabetes mellitus or any other prolonged illness. There was no history of significant loss of weight and appetite. On examination patient was conscious, cooperative and average built. His pulse rate HSR Proceedings in Intensive Care and Cardiovascular Anesthesia 2011, Vol. 3 was 78 per minute and blood pressure was 120/70 mmHg. There was no significant respiratory distress at that time. Systemic examination revealed decreased air entry on auscultation at the lung bases. Other systems were within normal limits. Routine blood investigations including complete blood count (CBC), renal and liver function tests were within normal limits. His coagulation profile including platelet count, prothrombin and activated partial thromboplastin time were also within normal limits. Chest radiograph revealed bilateral pleural effusion that was later confirmed by ultrasonography of the chest. Echocardiography revealed localized pericardial effusion adjacent to right atrium (RA) with multiple echodense nodular masses seen within pericardial cavity with constriction pathology. There were significant respiratory variations across mitral and tricuspid valve. Inferior venacava (IVC) was dilated and collapsing less than 50% with inspiration. There was mild dilation of RA with no regional wall motion abnormality (RWMA). Computerized tomography (CT) scan thorax revealed large eccentric sarcomatous mass lesion, aneurysmal dilatation of RA and bilateral pleural effusion. Based on these results pleural tapping was done which revealed haemorrhagic fluid. Fibreoptic bronchoscopy was also done. Fine needle aspiration cytology (FNAC) and pleural cytology revealed no active malignant or inflammatory cells. A provisional diagnosis of anterior mediastinal tumor was made and exploratory biopsy via midsternotomy was planned to confirm the diagnosis and informed consent was obtained for the same. Patient was premedicated with lorazepam 2mg and pantoprazole 40 mg a night before the surgery. On arrival in the operat-ing room, pulse oximetry revealed oxygen saturation of 96% on air. After securing venous access, radial artery cannulation was performed. Internal jugular vein cannulation was performed with trilumen catheter (7F). Anaesthesia was induced with sodium thiopentone (3 mg/kg) and fentanyl citrate (5 mcg/kg). Vecuronium bromide (0.15 mg/kg) was administered intravenously to facilitate tracheal intubation. Anaesthesia was maintained with intermittent doses of fentanyl citrate (1 mcg/ kg), vecuronium bromide (0.02 mg/kg), and isoflurane in 50% oxygen in air. Intraoperative monitoring included EKG with ST segment analysis, oxygen saturation, end tidal CO2, temperature and urine output, direct arterial blood pressure, central venous pressures, cardiac output and derived parameters, arterial blood gas analysis, and activated coagulation time (ACT). Transesophageal echocardiography (TEE) revealed large cavity around RA with communication with it and there was homogenous opacity inside the cavity. Heparin sulphate 4 mg/kg was administered to maintain ACT more than 450 seconds before initiating cardiopulmonary bypass (CPB). Moderate hypothermia with cooling upto 28 °C was maintained. Total bypass time was 98 minutes. Aortic crossclamping was not done during dissection of the mass. A large necrotic vascular mass was seen in the anterior pericardium with invasion of RA and right ventricle. During dissection of the mass a rent in RA occurred which was repaired with bovine pericardial patch. Weaning from CPB was uneventful. Heparin was reversed with protamine monitoring the ACT. A total of three units of packed red blood cells (PRBC) and four units of platelet concentrates (PC) and fresh frozen plasma (FFP) were transfused intraoperatively. Haemostasis was done however; patient was bleeding from the rough surgical sur-faces. Tranexamic acid 10 mg/kg as bolus and 1 mg/kg/hr infusion was also started. Thereafter, mediastinum was packed with sponges and only skin layer was closed. Patient was shifted to the postoperative recovery room for mechanical ventilation and monitoring of haemodynamics and bleeding from chest drains. In the recovery room patient was haemodynamically stable however constant bleeding was there from the mediastinum. He bled about 600ml in first and about 300 ml in next 24 hours. Echocardiography did not reveal any evidence of cardiac tamponade at that moment. He was planned for sternal closure on the second postoperative day after stabilisation of bleeding. However, during reopening significant oozing was still there from the surgical sites. Mediastinum was again packed with only skin closure and patient was shifted to the recovery room. Histopathology of the mediastinal sample gave impression of angiosracoma with microscopic findings of highly vascular mitotically active spindle cell neoplasm composed of freely anastomosing vascular channels lined up by atypical cells. He was haemodynamically stable in the postoperative period. However, constant bleeding from the chest drain was still there at 150-200 ml per day for next three days. On sixth postoperative day drainage was 40 ml in 24 hrs. Therefore, he was planned for removal of pack and secondary closure of the chest. During secondary closure, significant bleeding was still there from the raw surgical surface that was not controlled even by adequate haemostasis and electrocautry. Thromboelastography (TEG) was within normal limits. A total of nine packed red blood cells, four units of fresh frozen plasma and four units of platelet concentrate and apheresis were transfused in the recovery room. Keeping in view of the in-tractable bleeding, massive transfusion and second reopening it was decided to administer rFVIIa (Novoseven, Novo Nordisk India Ltd) in the doses of 90 mcg/kg to this patient. After administering recombinant factor seven oozing from the surgical surface settled down. Secondary closure of the chest was done and patient was shifted to the recovery room. Trachea was extubated after four hours and rest of the postoperative course was uneventful. After stabilization and discharge from the hospital patient was referred to oncology unit for further management and he is on a regular follow up for radiation therapy.
# Discussion
Primary angiosarcomas of the anterior mediastinum are rare tumors that need to be considered in the differential diagnosis of primary anterior mediastinal neoplasms. Despite their histologic similarity to angiosarcomas at other sites, primary angiosarcomas of the anterior mediastinum appear to follow a more protracted clinical course than their counterparts in other organ systems (1). Angiosarcomas tend to be aggressive and are often multicentric. Approximately 50% of angiosarcomas occur in the head and neck. These tumors have a high local recurrence rate and metastasis because of their intrinsic biologic properties and because they are often misdiagnosed, leading to a poor prognosis and a high mortality rate. Malignant vascular tumors are clinically aggressive, difficult to treat, and have a reported 5-year survival rate of around 20% [bib_ref] Malignant vascular tumors in young patients, Lezama-Del Valle [/bib_ref]. Some authors reported a mean survival of 14 months after diagnosis of angiosarcoma. Primary heart tumours are extremely rare in clinical practice, occurring at a frequency of 0.11-0.30% in surgical series and only HSR Proceedings in Intensive Care and Cardiovascular Anesthesia 2011, Vol. 3 25% of them are malignant. AS (75%) is the most common primary cardiac malignant tumor and is an extremely rare, rapidly spreading vascular tumor. It is seen more commonly in males than in females, usually presenting between the third and fifth decade of life. Clinical presentation is usually unspecific and can vary depending on the dimension and the site of the tumour. Increasing dyspnoea, peripheral oedema, superior venacava syndrome, lung oedema, arrythmia, weakness, syncope and dizziness are the most common symptoms. Potentially mortal complications are represented by pericardial effusion and thromboembolic events. AS are commonly located in the RA with a high incidence of pericardial involvement. This predilection of the tumor affecting the right heart often produces right-sided congestive heart failure, superior vena cava obstruction and pericardial effusion [bib_ref] Primary sarcomas of the heart, Burke [/bib_ref]. In this case; the tumor was found arising from the RA presenting with pericardial effusion and symptoms of right heart failure. Imaging studies like chest radiograph, computerized tomography (CT) and magnetic resonance imaging (MRI) can reveal the extent and severity of tumor preoperatively. Echocardiography specially TEE confirms the intracardiac extent of the tumor and associated cardiac functions. Histopathology confirms the diagnosis and severity of the disease. Standard treatment includes extensive surgical resection followed by radio and chemotherapy to improve the survival as done in this case. Perioperative management of these case warrant special care due to the presence of huge areas of necrosis and extensive locoregional spread. Pre-operative histological diagnosis and careful evaluation of resectability is recommended. Anaesthetic considerations in patients with tumor involving RA include hypox-emia, low cardiac output, possible right to left shunt, and potential pulmonary emboli. These patient's symptoms can be exacerbated by changes in body position. Optimal anaesthetic management depends on the judicious administration of anaesthetics, avoidance of cardiac depression and maintenance of preload. Invasive haemodynamic monitoring and good venous access for rapid transfusion is necessary in these cases. The primary problem in this case was refractory bleeding which was controlled even by reopening and adequate haemostasis. rFVIIa administered during second reopening helped in haemostasis and closure of chest. rFVIIa (eptacog alpha activated) was introduced to clinical medicine in the 1980s as a prohemostatic agent. It is thought to act locally at the site of tissue injury and vascular wall disruption, by binding to exposed tissue factor and generating small amounts of thrombin that are sufficient to activate platelets. The activated platelet surface can then form a template on which rFVIIa can directly or indirectly mediate further activation of coagulation, resulting in the generation of much more thrombin and, ultimately, fibrinogen to fibrin conversion. Clot formation is stabilized by inhibition of fibrinolysis, due to factor VIIa-mediated activation of thrombin-activatable fibrinolysis inhibitor [bib_ref] Mechanism of factor VIIa-dependent coagulation in hemophilia blood, Butenas [/bib_ref]. Various authors have used rFVIIa safely as a rescue for refractory bleeding after cardiac surgery and found that it is associated with reduced blood loss, rapid improvement of coagulation variables, and decreased need for blood products (7-10). Some authors reported use of rFVIIa in the setting of pulmonary haemorrhage secondary to chemotherapy for choriocarcinoma thereby preventing rapid deterioration in clinical condition and allowed salvage ther-apy [bib_ref] Recombinant factor VIIa in the management of pulmonary hemorrhage associated with metastatic..., Wheater [/bib_ref]. It has also been used as rescue in bleeding associated with obstetrics, hepatic resection, trauma, intracerebral bleed and even dental extractions apart from coagulation disorders [bib_ref] The role of tissue factor and factor VIIa in hemostasis, Mackman [/bib_ref]. However, therapy with rFVIIa may lead to thromboembolic complications (10%) and the patient may not respond. This may lead to fatal complications. Moreover, cost is also a limiting factor in its use. A meta-analysis showed that, in 73% of patients, rFVIIa achieved a reduction of bleeding and the probability of survival increased. rFVIIa was not associated with an increased risk of thromboembolism compared with placebo (13). However, data from 35 randomized controlled trials suggested that rate of arterial thromboembolic events is higher in patients who received rFVIIa than among those who received placebo (5.5% vs 3.2%, P=0.003) [bib_ref] Safety of Recombinant Activated Factor VII in Randomized Clinical trials, Levi [/bib_ref]. In the present case we have found significant reduction in bleeding after administering rFVIIa and secondary closure of the chest was possible. There was no evidence of any thromboembolic or other complications.
In conclusion, rFVIIa may be used as a rescue therapy in patients with malignant vascular tumors in the setting of intractable bleeding following surgical resection.
|
Patient and interest organizations’ views on personalized medicine: a qualitative study
Background: Personalized medicine (PM) aims to tailor disease prevention, diagnosis, and treatment to individuals on the basis of their genes, lifestyle and environments. Patient and interest organizations (PIOs) may potentially play an important role in the realization of PM. This paper investigates the views and perspectives on PM of a variety of PIOs. Methods: Semi-structured telephone interviews were conducted among leading representatives of 13 PIOs located in Europe and North-America. The data collected were analysed using a conventional content analysis approach. Results: The PIO representatives supported the realization of PM but feared that many financial, structural and organizational challenges may delay its realization. They encouraged strategies to modernize drug licencing mechanisms, develop research and data sharing infrastructures, and educate patients and health care professionals in PM. Notably, they emphasized the importance of developing PM in an equitable way and taking into consideration the patients' needs, values and personal situation. Despite varying levels of awareness regarding PM, the PIO representatives expressed willingness to engage in the PM agenda and recommended that PIOs work closely with policy-makers to design PM in a way that truly addresses the needs and concerns of patients. Conclusions: PIOs have the potential to become central drivers of the PM agenda. Collaborations should be further developed between PIOs, researchers, drug developers and health care authorities.
# Background
Recent advances in biomedical research and biotechnology offer new possibilities to tailor prevention, diagnostic and treatment to the specific needs of patients. By utilizing information about the patients' genetic and molecular profile combined with their family history and clinical and environmental data, it is hoped that health care providers will be able to start interventions earlier and select therapies that are more precise, efficient and provide less side-effectsstrategies broadly known as personalized medicine (PM) [bib_ref] Personalizing health care: feasibility and future implications, Godman [/bib_ref]. Large investments are being made worldwide to support such strategies. In the United States, the $215 million Precision Medicine Initiativewas recently launched with the objective to accelerate biomedical discoveries and improve the effectiveness of treatment. In Europe, the European Commission has since 2007 committed over €1 billion of health research funding to the development of '-omics' technologies and targeted therapies, and more funds are expected to be released in the coming years . PM has been extensively described in recent reports from national and international medical organizations and funders. These reports emphasise the importance of engaging a variety of stakeholders such as health care providers, biomedical researchers, regulators, drug developers, patients and patient organizations in driving PM forward. Among these stakeholders, patient organizations are particularly valuable players regarding their role in patient education, assessing new biomedical developments, and supporting research projects. However, with the exception of recent surveys in which patient organizations identified potential barriers in PM implementation [bib_ref] An index of barriers for the implementation of personalised medicine and pharmacogenomics..., Horgan [/bib_ref] and key priority areas within cancer , little empirical data are available to document the views and perspectives of patient organizations on PM.
Investigating how patient organizations perceive PM and which role they are willing to play in its realization is important for several reasons. First, patient organizations have close contact with the communities of patients they represent. They are therefore well-placed to provide qualified opinions regarding how to implement PM in a way that addresses the needs and concerns of these communities. For instance, PM may require that patients learn about their genetic risk profile, engage more actively in the medical decision-making process and share their health data with researchers and clinicians more broadly than conventionally practised [bib_ref] Ask not what personalized medicine can do for you -ask what you..., Budin-Ljosne [/bib_ref]. Patient organizations are well-qualified to familiarize patients and communities with such developments. Second, patient organizations may provide useful guidance in addressing potential ethical and societal challenges that may arise when new medical approaches are implemented. For instance, PM may imply that patients are offered differential access to treatment depending on their genetic profile [bib_ref] The ethical framing of personalized medicine, Joly [/bib_ref]. Patient organizations may be well placed to explore how dialogue with patients and communities should take place when access to conventional treatment may be restricted. Third, patient organizations have comprehensive expertise within the areas of patient and health care professional capacity building [bib_ref] The role of patient advocacy organizations in shaping genomic science, Koay [/bib_ref] [bib_ref] Generating sociability to drive science: patient advocacy organizations and genetics research, Panofsky [/bib_ref] , the development of online patient communities [bib_ref] Involvement of Patient Organisations in Research and Development of Orphan Drugs for..., Mavris [/bib_ref] , the design of regulatory frameworks and patenting policies [bib_ref] Gene patenting and licensing: the role of academic researchers and advocacy groups, Ledbetter [/bib_ref] , the financing and management of clinical trials and biobanks [bib_ref] Involvement of Patient Organisations in Research and Development of Orphan Drugs for..., Mavris [/bib_ref] [bib_ref] How disease advocacy organizations participate in clinical research: a survey of genetic..., Landy [/bib_ref] [bib_ref] Patients with rare diseases work to jump-start research, Marcus [/bib_ref] [bib_ref] Financial stability in biobanking: unique challenges for diseasefocused foundations and patient advocacy..., Bromley [/bib_ref] , and more recently, in developing databases and web-based platforms for patient-driven data sharing [bib_ref] The role of patient advocacy organizations in shaping genomic science, Koay [/bib_ref] [bib_ref] Huntington's disease: advocacy driving science, Wexler [/bib_ref]. Making use of these skills and experiences may accelerate the realization of PM.
Although patient organizations are often referred to as one homogenous group, they are rather a constellation of entities that vary considerably in size, organizational structure, denomination and mandate [bib_ref] The role of patient advocacy organizations in shaping genomic science, Koay [/bib_ref]. For instance, patient organizations may be publicly or privatelyfunded, small entities which represent only a limited network of families suffering from a particular condition, or they may be large-scale, international organizations bringing together national organizations which focus on one specific disease or specialize in improving the living conditions of all patients. We conducted a qualitative interview study among leading representatives of nonprofit patient organizations which are concerned with one specific disease or disease area and define themselves as patient advocacy organizations, umbrella organizations for patient advocacy organizations or interest organizations, hereafter referred to as patient and interest organizations (PIOs). The objective of our study was to collect information about three main issues: 1) the PIOs' views and perspectives on PM, 2) recommendations for facilitating the realization of PM, and 3) the role they foresee that they may play in the realization of PM.
# Methods
## Recruitment of study sample
To collect information about the views of a variety of PIOs, we proceeded in two steps. First, in July 2014, we sent an e-mail invitation to the leaders of a small number of PIOs that co-operate with the research network of the Norwegian Cancer Genomics Consortium (NCGC), a national research platform aiming to establish personalized clinical strategies for cancer treatment in Norway. The email invitation provided an outline of the study objectives and a description of what participation in the study entailed. In the invitation, PM was described as "an emerging practice of medicine that uses an individual's genetic profile to guide decisions made with regard to the prevention, diagnosis, and treatment of disease" as defined in the Talking Glossary of Genetic Terms of the National Human Genome Research Institute. The email invitation included a request for written informed consent. The study was approved by the Norwegian Social Science Data Services.
The first PIO leaders who accepted our invitation helped us identify other PIO representatives that might want to participate in the study. Through such snowball sampling, we were able to recruit representatives from 8 PIOs concerned with one specific disease or disease area (6 PIOs in Norway, 1 in the United Kingdom and 1 in the USA) between July 2014 and January 2015. In parallel and in order to extend our sampling, we sent email invitations to 20 leaders of disease-specific PIOs members of the European Patients' Forum (EPF), an umbrella organisation of pan-European patient advocacy organizations. The PIO representatives were recruited in the study until a point of saturation was attained, i.e. when no significant new information was collected. Five PIO representatives agreed to join the study, six responded that they did not have time to participate or did not want to participate, and nine did not respond to our invitation. In total, thirteen PIOs working within the areas of cancer (4), hereditary and genetic disorders (3), mental health (1), diabetes (1), psoriasis (1), AIDS (1), lupus (1), and primary immunodeficiencies (1) participated in the study [fig_ref] Table 1: PIOs participating in the study [/fig_ref]. The representatives interviewed were organizational leaders (10) (e.g. CEO, secretary general, director) and senior managers (3) employed by the PIOs. As leading representatives of their PIOs, they were able to speak on behalf of their organization although many specified that their organization did not have an official position on PM. The size of the PIOs varied from small PIOS gathering families suffering from a rare genetic disorder to large PIOs gathering more than a thousand organizational members.
## Data collection and analysis
Semi-structured telephone interviews were conducted with the thirteen PIO representatives, lasting forty minutes on average. An interview guide was used to lead the conversation, which included open-ended questions about the PIOs' perspectives regarding PM, PM-related activities, perceived challenges with regard to the realization of PM, recommendations for the adoption of PM, and the potential roles the PIOs may play in the realization of PM. The interviews were conducted in English or Norwegian by the first author trained in qualitative methods and fluent in both languages. The interviews were audio recorded and transcribed verbatim in English and Norwegian. The transcripts were analysed manually by the first author using a qualitative content analysis approach in an inductive way [bib_ref] The qualitative content analysis process, Elo [/bib_ref]. First, a thematic analysis of the interview texts was conducted to identify overarching themes that emerged from the responses to each open-ended question formulated in the interview guide. Next, the substantive content of the text was extracted, coded and categorized according to the overarching themes. If new codes were identified, the coding frame was updated accordingly. Then, the text pertaining to each of these themes was condensed to reflect the main points raised by the PIO representatives. In August 2015, a two-page report summarizing the main findings was sent by email to the PIOs representatives for validation. Their comments and corrections to the report were integrated in the results.
# Results
The PIO representatives described their interest in PM and potential challenges that may impede or delay the realization of PM. Particular emphasis was put on possible side effects of focusing on PM strategies. Then, the representatives made a number of recommendations for the realization of PM and discussed how their organizations may contribute to the PM endeavour.
## Pios interest in pm
Overall, the PIO representatives expressed interest in PM. They explained that the current medical needs of their patient groups are largely unmet and medical strategies which address issues of side effects, overtreatment and undertreatment are strongly needed. As outlined by this representative:
(1) If you create a medicine that directly targets the individual, you will then be able to avoid all these unsuccessful attempts to find the right treatment without getting better, having to deal with the side effects, the downturns, the disappointments (…) when you use a medicine that has side effects for a long period, you are sick for months and then it turns out that [the medicine] does not help anyway. Developing medicines that will perhaps work at once and not after the fifth attempt, (…) we see this as perhaps the most important thing to achieve.
However, several representatives confessed that PM was a topic that had not been thoroughly discussed in their organization. Most PIOs did not use the terms PM and did not specifically mention PM in their strategic documents with the exception of some cancer PIOs. Several representatives also explained that the concept of PM and its possible practical implications were still unclear. As an illustration, one representative believed that PM meant that health care professionals allocate more face-to-face time with each patient.
## Pm challenges high cost of pm
Most representatives feared that PM may be too expensive for many health care systems which are currently dealing with significant financial constraints. They experienced that patient access to conventional treatment is increasingly restrained due to cost issues, and observed that new targeted drugs that are launched on the market are so highly-priced that patients can hardly afford them unless their cost is fully covered by payers. As expressed by this representative:
(2) If we end up with the same level of cost to get an authorization for a targeted drug that would effectively treat 10% of a population of people with a given disease, rather than being used to treat a 100% of the population with a given disease, then the cost for (…) personalized medicine per patient treated will have to be higher.
Although the representatives acknowledged PM's potential to save costs as treatment becomes more targeted and waste is avoided, they expected PM to require significant up-front capital investments in equipment and infrastructures that most countries cannot afford. They also suspected that PM may lead to increased medical followup as genetic tests are more frequently used in clinical settings to confirm a diagnostic. Another concern was that patients who purchase genetic testing kits from direct-toconsumer genetic testing companies may seek medical advice from their physician to interpret the results, thus creating additional burdens on health care services.
## Lack of health literacy
Several PIO representatives expressed scepticism regarding the patients' ability to understand and endorse PM strategies. They noted that patients often struggle to understand basic medical information:
(3) People do not understand information well enough, this is where we see a big job for the patient organizations and patient groups in each of the countries in Europe, (…) educate, bring awareness of how important (…) the different kinds of medications are, when they are supposed to be used and for what, (…) there is a reason why it has been prescribed (…). There is a huge communication aspect that we have just picked up.
Another representative suspected that many patients are not ready to learn about their genetic profile, in particular if unexpected findings are discovered:
(4) Genetic testing for rare genetic conditions and other disorders can now be performed in hospitals, you can get some answers that are incidental and that no one asked for and for which a decision must be made. We think that this is a challenge (…), the patients are not aware of this. They can get a lot of information that is hard to digest.
These views were however nuanced by the perspectives of other representatives who emphasized that patients who have been waiting for a diagnostic for many years are often eager to learn about their genetic profile. Several PIO representatives also explained that general practitioners often do not understand the specificities of disease, and suspected that many are insufficiently trained in genetics to use PM strategies in their medical practice. Similarly, several representatives explained that policy-makers often do not understand why patients need specific types of treatment rather than a "one-sizefits-all" treatment.
## Lack of mechanisms and infrastructures to support pm
Several PIO representatives explained that current drug licensing mechanisms unnecessarily delay the launching of personalized treatments. Receiving drug approval from regulatory authorities often takes years and is particularly cumbersome if such approval is needed across borders. Although the representatives largely believed that data protection regulations are needed to protect the rights and interests of patients, some representatives suspected that such regulations may be too stringent and hinder the conduct of biomedical research. As expressed by this representative:When it comes to access to this information in relation to biobanks, there are so many rules already dealing with this, I know that many are concerned about this (…), but what is it we are scared about? There are rules for how [information] can be managed and used, I do not see any danger that [information] will be misused.
## Unwanted consequences of pm
In general, the PIO representatives worried that PM, because of its high costs, may reinforce already existing health care disparities among social groups unless specific measures are taken by policymakers to enable equitable access to PM. Some representatives also expressed concerns that patient groups may be forgotten if they represent a genetic subset that is not considered lucrative by pharmaceutical companies. As expressed by one representative:If it is really individualized, if it's really about genetics, there will be populations that will be marketwise less interesting, maybe population-wise they are, meaning that you should not only look at the benefit for the rich white community but also for the populations that are maybe research-wise less interesting to include but might really benefit from more research.
Some representatives feared that the increased use of technology in health care may lead professionals to reduce patients to their genetic profile. As described by this representative:
(7) I think it is essential that the patient remains at the center and that he doesn't become a data or an entity; it's really a patient and a face and the [health care professionals] need to treat that patient, that's the main goal.
Genetic discrimination was a shared concern of some representatives. An example was provided of patients being denied health insurance because of their medical history even if the law prohibits such practice. As expressed by one representative:
(8) If we, sometime in the future, find genes that you would like to have and genes you would prefer not having, it is clear that it may contribute to creating A-people and B-people. From the moment our genes become a question of value, it may create differences between people. Some people may be worth more than others and then we are back to the 30's, so from a historical point of view, this is important.
Finally, one representative worried that priority may be given to genetic research, not because it may lead to the most useful results, but because it is technologically exciting and may benefit the commercial interests of pharmaceutical companies. This representative emphasized the importance of also conducting other types of research such as social and behavioural research.
## Main recommendations for the realization of pm
The PIO representatives made recommendations for the realization of PM that can be assigned to five main categories as described below and summarized in [fig_ref] Table 2: PIOs' recommendations for the realization of PM Recommendations Policy-making • Establish principles... [/fig_ref].
## Policy-making
Overall, the representatives emphasized the importance of providing equitable access to PM. They explained that access to genome sequencing technologies and targeted drugs should not be limited by financial constraints but offered according to specific priority criteria that are jointly established by policymakers, health care professionals and patient representatives. They believed that modernizing drug licensing mechanisms is necessary to enable quicker access to targeted drugs. Adaptive pathways mechanisms were mentioned as a potential approach to improve timely access for patients to new medicines. The representatives also recommended allocating specific funding to the implementation of PM, for instance to develop necessary biobanks and data sharing infrastructures. Some representatives suggested
## Patient-centered health care
Most representatives emphasized the importance of an open dialogue between health care professionals and patients. They believed that patients should be provided with information that is simple, understandable, actionable and adapted to the patient's needs, values, level of health literacy and way of dealing with disease. They explained that the value of an intervention should not only be measured in medical, biological or scientific terms but also in terms of how it impacts the patients' quality of life and ability to function in society, and recommended that these aspects are taken into consideration when assessing patient needs. They strongly recommended providing genetic counselling to patients when genetic information is available. In general, the representatives believed that the medical literacy of patients should be improved. However, they explained that patients differ in their willingness to learn about their disease and engage in their health care, and that their personal preferences, for instance regarding access to their genetic information, should be respected. In addition, the representatives considered it important to educate health care professionals in PM strategies. In contrast, educating the general public about PM was seen as useful but not an absolute priority. The representatives believed that public awareness about PM may gradually develop as targeted treatments are becoming more broadly available in health care.
## Increased, inclusive research
In general, the representatives believed that more research should be conducted to understand the causes underlying disease, investigate the mechanisms of side effects, explore the consequences of living a long time with a specific treatment and develop strategies to improve the quality of life of patients. They explained that patients are willing to participate in medical research but are often denied such opportunity because of strict eligibility criteria. They recommended engaging patients and patient representatives as early as possible in the planning and design of clinical trials and broadening eligibility criteria to include more varied populations across socio-economic groups.
## Data sharing and protecting privacy
The representatives supported the development of biobanks and data sharing infrastructures that are governed by solid and transparent privacy protection frameworks as these are essential to maintain the trust of patients and protect their interests. Privacy protection was seen as particularly important when the health data that are shared originate from patients suffering from socially stigmatizing diseases. The representatives explained that more work is needed to make people comfortable with sharing their personal health data.
## Pios' active participation in agenda setting
The representatives believed that PIOs should be involved as early as possible in the planning and design of the PM agenda. They explained that the main role of PIOs in PM includes helping to: a) increase literacy in PM among patients and health care professionals, b) identify the areas in which PM is most needed, and c) support the development of research projects that are attractive to patients. The representatives envisioned increased collaboration between PIOs and research groups, medical bodies, drug developers and policy-makers, nationally and internationally to encourage the development of PM. They also believed that patients and lay representatives should be more frequently invited to join research ethics committees and drug approval boards. The representatives recommended the development of inter-PIOs collaborations to enable the organizations which are more knowledgeable about PM to help others join the PM endeavour. Finally, they called for initiatives to educate PIOs in PM as finding the right information about PM may be challenging and many PIOs do not have a clear understanding of how they can contribute to the successful realization of PM.
# Discussion
Our findings show that PIO representatives had a positive although cautious attitude toward PM. They believed that PM is needed but suspected that many financial, structural and organizational challenges may delay its realization. Although the recommendations forwarded by the representatives to support the realization of PM are largely congruent with those made by the promoters of PM, they also shed light on specific ethical and societal challenges that may arise in the process of adopting PM. Importantly, attention to these considerations has not been emphasized in the public debate but they should be addressed to enable the successful realization of PM in a way that truly addresses the needs and concerns of patients. Developing PM in an equitable way, and involving discussions with decision-makers about how such equity may be achieved, was seen as the main priority. The PIO representatives based their rationale on the types of problems that their patients encounter daily in health care systems. For instance, the prices of new targeted drugs have steadily increased during the last decade [bib_ref] Trends in the Cost and Use of Targeted Cancer Therapies for the..., Shih [/bib_ref] and health care systems frequently deny patients access to such drugs because of financial constraints unless the patients can pay out-of-pocket [bib_ref] Big pharma moves from 'blockbusters' to 'niche busters', Dolgin [/bib_ref]. The PIO representatives expressed concern that issues of social injustice may arise if individuals who are more socioeconomically advantaged benefit most from new treatments because they have the financial ability to pay for them. The fundamental principle of equitable access to health care may be threatened when "niche buster drugs" designed for smaller groups of patients are more expensive and therefore less accessible than blockbuster drugs [bib_ref] Big pharma moves from 'blockbusters' to 'niche busters', Dolgin [/bib_ref]. This issue has already been raised by coalitions of patient organizations which have called for an equitable and universal patient access to PM independent of the patient's socio-economic status and geographical location [bib_ref] An index of barriers for the implementation of personalised medicine and pharmacogenomics..., Horgan [/bib_ref].
Although the PIO representatives emphasized that equitable access to PM was critical, they did not articulate in detail what such equity should entail. Equity is a broadly acknowledged ethical concept that is concerned with the fair and impartial "distribution of benefits and costs to distinct individuals or groups" [bib_ref] Ethical Issues in Resource Allocation, Research, and New Product Development, Brock [/bib_ref]. Although most health care systems strive to achieve equity, it has shown to be difficult to operationalize. For instance, it is usually considered as equitable to allocate health care resources on the basis of consideration of the actual needs of patients [bib_ref] Justice and Access to Health Care, Daniels [/bib_ref]. However, how to determine which needs are most important, and how they should be prioritized is often challenging. As an illustration, should the needs of terminally ill patients be addressed in priority, or those of patients who may yield significant gains in life expectancy if treatment is provided [bib_ref] Just caring: assessing the ethical and economic costs of personalized medicine, Fleck [/bib_ref] ? Two PIO representatives mentioned that patients for whom no standard treatment exists, for instance some groups of cancer patients, should be prioritized to receive access to genome sequencing in the hope that a potential treatment may be found. However, neither the importance nor the implications of providing genome sequencing to cancer patients if little is known about the genes involved in their cancer types are well understood. Other criteria may also have to be taken into consideration for resource allocation such as the efficacy and effectiveness of the intervention (its ability to lead to a positive outcome for the patient) and its cost-effectiveness (ability to provide value for money) [bib_ref] From efficacy to equity: Literature review of decision criteria for resource allocation..., Guindo [/bib_ref]. Even if it is seen as morally justifiable to give priority to the patients who are in greatest need, it may not be wise from a costeffectiveness perspective if the intervention is not expected to benefit the patient. Decision-makers are faced with having to balance between these different considerations (equity, efficacy, cost-effectiveness) and make decisions that do not unjustifiably discriminate some groups of patients. In the absence of a clear normative framework providing guidance on how to distribute resources fairly [bib_ref] Criteria for fairly allocating scarce health-care resources to genetic tests: which matter..., Rogowski [/bib_ref] , it may be useful to investigate processes that are deliberative, transparent, and "appeal to rationales that all can accept as relevant to meeting health needs fairly" [bib_ref] Accountability for reasonableness, Daniels [/bib_ref]. This is where PIOs may play an important role by collaborating with decision-makers in order to find pragmatic solutions for an equitable and reasonable resource allocation. As an illustration, the Scottish Medicines Consortium convenes clinicians, health economists and patient representatives to identify and prioritize those medicines and interventions which represent good value for money to patients and should be launched on the market. Such collaboration may be good example of how a fair process may take place.
Taking into consideration patients' interests and values to a greater extent than traditionally practised was also seen as critically important by the PIO representatives. PM is often described as a holistic approach to medicine that utilizes information about the patient's clinical and family history, genetic susceptibility to disease, response to certain drugs, and lifestyle to assess the patient's health. The PIO representatives interpreted the meaning of "personalized" medicine slightly differently from the promoters of PM. Although the PIO representatives largely agreed that focusing on the biological and environmental factors that may affect the health of the patient is important, they also believed that "personalizing" health care means that greater attention should also include taking the time to listen to the patient, learn about her values, assess her mental and spiritual well-being, understand her personal circumstances, and improve her quality of life and ability to function in society with a condition. The views expressed by the PIO representatives are in concordance with recent calls for a more "humanistic" medicine that approaches the patient as a whole person with a personal history, emotions, beliefs and sufferings. A humanistic doctor takes the time to listen to the patient's story, shows compassion and empathy, makes sure that the patient feels valued and respected, and takes into consideration the patient's personality and experience in the process of medical care. As the PIO representatives repeatedly emphasized during our interviews, no technological advancement or biological instrument can substitute the importance of the doctorpatient relationship.
PM aims to strengthen dialogue and collaboration between doctor and patient through increased patient engagement. In accounts of PM, patients are encouraged to discuss various treatment options with their healthcare provider, communicate about their preferences, values and lifestyle, and provide feedback on health care processes and outcomes. At such, patient engagement may be a critical strategy to enable the provision of health care that is responsive to the patient's specific needs and situation as envisioned by the PIO representatives. Patient engagement enables a better power balance between patients and doctors, thus making it easier for patients to express their concerns and preferences, and bring perspectives on their own care. However, while patient engagement strategies may work for those patients who are eager to take responsibility for their health care, seek knowledge and understand the details of their clinical situation, they may not be suitable for those who prefer to delegate the majority of the responsibility for their health care to their health care professional. Similarly, some patients may not have the ability to engage in their health, for instance because they lack the necessary health literacy to understand what is at stake [bib_ref] Health literacy in Europe: comparative results of the European health literacy survey..., Sorensen [/bib_ref]. In this context, the PIOs that have available resources and are willing to engage in PM may have an important role to play in enhancing the health literacy of patients and familiarizing them with engagement strategies. Other patients may be so physically and mentally ill that they do not have the capability to engage. For these patients, providing health care with compassion and empathy may be particularly important; time should be allocated in the PM agenda to enable health care professionals to interact in a "humanistic" manner with their patients.
Finally, our study results indicate that levels of awareness regarding PM vary considerably between PIO representatives. In general, the representatives from the larger PIOs were more familiar with the concept of PM and worked more mindfully toward PM than representatives from smaller PIOs. Similarly, cancer PIOs were more likely to integrate the concept or aspects of PM in their strategy independent of their organizational size. Larger PIOs may have more resources to investigate new medical strategies than smaller PIOs which struggle with basic issues such as guiding their patients through the health care bureaucracy or educating health care professionals. Engaging in PM may be more challenging when the organizations' resources are limited. Similarly, the general attention given to cancer in the PM discourse and the fact that most targeted drugs that have been marketed are cancer drugsmay explain why cancer PIOs were generally more aware of PM than non-cancer PIOs. As an illustration, several representatives from non-cancer PIOs believed PM to be primarily focusing on cancer to the disservice of other disease areas; a belief previously observed among health care professionals [bib_ref] Re-focusing the ethical discourse on personalized medicine: a qualitative interview study with..., Schleidgen [/bib_ref] and patient representatives [bib_ref] An index of barriers for the implementation of personalised medicine and pharmacogenomics..., Horgan [/bib_ref]. It is therefore important that decision-makers work to increase levels of awareness regarding PM among PIOs, and provide support to those PIOs that have limited resources and competency but are willing to contribute to the PM agenda.
# Conclusions
The development and realization of PM requires the involvement of PIOs. Historically, PIOs have shown an impressive ability to engender important changes on a large scale that benefit patients. If the PIOs become sufficiently convinced that PM is the future of medicine and will benefit their patients, then they have the potential to play a significant role in driving the PM agenda forward. It is therefore important that researchers, health care funders, drug developers and policy makers invite PIOs to the table, not only the biggest and most influential PIOs but also the smaller ones should be engaged so that PM is developed in ways that address the "health care needs of patients. It is also important that PIOs work together to become effective advocates of PM. The European Patients' Academy (EUPATI), a pan-European platform for patients and PIO engagement, may be a potential springboard for intra-PIO collaborations. The European Alliance for Personalised Medicine (EAPM), a coalition bringing together European healthcare experts and patient advocates, may also be one of the central arenas where PIOs could discuss with key stakeholders issues related to priority setting and resource allocation. Such models of collaboration and engagement are needed to give PIOs an increased opportunity to contribute to the decision making process regarding the design of PM.
# Study limitations
The results from our study provide insights into how PIOs working within a range of diseases perceive PM and reflect upon the PM agenda. Deciding on which PIOs to include in our study was challenging: there are thousands of active PIOs and they vary widely in organizational structure, areas of activity and funding. It was difficult to learn about the PIOs organizational and financial structure on the basis of the information they provide on their web site. In the absence of obvious criteria to select PIOs, we decided to restrict our sampling to non-profit PIOs that were disease-specific and were either member of the European Patients' Forum, or were within the network of the Norwegian Cancer Genomics Consortium. This was a pragmatic choice that enabled us to come in contact with a variety of PIOs. However, our sampling may not be representative of the wider spectrum of PIOs that exist, and the PIOs situation (e.g. relationship to biopharmaceutical companies) could have to some extent influenced their interest in and level of knowledge about PM.
For many of the PIO representatives, this study was the first time they discussed PM and they were therefore not able to provide any concrete examples of issues related to their experience with PM, or specific recommendations regarding their potential role in the realization of PM. More work may be needed to investigate how a larger spectrum of PIOs which have specifically endorsed the concept of PM in their strategies, and work actively toward it, envision their role in the realization of PM agenda.
[table] Table 1: PIOs participating in the study [/table]
[table] Table 2: PIOs' recommendations for the realization of PM Recommendations Policy-making • Establish principles and criteria for equitable patient access to PM • Modernize drug licensing mechanisms, e.g. through adaptive pathways • Allocate specific funding to the implementation of PM • Implement PM gradually, e.g. through pilot projects including the most needy patients Invite patients and PIOs early in the planning and design of clinical trials • Broaden eligibility criteria to recruit more patients in research Data sharing and protecting privacy • Develop privacy-solid biobanks and data sharing infrastructures • Ask for permission before using personal health data • Develop flexible and interactive consent mechanisms PIOs' active participation in agenda setting • Engage as early as possible in the development of PM • Contribute to educate stakeholders in PM • Contribute to design the PM agenda, e.g. by identifying priority areas • Support the development of collaborative research projects • Respond to national hearings of relevance for PM • Join research ethics committees and drug approval boards • Develop partnerships with medical professions, e.g. through medical expert panels • Develop partnerships between PIOs nationally and internationally [/table]
|
Central Venous Oxygen Saturation/Lactate Ratio and Prediction of Major Adverse Events After Pediatric Heart Surgery
Introduction: Major adverse events (MAE) are unexpected but undesirably frequent after pediatric congenital heart surgery and contribute to poorer outcomes. The aim of this study was to test the predictive value of a ratio between central venous oxygen saturation and arterial lactate (ScvO2/lactate) for MAE after pediatric congenital heart surgery in a Brazilian university hospital.Methods: We conducted a retrospective observational study in a tertiary care university hospital, including 194 infants and children submitted to surgery for congenital heart disease. The predictive value of ScvO2, lactate, and ScvO2/lactate ratio were assessed by the area under the receiver operating characteristics curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).Results: The incidence of MAE was 16% -cardiac arrest/death, unplanned reoperation, and low cardiac output syndrome were the most common events. Overall, ScvO2/lactate ratio discriminated patients with and without MAE very well (AUC 0.842), performing better than either variable alone, with sensitivity of 48%, specificity of 94%, PPV of 60%, and NPV of 91%.Conclusion: A ScvO2/lactate ratio > 5 can accurately identify patients at low risk of MAE after pediatric congenital heart surgery, with very good specificity and NPV, but poor sensitivity and PPV.AUCCI CPB cTnI IL LCOS MAE NPV PPV RACHS-1 ROC ScvO 2 SD VIS = Area under the receiver operating characteristics curve = Confidence interval = Cardiopulmonary bypass = Cardiac troponin I = Interleukin = Low cardiac output syndrome = Major adverse events = Negative predictive value = Positive predictive value = Risk Adjustment for Congenital Heart Surgery 1 = Receiver operating characteristics = Central venous oxygen saturation = Standard deviation = Vasoactive-inotropic score Brazilian Journal of Cardiovascular Surgery Braz J Cardiovasc Surg 2021;36(6):736-42 Rocha V H S, et al. -ScvO 2 /Lactate Ratio and Major Adverse Events
# Introduction
Major adverse events (MAE) are unexpected but undesirably frequent after pediatric congenital heart surgery and contribute to poorer outcomes. A MAE can be defined as cardiac arrest (with or without extracorporeal life support), emergency chest reopening or reoperation, and death . Identifying patients at risk for MAE is challenging, but it could help physicians and nurses to monitor and allocate more resources to specific patients to prevent or rapidly address and treat a MAE. This has been attempted with the use of clinical examination (capillary refill time, pulses, urine output, core-toe temperature gradient), classical (heart rate, arterial blood pressure, central venous pressure, etc) and more advanced (cardiac index, systemic or pulmonary vascular resistance) hemodynamic variables, and laboratorial tests (lactate, base excess, arterial or central venous already in place or from arterial puncture. The ScvO 2 /lactate ratios were calculated using paired measurements (within the same hour). The first values obtained within 0-6, 6-12, 12-24, and 24-48 hours after surgery were used for analysis, as well as the worst values within the first 48 hours (lowest for ScvOand ScvO 2 /lactate ratio, and highest for lactate). In all patients, only measurements made before the occurrence of MAE were considered in the analysis.
# Statistical analysis
All results were described by means and standard deviations, medians and interquartile ranges, or counts (percentages). The continuous independent variables were categorized, and the best cutoff points for each one were determined by adjusting receiver operating characteristics (ROC) curves. The areas under the ROC curves (AUC), sensitivity, specificity, PPV, and negative predictive value (NPV) were also calculated. The individual contributions of each independent variable to the prediction of MAE were assessed by adjusting a multiple log-binomial regression model.
# Results
A total of 197 patients were eligible for the study, but only 194 (98%) were included. The reason for exclusion of three patients was death at the operation room. The overall incidence of MAE was 16% (31 patients). The demographic, clinical, and surgical characteristics of all patients are depicted in. In short, patients who experienced a MAE were younger, more complex, and had higher VIS at the end of surgery and 12 hours after that. The types of MAE we observed are listed in .
Despite experiencing MAE, some patients also had minor/ moderate adverse events and complications (92 events)pleural effusion, cardiac arrhythmia, reintubation, pneumonia, and mechanical ventilation > 7 days were the most frequent events. They are described in.
The mortality for patients who experienced a MAE was significantly higher than that of patients in the control group (45.1% vs. 6.7%, P<0.001). Patients who experienced MAE had significantly higher lactate levels and significantly lower ScvO 2 and ScvO 2 /lactate ratio values than those not experiencing MAE at all time frames after surgery, except for lactate between six and 12 hours . For ScvO 2 , the AUC in all time frames were good discriminants of MAE. The best AUC for ScvO 2 was obtained with the lowest value between zero and 48 hours after surgery (0.763). For lactate, all AUC were good discriminants of MAE, and the best AUC was obtained with the values between 12 and 48 hours (0.806). For ScvO 2 /lactate ratio, values in all time frames were also good discriminants of MAE. The best AUC for ScvO 2 /lactate ratio was the lowest value between zero and 48 hours after surgery (0.841). ScvO 2 /lactate ratio had the higher AUC for prediction of MAE. oxygen saturation [ScvO 2 ]), but these approaches lacked sensitivity .
Arterial lactate and ScvO 2 are indirect markers of hypoxia, reflecting poor tissue perfusion, and exhibit higher sensitivity and specificity for predicting MAE than other variables do, especially after pediatric congenital heart surgery. However, when isolated, they still have low accuracy, but a combination of these two markers (a ratio between ScvO 2 and arterial lactate, named ScvO 2 /lactate ratio), when < 5, showed a positive predictive value (PPV) of 93.8% for MAE, as well as 78.9% sensitivity and 90.5% specificity . More recently, the vasoactive-inotropic score (VIS) was associated with poor outcomes after pediatric congenital heart surgery.
However, the predictive value of ScvO 2 /lactate ratio for MAE after pediatric congenital heart surgery, with a cutoff value of 5, has not been validated in other populations. Therefore, we aimed to retrospectively assess the predictive value of ScvO 2 / lactate ratio for MAE after pediatric congenital heart surgery in a Brazilian university hospital. We hypothesized that a low ScvO 2 / lactate ratio would have a PPV of 90% or more for MAE.
# Methods
This was a retrospective observational study carried out in a single center, a Brazilian tertiary care university hospital, the Hospital das Clínicas, Faculdade de Medicina de Ribeirão Preto, Universidade de São Paulo. The study followed the Brazilian regulations for research on human subjects, it was approved by the local institutional review board (Research Ethics Committee HCFMRP-USP, protocol #CAAE: 99316918.0.0000.5440), and informed consent was waived. This study followed the STROBE recommendations for observational studies (https://www.strobe-statement.org/).
All patients < 18 years old submitted to congenital heart surgery between 2015 and 2017 were eligible for the study. The inclusion criteria were age < 18 years, gestational age ≥ 37 weeks, body weight ≥ 2 kg, and being submitted to open heart surgery. The exclusion criteria were death in the operation room, lack of information on medical records, and not having ScvO 2 or lactate measured within 12 hours after surgery.
Data were retrieved from the electronical medical records, including demographic, clinical, surgical, and laboratorial data. An anonymized database was created using the Research Electronic Data Capture (or REDCap). Our primary outcome was the occurrence of a MAE within 48 hours after surgery -at least one of the following: cardiac arrest, unplanned reoperation, emergency chest reopening, low cardiac output syndrome (LCOS), or death. Unplanned reoperation was defined as the need for an additional, unanticipated surgical procedure or revision as a result of a significant postoperative residual lesion. Emergency chest reopening was defined as the need of sternum opening for exploration, bleeding control, or to alleviate pressure on the mediastinum. LCOS was clinically defined as the presence of altered mental status, mottled skin, cold extremities, prolonged capillary refill time (> 2 seconds), and weak pulses.
All ScvO 2 measurements were done in venous blood drawn from catheters placed in the superior vena cava or in the right atrium (in the absence of left-to-right shunt). Lactate was measured in blood drawn from indwelling arterial lines presents a comparison between different cutoff points for ScvO 2 /lactate ratio. All cutoff points rendered good NPV, but very low PPV and variable sensitivity and specificity. The lower cutoff (< 5) increases specificity at the cost of a low sensitivity, while the higher cutoff (< 17) increases sensitivity at the cost of low specificity. The intermediate cutoff point (< 9) yields very good NPV, and low sensitivity (61%) with fair specificity (85%). Low cardiac output syndrome 6 (3.1%)
Counts of MAE sum up to more than 31 because one patient may have had more than one MAE. Percentages refer to the whole sample of 194 patients.
The cutoff < 5 had the best accuracy (87%). According to each cutoff, the relative risks of a patient with a ScvO 2 /lactate ratio < 5, < 9, or < 17 experience a MAE were 6.33 (95% confidence interval [CI] 3.60-11.2), 5.40 (95% CI 2.85-10.2), and 3.64 (95% CI 1.77-7.47), respectively. Based on these numbers, having a ScvO 2 /lactate ratio < 5 at any time within 48 hours after surgery is associated with a six-fold increased risk of MAE, with accuracy of 87% and specificity and NPV > 90%.
# Discussion
We have shown that ScvO 2 /lactate ratio can accurately predict the occurrence of MAE within 48 hours after pediatric congenital heart surgery, with very good NPV, but poor PPV. Thus, our hypothesis was not confirmed. This means that ScvO 2 / lactate ratio > 5 is best suitable for identifying patients at low risk of MAE. We have also confirmed that ScvO 2 /lactate ratio is superior to either variable alone for this purpose. The need for predicting tools for the occurrence of MAE is not new. Many studies have focused on lactate clearance, time of elevated lactate, cardiopulmonary bypass (CPB) duration, inflammatory markers, and other markers, and emphasis have been put in predicting LCOS, but an ideal predictive tool has not been found yet, as of today.
One of the first attempts was made in a study with 63 neonates undergoing heart surgery for transposition of the great arteries, in which high postoperative levels of cardiac troponin T, interleukin (IL)-6, and IL-8 were associated with LCOS. In 46 children submitted to open heart surgery, our group showed that preoperative N-terminal pro-B-type natriuretic peptide plus postoperative IL-8 or platelet count were independent predictors of LCOS (sensitivity 93%, specificity 75%, PPV 87%, and NPV 86%), while CPB duration or postoperative IL-8 plus postoperative cardiac troponin I (cTnI) were independent predictors of death (sensitivity 100%, specificity 65%, PPV 38%, and NPV 100%). In 99 consecutive children undergoing open heart surgery for congenital heart disease, a cTnI level > 13 ng/mL four hours postoperatively predicted LCOS with sensitivity of 0.78, specificity of 0.72, and AUC of 0.75. In addition, cTnI was the only predictive variable associated with LCOS in multivariate logistic regression analysis. In a study of 312 post CPB neonates, an oxygen delivery index had an AUC of 0.79, showing good prediction of low ScvO 2 , as a surrogate for LCOS. In 117 children submitted to open heart surgery, the VIS score at two hours after CPB was an independent predictor of LCOS. A LCOS score (tachycardia, oliguria, toe temperature < 30 °C, volume administration > 30 mL/ kg/day, decreased near infrared spectroscopy measurements, hyperlactatemia, and need for vasoactive/inotropes in excess of milrinone at 0.5 μg/kg/min) showed an AUC for a composite morbidity (prolonged mechanical ventilation, new infection, cardiopulmonary arrest, neurologic event, renal dysfunction, necrotizing enterocolitis, and extracorporeal life support) of 0.83 in 55 patients undergoing pediatric open heart surgery. Later, a modified LCOS score was superior to the inadequate oxygen delivery index in predicting adverse events linked to LCOS in 536 bypass pediatric operations for congenital heart defects.
Only a few studies focused on the detection of MAE, a more comprehensive outcome than LCOS. Duke et al.studied a cohort of 90 children submitted to heart surgery and investigated the predictive value of several variables for MAE. They found an incidence of MAE of 13.3% (ours was 18.9%), and the independent predictors of MAE were lactate (measured at intensive care unit admission), CPB duration, lactate and CO 2 difference (both measured four hours after admission), and lactate and base deficit (both measured eight hours after admission). They did not study ScvO 2 /lactate ratio specifically. Kalyanaraman et al.studied 129 children submitted to heart surgery and evaluated the predictive value of repeated measurements of arterial lactate for MAE. Moreover, they looked at the amount of time that lactate was > 2 mmol/L ("lactime"). They concluded that "lactime" had 98% sensitivity, but only 60% PPV for a postoperative MAE. Later, Schumacher et al.studied 231 infants undergoing cardiac surgery, and found that a lactate increase rate of 0.6 mmol/L/h had very good discriminatory ability (AUC 0.89) with a sensitivity of 90%, specificity of 84%, PPV of 34%, and NPV of 99% for a poor outcome (death, need for extracorporeal support, and dialysis). More recently, in a cohort of 257 pediatric patients undergoing cardiac surgery, the incidence of MAE was 19% and mortality was 13%. The predictors of MAE were cyanotic congenital heart disease, longer CPB duration (> 120 minutes), high inotropes (at least two, at high doses), and increase in lactate > 0.75 mmol/L/h or more in the first 24 hours (odds ratio 37.1, 95% CI 10.1-136.3).
Before our study, only one study assessed the predictive value of the ScvO 2 /lactate ratio for MAE. Seear et al. showed, in 52 children submitted to heart surgery, that a ScvO 2 /lactate ratio < 5 predicted MAE with PPV of 93.8%, sensitivity of 78.9% and specificity of 90.5% (NPV was not reported). They also showed that the PPV of ScvO 2 /lactate ratio was higher than those of a ScvO 2 < 40% (58.3%) or a lactate > 8 mmol/L (63.6%) alone. In our study, using the same cutoff point (< 5), NPV was higher than PPV, specificity was high, but sensitivity was too low. These results suggest that calculating the ScvO 2 /lactate ratio when caring for a pediatric patient right after open heart surgery is an easy, quick, and inexpensive tool that may help physicians to identify patients at low risk of MAE, allocating more resources to those at higher risk.
# Limitations
The limitations of our study include: (a) retrospective data, implying that some MAE may have not been properly recorded and that ScvO 2 and lactate were not consistently measured; and (b) the significant heterogeneity of heart conditions and surgeries, which compromises generalizability to other centers and populations. Therefore, our results need to be taken with caution.
# Conclusion
In conclusion, ScvO 2 /lactate ratio > 5 can accurately identify patients at low risk of MAE after pediatric congenital heart surgery, with very good specificity and NPV, but poor sensitivity and PPV.
# Acknowledgments
We thank to the team of physicians, nurses, respiratory therapists, and other allied staff from the Pediatric Critical Care Unit, Hospital das Clínicas, Faculdade de Medicina de Ribeirão Preto, Universidade de São Paulo, for their job in caring for these patients; and to Mr. Carlos Alberto Siqueira Lima Jr, for technical assistance with data extraction. . Comparison of diagnostic values of different cutoff points for the lowest central venous oxygen saturation (ScvO 2 ) to arterial lactate ratio for prediction of major adverse events.
## Cutoff
Sensitivity (%) Specificity (%) PPV (%) NPV (%) Accuracy (%)
|
KIR diversity in three ethnic minority populations in China
Background: Killer cell immunoglobulin-like receptors (KIRs) show extensive variation in genetic content and allelic polymorphisms among different populations.Materials and methods:We analyzed the distribution of KIR genes in the Tibetan ethnic minority of Lhasa city, the Uyghur and Kazakh ethnic minorities of Urumqi city populations in China. Genotyping of 16 KIR genes was tested in 479 randomly selected individuals using the multiple PCR-SSP method.Results: A total of 42 KIR genotypes were detected, of which, 29 were predicted to be AB genotypes, 12 were BB genotypes and one was AA genotypes. 27 KIR genotypes were identified in Kazakhs, 30 KIR genotypes were identified in Uyghurs and 20 KIR genotypes were identified in Tibetans. The predominant genotype 1(AA genotypes) occurred most frequently in Tibetans (52.7%, 118/224), Kazakhs (43.2%, 54/125) and Uyghurs (34.9%, 45/130). Not only the four framework genes were present in all individuals, but the pseudogene 2DP1 could also be detected in all Uyghur individuals. Tibetans were different from Kazakh and Uyghur groups in KIR genetic content and KIR allelic variation. Intriguingly, Tibetans (29.5%, 66/224) had lower frequencies of 2DS4-v when compared with Uyghurs (60.8%, 79/130) and Kazakhs (59.2%, 74/125). Uyghurs (25.4%, 33/130) displayed higher frequencies of Bx genotypes with C4Tx (absence of KIR3DS1-2DL5-2DS5-2DS1) than both Kazakhs (11.2%, 14/125) and Tibetans (3.6%, 8/224).Conclusions:The study showed that profile of KIR genotypes in three ethnic minority populations in China displayed ethnic diversity. It could be valuable for enriching the ethnical information resources for KIR gene, as well as facilitating further research on KIR-related diseases.
# Background
Killer cell immunoglobulin-like receptors (KIRs) are a group of glycoproteins expressed on the surface of Natural killer (NK) cells [bib_ref] KIR gene content diversity in four Iranian populations, Ashouri [/bib_ref] [bib_ref] NK cells: new insights on physiology and clinical implication in diseases, Schleinitz [/bib_ref]. NK cells are important components of the innate immune system that take part in various immune responses to different infectious agents [bib_ref] HIV-1 adaptation to NK-cell-mediated immune pressure, Alter [/bib_ref].
KIR genes are located on human chromosome 19q13.4. Up to now, 16 KIR gene loci have been identified, including two pseudogenes (KIR 2DP1 and 3DP1).
Four framework KIR genes (2DL4, 3DL2, 3DL3 and 3DP1) exist in almost all populations [bib_ref] Plasticity in the organization and sequences of human KIR/ILT gene families, Wilson [/bib_ref]. Fourteen functional KIR genes have been identified and confirmed, of which seven are inhibitory (KIR2DL1-3, 2DL5 and 3DL1-3) and six are activating (KIR2DS1-5 and 3DS1), with only KIR2DL4 showing both inhibitory and activating potential [bib_ref] KIR gene content diversity in four Iranian populations, Ashouri [/bib_ref] [bib_ref] KIR2DL4 (CD158d), an NK cell-activating receptor with inhibitory potential, Faure [/bib_ref] [bib_ref] Characterization of killer cell immunoglobulin-like receptor (KIR) genotypes and haplotypes in Chinese..., Bao [/bib_ref]. According to the numbers and types, KIR haplotypes are broadly classified into two groups: group A haplotypes have a fixed KIR gene (2DL1, 2DL3-4, 3DL1-3, 2DS4, 2DP1 and 3DP1), in contrast, group B haplotypes have variations in genetic content which is comprised of several genes and alleles that are not found in A haplotypes. Especially, KIR2DS1-3, 2DS5, 2DL2, 2DL5 and 3DS1 are associated only with group B haplotypes, thus B haplotypes generally encode more activating KIR receptors than A haplotypes which encode a single activating receptor (KIR2DS4). Both group A and B are present in all populations, but their frequencies vary considerably [bib_ref] KIR gene content diversity in four Iranian populations, Ashouri [/bib_ref] [bib_ref] MHC class Imolecules and KIRs in human history, health and survival, Parham [/bib_ref]. Many individuals carry two copies of A haplotypes (AA genotypes). The remainders carry Bx genotypes, which are consisted of either one copy of A haplotype and one copy of B haplotype (AB genotypes) or two copies of B haplotypes (BB genotypes). Based on the presence and absence of two distinct gene clusters (C4, KIR2DS2-2DL2-2DS3-2DL5; T4, KIR3DS1-2DL5-2DS5-2DS1), Bx genotype carries are divided into four subsets: C4Tx (presence of C4 and absence of T4), CxT4 (absence of C4 and presence of T4), C4T4 (presence of both C4 and T4), and CxTx (absence of both C4 and T4) [bib_ref] KIR gene content diversity in four Iranian populations, Ashouri [/bib_ref] [bib_ref] Global diversity and evidence for coevolution of KIR and HLA, Single [/bib_ref] [bib_ref] Donor-recipient combinations of group A and B KIR haplotypes and HLA class..., Mcqueen [/bib_ref].
Usually, Group B haplotypes encode more activating KIR genes. In individuals with AA genotype, the alleles of KIR2DS4 can encode either functional (KIR2DS4-f) or non-functional (KIR2DS4-v) variants [bib_ref] KIR2DS4 allelic variants: differential effects on in utero and intrapartum HIV-1 mother-to-child..., Heather [/bib_ref] [bib_ref] Studies on the expression of the deleted KIR2DS4*003 gene product and distribution..., Middleton [/bib_ref]. KIR2DS4v alleles differ from the KIR2DS4-f alleles in that the former have a 22-bp deletion in exon 5, which leads to a frame shift mutation and produces a premature stop codon, preventing the formation of a functional membrane-bound receptor domain [bib_ref] Studies on the expression of the deleted KIR2DS4*003 gene product and distribution..., Middleton [/bib_ref]. Therefore, individuals with AA genotypes which harbor KIR2DS4-v alleles will not have a membrane-bound KIR2DS4 protein, but rather a soluble form of KIR2DS4. KIR2DS4-f and 2DS4-v frequencies vary amongst different populations. Increasing attention has been paid to the role of KIR genetic content and allelic variation on infectious diseases such as Hepatitis C and HIV, with some studies dedicating a specific focus on KIR 2DS4 [bib_ref] KIR2DS4 allelic variants: differential effects on in utero and intrapartum HIV-1 mother-to-child..., Heather [/bib_ref] [bib_ref] Studies on the expression of the deleted KIR2DS4*003 gene product and distribution..., Middleton [/bib_ref] [bib_ref] Investigation of killer cell immunoglobulin-like receptor gene diversity: II. KIR2DS4, Maxwell [/bib_ref] [bib_ref] An allelic typing method for 2DS4 variant used in study of haplotypes..., Bao [/bib_ref] [bib_ref] Expression of activating KIR2DS2 and KIR2DS4 genes after hematopoietic cell transplantation: relevance..., Gllez-Hawkins [/bib_ref] [bib_ref] Impact of a functional KIR2DS4 allele on heterosexual HIV-1 transmission among discordant..., Merino [/bib_ref]. KIR gene diversities have been studied in many different geographical populations as previously reported [bib_ref] KIR gene content diversity in four Iranian populations, Ashouri [/bib_ref] [bib_ref] KIR2DS4 allelic variants: differential effects on in utero and intrapartum HIV-1 mother-to-child..., Heather [/bib_ref] [bib_ref] Comparison of the rapidly evolving KIR locus in Parsis and natives of..., Badwe [/bib_ref] [bib_ref] Genetic polymorphism analysis of killer cell immunoglobulin-like receptor genes in the Chinese..., Wang [/bib_ref] [bib_ref] Relationship between KIR and HLA ligands in Han population of Sichuan Marrow..., Zhan [/bib_ref] [bib_ref] Zhejiang Han population of killer cell immunoglobulin-like receptor gene polymorphism, Zhu [/bib_ref] [bib_ref] Polymorph ism of killer cell immunoglobulin-like receptors gene in Yunnan Han population, Su [/bib_ref] [bib_ref] Characterization of killer cell immunoglobulin-like receptor (KIR) genotypes and haplotypes in Chinese..., Bao [/bib_ref] [bib_ref] ParkMH: haplotype analysis of killer cell immunoglobulin-like receptor genes in 77 Korean..., Whang [/bib_ref] [bib_ref] Predominance of group A KIR haplotypes in Japanese associated with diverse NK..., Yawata [/bib_ref]. There are 56 ethnic groups in China . Most of the studies on KIR gene diversity have been reported on Han populations in China [bib_ref] Relationship between KIR and HLA ligands in Han population of Sichuan Marrow..., Zhan [/bib_ref] [bib_ref] Zhejiang Han population of killer cell immunoglobulin-like receptor gene polymorphism, Zhu [/bib_ref] [bib_ref] Polymorph ism of killer cell immunoglobulin-like receptors gene in Yunnan Han population, Su [/bib_ref] [bib_ref] Characterization of killer cell immunoglobulin-like receptor (KIR) genotypes and haplotypes in Chinese..., Bao [/bib_ref]. A large-scale survey on transfusion-transmitted HIV-1/2 infection among Chinese blood donors conducted by our institute showed that the positive rate for HIV infection was higher in some ethnic minorities (including data accumulated at Urumqi Blood Center, Xinjiang) than Han majority donors [bib_ref] Prevalence, incidence, and residual risks for transfusion transmitted HIV-1/2 infection among Chinese..., Wang [/bib_ref]. Expanding our understanding of the ethnic intermarriage and possibly random demographic factors could help us to determine the variation of KIR gene frequencies, which might be useful for future research on ethnicity-based diseases. Hence, in this study we chose the main minorities in Urumqi city of Northwest China, the Uyghur (comprised 10,069,346 persons, presented a typical mixture of Eastern and Western anthropometric traits [bib_ref] Analysis of genomic admixture in Uyghur and its implication in mapping strategy, Xu [/bib_ref] and Kazakh (comprised 1,462,588 persons, and the Tibetan (the major ethnic minority in Tibet, comprised 6,282,000 persons, mostly living in Lhasa city of Southwest China, as the research objects. (Additional file 1: [fig_ref] Figure 1: KIR gene content diversity of three Chinese populations, within 479 unrelated individuals,... [/fig_ref].
# Methods
## Samples and dna isolation
## Kir pcr-ssp genotyping
KIR genes were typed for the presence or absence of the 14 KIR genes, including KIR2DL1-5, 2DS1-5, 3DL1-3 and the 2 pseudogenes (3DP1 and 2DP1), using multiplex PCR-SSP. In order to ensure accuracy, two different sets of primers for amplification (except 3DP1 and 2DS1) were used while HLA-DRB1 intron amplification primers were added to the third well as a positive control. Primer designs were adopted from those previously reported by Campbell et al.. The primer arrangement was referred in [fig_ref] Table 1: Multiple PCR primers combination pattern a The length of 3DP1 is 399... [/fig_ref] PCR master-mix for the multiplex PCR-SSP which contained 3 μL of primer (10 μM) pairs, 1.5 μL of ddH 2 O, 0.5 μL of DNA, 5 μL of GoTaq green master mix (Promega, Cat: #M7123), for a total of 10 μL. The conditions for PCR cycles were as followed: five cycles at 94°C for 3 min, 94°C for 15 s, 65°C for 15 s, 72°C for 30 s; twenty-one cycles at 94°C for 15 s, 60°C for 15 s, 72°C for 30 s; four cycles at 94°C for 15 s, 55°C for 1 min,
# Statistical analysis
Genotypic frequency = n/N (n represented numbers of positive genotypes and N was the total number of individuals tested). The carrier frequencies (CF) of the KIR genes were determined by dividing the number of positive typing reactions by the total number of typed individuals. The estimated gene frequencies were calculated using the formula GF = 1 − (1 − CF) 1/2 . The frequencies of A and B haplotypes were calculated using the following formula: group A = 2nAA + nAB/2N and group-B = 2nBB + nAB/2N, where nAA, nAB, and nBB were the numbers of AA, AB and BB genotypes. Differences in the carrier frequencies between the studied population and other populations previously published were assessed by the standard Chi square test (x 2 ) using statistical software PEMS3.1 Medicine, and p < 0.05 was considered to be statistically significant. Subsets of KIR genetic content profiles and their frequencies in populations were carried out using statistical software Graph Pad Prism5. Principal components analysis (PCA) of 22 populationswas performed using an R package: psych software. Genetic distances were calculated by Nei's method using phylip [bib_ref] PHYLIP-Phylogeny Inference Package (Version 3.2), Felsenstein [/bib_ref]. On the basis of the Nei's genetic distance, a dendrogram was constructed by the neighbor-joining method [bib_ref] The neighbor-joining method: a new method for reconstructing phylogenetic trees, Saitou [/bib_ref] and visualized with MEGA software in order to compare the frequencies of KIR genes in 22 populations, using bootstrap values are calculated from 100 trees.
# Results
## Kir genotypes and haplotypes
A total of 479 individuals, 42 genotypes were found to carry a different number and combination of 16 KIR genes [fig_ref] Figure 1: KIR gene content diversity of three Chinese populations, within 479 unrelated individuals,... [/fig_ref] , of which, 29 were predicted to be AB genotypes, 12 were predicted to be BB genotypes and one was AA genotype. [bib_ref] Analysis of genomic admixture in Uyghur and its implication in mapping strategy, Xu [/bib_ref].
## Genetic relationships between populations
The PCA was generated depending on the basis of KIR gene frequencies of the three groups with other previously studied populations (as shown in Additional file 2: [fig_ref] Table 1: Multiple PCR primers combination pattern a The length of 3DP1 is 399... [/fig_ref] , PC1 and PC2 of 22 populations were given in Additional file 3: . Different geographic clusters of Asian and European were noted on the PCA plot. The most divergent were the clusters of African, Senegalese and Indian, as they did not cluster in a single group . Tibetans was plotted within the Asian group, the Uyghur was plotted within the European group, while the Kazakh was mapped between Asian and European groups. The neighbor-joining dendrogram was constructed, using the allelic frequency data of 12 KIR loci (KIR2DL1-3, 2DL5, 3DL1, 2DS1-5, 3DS1 and 2DP1) to
## Table 2 comparison of genotypes, haplotypes and linkage groups in tibetan, kazakh and uyghur populations
The haplotype A and B were determined by using the following formula: group A = 2nAA + nAB/2N and group B = 2nBB + nAB/2N, where nAA, nAB, and nBB are the numbers of AA, AB, and BB genotypes, N = total number of population. The p values are given only for those pairwise comparisons indicating significant (P < 0.05) differences.show a relationship between the three groups and other previously reported populations(Additional file 2: [fig_ref] Table 1: Multiple PCR primers combination pattern a The length of 3DP1 is 399... [/fig_ref]. In this dendrogram, there were four main clusters [fig_ref] Figure 3: The neighbor-joining dendrogram represented the genetic distances and was created on the... [/fig_ref] : European, African, Asian and Indian cluster. The Uyghur population was located in the European cluster, the Tibetan was located in the Asian cluster, while the Kazakh was located between European and Asian clusters. Genetic distance was shown in [fig_ref] Figure 3: The neighbor-joining dendrogram represented the genetic distances and was created on the... [/fig_ref].
## Types
## Genetic relationships between three ethnic populations centromeric and telomeric clusters of kir genes (cxtx)
We found that the three groups carried all four subsets of Bx genotypes (CxT4, C4Tx, C4T4, and CxTx, [fig_ref] Figure 1: KIR gene content diversity of three Chinese populations, within 479 unrelated individuals,... [/fig_ref] [fig_ref] Figure 1: KIR gene content diversity of three Chinese populations, within 479 unrelated individuals,... [/fig_ref]. The Uyghurs (25.4%, 33/130)
## Table 3 comparison of carrier frequency of kir genes in three ethnic minority populations in china
Frequency (%CF) of carriers of each gene is expressed as a percentage and defined as the number of donors carrying the gene (n) divided by the number of donors studied (N) in the given populations. The p values are given only for those pairwise comparisons indicating significant (P < 0.05) difference. [fig_ref] Figure 4: Subsets of KIR genetic content profiles and their frequency in populations [/fig_ref].
## Kir
# Discussion
KIR genotypes show extensive variation in populations according to different geographical regions and different ethnic groups, due to the presence or absence of different KIRs in individual haplotypes and allelic polymorphism of KIR genes [bib_ref] Genetic polymorphism analysis of killer cell immunoglobulin-like receptor genes in the Chinese..., Wang [/bib_ref]. In this study, Tibetans were detected to display a lower level of genotypic diversity, with the predominant genotypes ID1 (AA genotypes)and ID2. The cumulative frequency of the top five genotypes 1-4, 6 was 84.8% (190/224), which showed that KIR genotype distribution in Tibetan populations was relatively concentrated. Meantime, Uyghurs and Kazakhs displayed a higher genotypic diversity, with the predominant genotypes ID1. Intriguingly, Uyghur displayed consistency with Caucasians on the level of KIR allelic frequency [bib_ref] KIR2DS4 allelic variants: differential effects on in utero and intrapartum HIV-1 mother-to-child..., Heather [/bib_ref] , which comprised high frequencies of Bx genotypes (absence of KIR3DS1-2DL5-2DS5-2DS1), while Tibetans shared a lower frequencies consistent with Northeast Asian (Han Chinese [bib_ref] Characterization of killer cell immunoglobulin-like receptor (KIR) genotypes and haplotypes in Chinese..., Bao [/bib_ref] , Korean [bib_ref] ParkMH: haplotype analysis of killer cell immunoglobulin-like receptor genes in 77 Korean..., Whang [/bib_ref] and Japanese [bib_ref] Predominance of group A KIR haplotypes in Japanese associated with diverse NK..., Yawata [/bib_ref]. Our data on Kazakhs showed that the KIR allelic frequency is between that of Caucasians and Northeast Asians. Many studies on KIR gene diversity have been [fig_ref] Table 1: Multiple PCR primers combination pattern a The length of 3DP1 is 399... [/fig_ref].
reported on Han populations in China [bib_ref] Relationship between KIR and HLA ligands in Han population of Sichuan Marrow..., Zhan [/bib_ref] [bib_ref] Zhejiang Han population of killer cell immunoglobulin-like receptor gene polymorphism, Zhu [/bib_ref] [bib_ref] Polymorph ism of killer cell immunoglobulin-like receptors gene in Yunnan Han population, Su [/bib_ref] [bib_ref] Characterization of killer cell immunoglobulin-like receptor (KIR) genotypes and haplotypes in Chinese..., Bao [/bib_ref]. Expanding our understanding of ethnic intermarriage and possibly random demographic factors for ethnic minorities could help us to determine the variation of KIR gene frequencies, which might be useful for future research on ethnicity-based diseases. Activating KIR genes were found to show greater variability in frequencies than inhibitory KIR genes, which was similar to those previously reported in other populations [bib_ref] KIR gene content diversity in four Iranian populations, Ashouri [/bib_ref]. KIR2DS4 is the only activating gene in AA genotypes, with the other 5 activating genes (KIR2DS1-3, 2DS5 and 3DS1) being absent [bib_ref] KIR2DL4 (CD158d), an NK cell-activating receptor with inhibitory potential, Faure [/bib_ref] [bib_ref] Characterization of killer cell immunoglobulin-like receptor (KIR) genotypes and haplotypes in Chinese..., Bao [/bib_ref]. In this study, KIR2DS4-v was found to have a higher frequency in Uyghurs and Kazakhs, which was consistent with the frequency in Caucasians [bib_ref] KIR2DS4 allelic variants: differential effects on in utero and intrapartum HIV-1 mother-to-child..., Heather [/bib_ref]. Tibetans had a lower KIR2DS4v frequency when compared with Kazakhs and Uyghurs, more similar to Asian populations (Han Chinese, Japanese and Koreans) [bib_ref] Characterization of killer cell immunoglobulin-like receptor (KIR) genotypes and haplotypes in Chinese..., Bao [/bib_ref] [bib_ref] ParkMH: haplotype analysis of killer cell immunoglobulin-like receptor genes in 77 Korean..., Whang [/bib_ref] [bib_ref] Predominance of group A KIR haplotypes in Japanese associated with diverse NK..., Yawata [/bib_ref].
The PCA and the generated dendrogram aligned with the origins of ancestry distributing across vast geographic regions. In our data, the Uyghur was located in the European cluster, showing a similarity to the origins of European Caucasian ancestry. This was consistent with previously report that Uyghur population, settled in Xinjiang, was a population presenting a typical admixture of Eastern and Western anthropometric traits [bib_ref] Prevalence, incidence, and residual risks for transfusion transmitted HIV-1/2 infection among Chinese..., Wang [/bib_ref]. The Kazakh was between Eastern Asian and European Caucasian groups, demonstrating an array of mixed anthropological features of Europeans and Asians. This could help to further understand the origins of ancestry, the pattern of human migration, the intermarriage between ethnicities and the genetic distances between different populations or ethnic groups.
# Conclusions
The KIR gene frequencies of the Tibetan and Uyghur populations were consistent with the geographic and ethnic distribution, while the Kazakhs showed unique genetic content and allelic polymorphisms. Better understanding the ethnic intermarriage and possibly random demographic factors could help determine the factors in the variation of KIR gene frequencies, which could be useful for future research on ethnicity-based diseases.
## Additional files
Additional file 1: [fig_ref] Figure 1: KIR gene content diversity of three Chinese populations, within 479 unrelated individuals,... [/fig_ref].
[fig] Figure 1: KIR gene content diversity of three Chinese populations, within 479 unrelated individuals, 42 genotypes that differed by the presence (shaded box) and absence (white box) of 16 KIR genes were detected. The frequency of each genotype is presented in percentage frequency (%F) and defined as the number of individuals carrying the genotype (n) divided by the number of individuals studied (N) in the provided population. [/fig]
[fig] Figure 3: The neighbor-joining dendrogram represented the genetic distances and was created on the basis of the frequencies of 12 KIR gene (KIR2DL1-3, 2DL5, 3DL1, 2DS1-5, 3DS1 and 2DP1) in order to show the genetic relationship between the three groups and other previously reported populations. The KIR gene frequency of populations was given in Additional file 2: [/fig]
[fig] Figure 4: Subsets of KIR genetic content profiles and their frequency in populations. The frequencies of unique KIR genotype subsets in Tibetan, Kazakh and Uyghur populations were elucidated. [/fig]
[table] Table 1: Multiple PCR primers combination pattern a The length of 3DP1 is 399 bp and the variant 3DP1v is 280 bp. b The length of 2DS4-f is 219 bp and the variant 2DS4-v is 197 bp. [/table]
|
From Spheroids to Organoids: The Next Generation of Model Systems of Human Cardiac Regeneration in a Dish
Citation: Scalise, M.; Marino, F.; Salerno, L.; Cianflone, E.; Molinaro, C.; Salerno, N.; De Angelis, A.; Viglietto, G.; Urbanek, K.; Torella, D. From Spheroids to Organoids: The Next Generation of Model Systems of Human
## Defining an organoid
The term "organoid" refers to mini clusters of growing cells able to self-organize in vitro and differentiate into functional cell types, resembling an organ 3D structure and function. The word "organoid" is mainly used to describe such structure derived from stem cells. In multicellular organisms, stem cells are undifferentiated or partially committed cells that can differentiate into various types of cells and proliferate indefinitely to produce more of the same stem cell (the so-called self-renewal). Stem cells are present in both embryonic and adult organisms, but they have slightly different properties in each. In general, they can generate all tissues of the developing embryo and maintain tissue homeostasis in adults. In the past decade, these prototypical stem cell features have been exploited to develop the organoids in vitro. Organoids are therefore stem cell-derived and self-organizing 3D cultures that phenocopy cell-type composition, architecture, and, to a Int. J. Mol. Sci. 2021, [bib_ref] Generation of Human PSC-Derived Kidney Organoids with Patterned Nephron Segments and a..., Low [/bib_ref] , 13180 2 of 25 certain extent, functionality of different tissues [bib_ref] Use and application of 3D-organoid technology, Artegiani [/bib_ref]. The developmental potential of the initiating stem cells influences how complex the organoid can be. The organoids resemble specific features of organs in vivo: an organoid must indeed contain more than one cell type of the organ it models; it should exhibit some function related to that organ; and the cells should be organized similarly to the tissue of the organ. Additionally, organoids' formation recapitulates characteristic processes of self-organization during development [bib_ref] Human organoids: Model systems for human biology and medicine, Kim [/bib_ref]. However, although most organoid cultures develop functional tissue units, they lack elements such as mesenchymal, stromal, immune, and neural cells that populate tissue in vivo. Failing to recapitulate the complexity of native organs, the (partial) absence of a mesenchymal compartment, vascularization, and microbiome represents therefore a limit of organoid technology. Yet, recent studies are dealing with trying to overcome this limitation through obtaining a tridimensional structure that more closely reproduces the whole cellular diversity of the tissue microenvironment [bib_ref] Organoids as an in vitro model of human development and disease, Fatehullah [/bib_ref].
Although still imperfect, organoids represent an attractive model for studying human biology and disease, carrying the potential to answer several unresolved questions. Organoids rely on artificial extracellular matrices (ECM) to facilitate their self-organization into structures that resemble native tissue. ECM components such as laminin, fibronectin and collagen engage the integrin receptors and support to maintain cell identity and function. Organoids are similar to primary tissue in their composition and architecture, harboring small populations of genomically stable, self-renewing stem cells that give rise to fully differentiated progeny comprising all major cell lineages. Among the advantages of their use, organoids can be expanded indefinitely, cryopreserved as biobanks, and easily manipulated using techniques such as those established for traditional 2D monolayer culture. The study of organoid formation can provide valuable information about the mechanisms underlying development and organ regeneration, underscoring their value for basic biological research in addition to their potential application in drug testing and molecular medicine. The potential of organoids to complement existing model systems and extend basic research and drug discovery is becoming more widely appreciated [bib_ref] Stem Cells: In Sickness and in Health, Bahmad [/bib_ref] [bib_ref] A Three-Dimensional Organoid Culture System Derived from Human Glioblastomas Recapitulates the Hypoxic..., Hubert [/bib_ref].
In this review, we discuss the evolution of biological model systems tracing the main methodologies that led us from 2D-to the 3D-cell cultures and organoids' generation. In particular, we thoroughly assess the current methodologies used to generate cardiac spheroids and 3D-cell structures derived from stem cells, and highlight the potential of cardiac organoids ("cardioids") as a model system for the understanding of heart development and for the study of human cardiac regeneration in a dish. Furthermore, we argue that cardioids are ideal human preclinical models, useful to simulate pathological processes as well as to test drug toxicity, highlighting their current limitations that remain to be addressed.
## A brief historical perspective of organoid development
The first effort in describing in vitro self-organization and differentiation goes back in time to the beginning of the 20th century, when Wilson showed that dissociated sponge cells, when kept under appropriate conditions, degenerate, resulting in small masses of undifferentiated tissue able to grow and differentiate into complete sponges [bib_ref] A New Method by Which Sponges May Be Artificially Reared, Wilson [/bib_ref]. Some decades later, Holtfreter performed dissociation-reaggregation experiments with dissociated amphibian pro-nephrons [7] . In 1960, Weiss and Taylor conducted the same experiments with different organs from chicken embryos [bib_ref] Reconstitution of Complete Organs from Single-Cell Suspensions of Chick Embryos in Advanced..., Weiss [/bib_ref] , and shortly later, Pierce and Verney described the differentiation of embryoid bodies (EBs) in vitro [bib_ref] An in vitro and in vivo study of differentiation in teratocarcinomas, Pierce [/bib_ref]. At the same time, Steinberg proposed a differential adhesion hypothesis according to which thermodynamic effects regulate the cell sorting and rearrangement in surface adhesion [bib_ref] On the Mechanism of Tissue Reconstruction by Dissociated Cells, Iii. Free Energy..., Steinberg [/bib_ref]. Stem cell research flourished when different groups isolated pluripotent stem cells from mouse embryos [bib_ref] Establishment in culture of pluripotential cells from mouse embryos, Evans [/bib_ref] [bib_ref] Isolation of a pluripotent cell line from early mouse embryos cultured in..., Martin [/bib_ref]. Subsequently, other groups focused on improving cell culture protocols to mimic the in vivo environment conditions. Li et al. showed that breast epithelium organizes into 3D ducts and increases milk protein secretion when grown in a tumor extracellular matrix [bib_ref] Influence of a reconstituted basement membrane and its components on casein gene..., Li [/bib_ref]. Jennings et al. demonstrated similar structures in alveolar type II epithelial cells [bib_ref] Functional differentiation of alveolar type II epithelial cells in vitro: Effects of..., Shannon [/bib_ref]. An important event in this field occurred in 1998 when Thompson et al. isolated and cultured the first embryonic stem cell line from human blastocysts [bib_ref] Embryonic stem cell lines derived from human blastocysts, Thomson [/bib_ref]. The watershed from 2D to spheroid and organoid culture (3D) occurred with the generation of 3D cerebral cortex tissue from pluripotent stem cells by Eiraku et al. [bib_ref] Self-organized formation of polarized cortical tissues from ESCs and its active manipulation..., Eiraku [/bib_ref]. Successively, intestinal 3D structures were generated from adult intestinal stem cells in Matrigel [bib_ref] Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal..., Sato [/bib_ref] , kicking off various works in other systems, including stomach, liver, pancreas, lungs, kidney, brain, and retina [bib_ref] Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro, Spence [/bib_ref] [bib_ref] Mouse embryonic stem cell culture for generation of three-dimensional retinal and cortical..., Eiraku [/bib_ref] [bib_ref] Cerebral organoids model human brain development and microcephaly, Lancaster [/bib_ref] [bib_ref] Kidney organoids from human iPS cells contain multiple lineages and model human..., Takasato [/bib_ref] [bib_ref] Generation of Human PSC-Derived Kidney Organoids with Patterned Nephron Segments and a..., Low [/bib_ref] [bib_ref] Liver organoids: From basic research to therapeutic applications, Prior [/bib_ref] [bib_ref] Organoids from the Human Fetal and Adult Pancreas, Balak [/bib_ref]. Moreover, in the recent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, to elucidate how the virus can damage lungs, liver, and kidney tissues, causing some of the severe complications seen in patients with coronavirus disease-2019 (COVID-19), several researchers have focused on organoids that could also facilitate screening for potential new drugs. Bronchial organoids with four distinct cell types, made from frozen cells from the bronchi's epithelium-the outer cell layer-have been developed [bib_ref] In Vitro and Animal Models for SARS-CoV-2 research, Takayama [/bib_ref]. By infecting these organoids with SARS-CoV-2, it was established that the virus primarily targets stem cells, or basal cells, that supply the epithelium. At the same time, it does not easily penetrate the protective secretory cells, the Clara cells [bib_ref] An organoid-derived bronchioalveolar model for SARS-CoV-2 infection of human alveolar type II-like..., Lamers [/bib_ref]. Human blood vessel organoids have been recently derived from human pluripotent stem cells (hPSCs) for modeling and identifying the regulators of diabetic vasculopathy [bib_ref] Human blood vessel organoids as a model of diabetic vasculopathy, Wimmer [/bib_ref]. Finally, very recently, a small three-dimensional model of the heart was obtained from stem cells in vitro. Heart organoids can reproduce specific functions of a heart chamber and of pathological conditions such as congenital heart defects and tissue damage after a heart attack [bib_ref] Cardioids reveal self-organizing principles of human cardiogenesis, Hofbauer [/bib_ref] [fig_ref] Table 1: Main methodologies of cardiosphere derivation with their characteristics [/fig_ref] Heart organoids are established from human pluripotent stem cells Bronchial organoids are developed for evaluation of SARS-CoV-2 infection of human alveolar type II-like cells 3D culture system are used to expanse adult human pancreatic tissue Human blood vessel organoids are used as model of diabetic vasculopathy Lung organoids are developed by manipulating signaling pathways of pluripotent stem cells generating ventral-anterior foregut spheroids Development of kidney organoids containing nephrons associated with a collecting duct network surrounded by renal interstitium and endothelial cells Cerebral organoids derived from human pluripotent stem cells grown in Matrigel and with agitation 3D retinal and cortical tissues are generated from mouse embryonic stem cell culture Single sorted Lgr5+ stem cells are used to derive intestinal crypt-villus organoids Self-organized formation of apico-basally polarized cortical tissues from embryonic stem cells are generated using 3D aggregation culture The first human embryonic stem cell line is isolated and cultured from human blastocysts Breast epithelia show self-organization into 3D ducts and ductules when grown on tumor ECM extract Alveolar type II epithelial cells are functionally differentiated in vitro on ECM extract Pluripotent stem cells are isolated from mouse embryos and termed "embryonic stem cells" The differential adhesion hypothesis (DAH) of cell sorting and rearrangement in surface ahesion is introduced by Steinberg Differentiation of Embyoid Bodies in vitro Dissociation-reaggregation experiments are conducted with multiple organs from embryonic chick Dissociation-reaggregation experiments are performed with dissociated amphibian pronephros First description of in vitro self-organization and differentiation: sponge cells self-organize to regenerate a whole organism Liver organoids are derived from in vitro expansion of single Lgr5+ liver stem cells induced by Wnt-driven regeneration . Milestones timeline of the methodologies for 3D cell cultures and organoids' generation. Since the beginning of the 20th century, several different methodological breakthroughs have been established in biology and clinical translation leading to the current achievements, challenges, and potential applications of organoids.
## The basis of 3d cellular structure formation
Understanding the principles that lead to organization during tissue development is fundamental to developmental biology. Cell sorting is the process by which cohering disorganized aggregates of cells establish structured tissues. These cellular aggregates from disorganized structures become homogeneous tissue domains [bib_ref] Cell sorting out: The self-assembly of tissues in vitro, Armstrong [/bib_ref]. The ability of disordered cell aggregates to restore normal tissue architecture suggests that understanding the mechanisms underlying cell sorting in vivo should prove instructive in understanding the processes that govern and stabilize the definitive relationships of the tissues associated with each other in the various organs. These cells can sometimes further divide to give rise to more differentiated progeny, which is further displaced. The basis of this organ self-assembly seems to arise from the segregation of cells with similar adhesive properties into domains that achieve the most thermodynamically stable pattern, known as Steinberg's differential adhesion hypothesis [bib_ref] Phases in Cell Aggregation and Tissue Reconstruction an Approach to the Kinetics..., Steinberg [/bib_ref]. Differential adhesion is mediated by cell surface adhesion proteins, for example, in separating vertebrate neural and epidermal ectoderm [bib_ref] Ectopic expression of N-cadherin perturbs histogenesis in Xenopus embryos, Fujimori [/bib_ref] [bib_ref] The effects of N-cadherin misexpression on morphogenesis in Xenopus embryos, Detrick [/bib_ref] , where differential epithelial and neural and cadherin expression mediates cell sorting out. The events of cell sorting do not mimic the pathways of the morphogenic movements, whereas during normal morphogenesis, the tissues of single organs do not sort out into their final form from random mixtures of the constituent cells. Despite the latter, an understanding of the mechanisms by which cultured heterotypic cell aggregates generate patterned arrays of tissues shows promise for analyzing the processes that the embryo employs to produce the definitive organization of tissues brought into association by prior morphogenetic cell movements.
A second mechanism that can influence tissue morphogenesis is the correct and spatially restricted progenitor cell fate decisions. Progenitor cells give rise to more differentiated progeny, which, because of spatial constraints of the tissue and division orientation, are forced into a more superficial position that promotes their differentiation. This stratification depends upon proper stem cell division orientation, the interplay of symmetric and asymmetric divisions, and the migration of differentiated daughter cells to defined locations within the tissue [bib_ref] From progenitors to differentiated cells in the vertebrate retina, Agathocleous [/bib_ref] [bib_ref] Mammalian inscuteable regulates spindle orientation and cell fate in the developing retina, Zigman [/bib_ref]. A common technique used in studying tissue organization is re-aggregation, which breaks the tissue down into its simplest components, the cells, and allows them to re-aggregate in a simplified environment, devoid of surrounding tissues. It enables the investigator to observe how these cells interact to form the tissue and manipulate the components to determine which are essential in this process. Re-aggregation studies have been used since the early 20th century, first using simple organisms such as sea urchins and sponges and later using more complex, multi-layered tissues such as limb buds and retina from chicks [bib_ref] Rotation-mediated histogenetic aggregation of dissociated cells. A quantifiable approach to cell interactions..., Moscona [/bib_ref] [bib_ref] The dissociation and aggregation of cells from organ rudiments of the early..., Moscona [/bib_ref]. These studies have revealed the innate ability of these multi-layered tissues to self-organize in vitro in the absence of many extrinsic cues and scaffolds. Stem cells (SCs), cultured in a dish, proliferate as a monolayer in two-dimensions (2D) and frequently require indeterminate or xenogeneic materials, including attachment substrates, cytokines, growth factors, as well as serum, to be effectively maintained and expanded in vitro. Xenogeneic contaminants from any non-human feeder cells or foreign components of the culture system hinder clinical application [bib_ref] Toward xeno-free culture of human embryonic stem cells, Mallon [/bib_ref]. Monolayer culture requires interaction to maintain self-renewal and potency of cells, cell differentiation, vitality, expression of genes and proteins, responsiveness to stimuli, drug metabolism, and other cellular functions, which are highly inefficient for large-scale expansion of cells [bib_ref] The third dimension bridges the gap between cell culture and live tissue, Pampaloni [/bib_ref] [bib_ref] Three-dimensional cell culture: The missing link in drug discovery, Breslin [/bib_ref]. In particular, 2D attachment influences cell shape and structure [bib_ref] Control of stem cell fate by physical interactions with the extracellular matrix, Guilak [/bib_ref] , leading to cell flattening and changes in the internal cytoskeleton and nuclear shape [bib_ref] Advances in 3D cell culture technologies enabling tissue-like structures to be created..., Knight [/bib_ref] , which in turn induces gene and protein expression changes [bib_ref] Engineering gene expression and protein synthesis by modulation of nuclear shape, Thomas [/bib_ref] [bib_ref] Adult cardiac stem cells are multipotent and robustly myogenic: C-kit expression is..., Vicinanza [/bib_ref]. After isolation from the tissue and transfer to the 2D conditions, cells start to lose their morphology, affecting their function [bib_ref] Geometric cues for directing the differentiation of mesenchymal stem cells, Kilian [/bib_ref] , the organization of the structures inside the cell, secretion, and cell signaling [bib_ref] Of extracellular matrix, scaffolds, and signaling: Tissue architecture regulates development, homeostasis, and..., Nelson [/bib_ref]. Further studies have shown that the composition and organization of the ECM can also send biochemical and mechanical signals for cell differentiation [bib_ref] Extracellular matrix: A dynamic microenvironment for stem cell niche, Gattazzo [/bib_ref]. Two-dimensional culture techniques and applications have been practiced for most primary and established cell lines and standardized for analytical assays ranging from microscopy and counting cells to the study of disease processes and drug testing [bib_ref] Organoid-on-a-chip and body-on-a-chip systems for drug screening and disease modeling, Skardal [/bib_ref]. When in 2D, cells have more surface area in contact with the plastic and culture media than with other cells [bib_ref] Deconstructing the third dimension: How 3D culture microenvironments alter cellular cues, Baker [/bib_ref] , forcing them into a polarization that does not reflect physiological conditions. Two-dimensional culture has been used to differentiate SCs into many specialized cells, including chondrocytes, osteocytes, adipocytes, cardiomyocytes, smooth muscle cells, and hepatocytes [bib_ref] Differentiation of embryonic stem cells into hepatocytes: Biological functions and therapeutic application, Yamamoto [/bib_ref] [bib_ref] Effects of extracellular matrixes and growth factors on the hepatic differentiation of..., Ishii [/bib_ref]. One of the advantages of monolayer culture is that it allows for uniform treatment for the differentiation of cells [bib_ref] Effects of extracellular matrixes and growth factors on the hepatic differentiation of..., Ishii [/bib_ref]. In some cases, 2D cultures, however, result in a lack of resulting functional derivatives [bib_ref] Controlled differentiation of stem cells, Hwang [/bib_ref]. Further studies have shown that 2D monolayer culture fails to reproduce animal physiology [bib_ref] 3D cell culture: A review of current approaches and techniques, Haycock [/bib_ref] appropriately and is insufficient to validate drug discovery [bib_ref] Engineering cellular microenvironments to improve cell-based drug testing, Bhadriraju [/bib_ref]. Cultures of pluripotent stem cells (PSCs) on dishes are coated with ECM components (such as Matrigel, laminin, collagen, or gelatin) and mouse embryonic fibroblasts (MEFs) feeder layer to support attachment [bib_ref] Preparation of mouse embryonic fibroblast cells suitable for culturing human embryonic and..., Jozefczuk [/bib_ref]. Two-dimensional expansion of embryonic stem cells (ESCs) has been enhanced using completely defined xeno-free culture media and attachment substrates such as albumin-free E8 media and humanized recombinant protein vitronectin, respectively [bib_ref] Chemically defined conditions for human iPSC derivation and culture, Chen [/bib_ref] [bib_ref] Concise review: The evolution of human pluripotent stem cell culture: From feeder..., Villa-Diaz [/bib_ref]. However, homogeneous expansion of PSCs is still challenging to stabilize as 2D culture methods for propagation of PSCs are laborious, expensive, and require a high level of expertise. In general, 2D culture conditions favor the non-specific differentiation of PSCs.
The use of single-cell and multi-cell spheroids has proven to be an efficient system to optimize and overcome limitations associated with in vitro conventional systems. SCs are generally cultured in vitro under non-adherent conditions as spheres or adherent conditions in two-dimensional cultures or three-dimensional matrices. This method represents one of the simplest 3D culture techniques to achieve by forming multicellular aggregates, or spheroids, which allow 3D interactions with cells and the ECM in the absence of additional substrates [bib_ref] Scaffold-free culture of mesenchymal stem cell spheroids in suspension preserves multilineage potential, Baraniak [/bib_ref]. A spheroid culture system provides a similar physicochemical structure that closely mimics the in vivo tissue counterparts by facilitating cell-cell and cell-matrix interaction, which play a significant role in various cellular mechanisms, subsequently maintaining the cellular properties [bib_ref] From cell-ECM interactions to tissue engineering, Rosso [/bib_ref]. These spheroids have been utilized with a wide range of adherent cell types, formed by spontaneous or forced aggregation techniques including hanging drop, rotating culture, or low-adhesion culture plates in suspension culture [bib_ref] In vitro differentiation of embryonic stem cells, Keller [/bib_ref] [bib_ref] Spheroid Formation of Hepatocarcinoma Cells in Microwells: Experiments and Monte Carlo Simulations, Wang [/bib_ref]. During spheroid formation, dispersed cells aggregate due to long-chain ECM fibers consisting of RGD (the tripeptide Arg-Gly-Asp) motifs that allow binding cell-surface integrin, and this leads to upregulated cadherin expression. Cadherin accumulates on the surface of the cell membrane, and the hemophilic cadherin-cadherin binding between neighboring cells allows for tightening connections between cells, and spheroids are formed [bib_ref] Recent advances in three-dimensional multicellular spheroid culture for biomedical research, Lin [/bib_ref] [bib_ref] Advances in multicellular spheroids formation, Cui [/bib_ref]. Spheroids comprise highly proliferative, non-proliferative, and apoptotic cells with limited diffusion of oxygen and nutrients to the center of the spheroid, leading to an increasing hypoxic environment [bib_ref] Three-dimensional cell culture systems and their applications in drug discovery and cell-based..., Edmondson [/bib_ref]. Due to their heterogeneous nature, spheroids have been more successfully employed to study complex 3D cell structures, cell differentiation, and cancer biology than homogenous cell proliferation [bib_ref] Scaffold-free culture of mesenchymal stem cell spheroids in suspension preserves multilineage potential, Baraniak [/bib_ref]. However, long-term suspension culture of spheroids often results in aggregation of cells, leading to necrotic centers due to limited diffusion of nutrients and oxygen into and waste out of the aggregate. In contrast to ESCs, short-term spheroid culture has been employed to maintain and expand mesenchymal stem cells (MSCs) [bib_ref] Scaffold-free culture of mesenchymal stem cell spheroids in suspension preserves multilineage potential, Baraniak [/bib_ref]. When subcultured back to 2D culture conditions, spheroid-grown MSCs displayed an undifferentiated morphology and enhanced differentiation potential via increased ECM deposition compared to adherentgrown MSCs [bib_ref] 3D spheroid culture system on micropatterned substrates for improved differentiation efficiency of..., Wang [/bib_ref]. MSCs cultured in 3D spheroids exhibited increased clonal growth and multipotency [bib_ref] Three-dimensional spheroid culture of human umbilical cord mesenchymal stem cells promotes cell..., Li [/bib_ref] , altered miRNA expression, and increased acetylation in the promoter regions of pluripotency genes, OCT4, SOX2, and NANOG [bib_ref] Mammalian microRNAs predominantly act to decrease target mRNA levels, Guo [/bib_ref].
Organoids represent an essential bridge between traditional 2D cultures and in vivo mouse/human models. They are more physiologically relevant than monolayer culture models and are more amenable to manipulating niche components, signaling pathways, and genome editing than in vivo models. Organoids are classified into tissue and stem cell organoids, depending on the source from which they are formed. Stem cell organoids are generated from either ESCs, induced pluripotent stem cells (iPSCs), or primary stem cells such as neonatal tissue stem cells or tissue-resident adult stem cells. To date, several in vitro organoids have been established to resemble various tissues, including functional organoids for thyroid, pancreas, liver, stomach, intestine, vascularized cardiac patch, and cerebral.
A notable difference between organoids derived from primary tissue and ESCs/iPSCs is the presence of cell types other than the intended lineage in the latter. This is because the factors used for directed differentiation of ESCs/iPSCs are not entirely efficient in driving all the cells towards the lineage of choice; thus, many ectodermal and endodermal organoids, such as those of the intestine, stomach, and kidney, have limited presence of mesenchymal cell types [bib_ref] Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro, Spence [/bib_ref] [bib_ref] Kidney organoids from human iPS cells contain multiple lineages and model human..., Takasato [/bib_ref] [bib_ref] Modelling human development and disease in pluripotent stem-cell-derived gastric organoids, Mccracken [/bib_ref]. In addition, the capability of self-renewing to grow in a near-physiological manner provides us with an excellent model system for a wide range of basic research and translational applications. A significant advantage of this system is the ability to greatly expand tissue-specific stem cells and their differentiated progeny from minimal amounts of starting material such as biopsies, facilitating detailed analyses of stem cell behavior, drug screening, disease modeling, and genetic screening. Indeed, intestinal organoids have already been used extensively to analyze stem cell behavior, identify niche components, model pathogen-epithelia interactions, gene editing, disease modeling, and orthotopic transplantation [bib_ref] Organoids as an in vitro model of human development and disease, Fatehullah [/bib_ref]. As organoids generated from ESCs, iPSCs, and fetal tissues faithfully retain the features of their original developmental stage, we can obtain detailed information of embryonic development in a dish as differentiation of the cells is systematically induced. It also delivers invaluable mechanistic insight into the development of stem cells and their niches while providing an opportunity to monitor their differentiation into mature functional lineages.
Another exciting application of the organoids is modeling host-microbe interactions. Some 3D tissue models have been applied to the study of microbial pathogenesis, such as hemolytic uremic syndrome caused by Shiga-toxin-producing Escherichia coli [bib_ref] Oncogenic transformation of diverse gastrointestinal tissues in primary organoid culture, Li [/bib_ref]. Studying bacterial and viral infection will allow a greater understanding of the pathogenic mechanisms and lead to better treatment strategies. Other studies have employed CRISPR/Cas9mediated gene editing of healthy organoids to evaluate candidate gene function in tissue physiology and carcinogenesis directly. This advanced molecular technology system allowed the manipulation of specific genes enabling disease modeling and targeted gene therapy. Heart disease-associated human induced pluripotent stem cells (hiPSCs) have been derived from patients with cardiomyopathy, cardiomyopathy-associated Duchenne Muscular Dystrophy, familial long QT syndrome, prolonged QT interval, arrhythmia, hypertrophy, and myocardial infarction. In addition, through CRISPR/Cas9 technology, the possibility of creating hiPSCs with specific gene mutations or repairing known gene mutations to re-establish physiological cell function has become concrete. Moreover, this approach introduced serial mutations into healthy human colon organoids, converting them into cancer organoids capable of driving in vivo cancer formation following orthotopic or kidney capsule transplantation [bib_ref] Oncogenic transformation of diverse gastrointestinal tissues in primary organoid culture, Li [/bib_ref] [bib_ref] Sequential cancer mutations in cultured human intestinal stem cells, Drost [/bib_ref] [bib_ref] Modeling colorectal cancer using CRISPR-Cas9-mediated engineering of human intestinal organoids, Matano [/bib_ref]. Patient-derived organoids also represent an essential resource for developing personalized treatment regimes. A wide variety of active drugs and small compounds can be screened for targeting candidate signaling pathways to design more effective drug regimens in conjunction with other relevant diagnostic and prognostic factors. Furthermore, in combination with 4D microscopy, organoids can be tracked over time to assess cancer stem cell behavior and viability in response to active drugs to predict patient outcomes.
## Organoid generation
Stem cells are primitive or "unspecialized" cells that have the potential to differentiate into many different and specialized cell types such as blood cells, muscle cells, bone cells, spleen cells, and other cells with specific functions [bib_ref] Embryonic stem cell lines from human blastocysts: Somatic differentiation in vitro, Reubinoff [/bib_ref] [bib_ref] Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures..., Takahashi [/bib_ref]. Different methods exist for generating tissues from human pluripotent stem cells such as ESCs and iPSCs, and adult Stem Cells (aSCs) by mimicking the biochemical and physical cues of tissue development and homeostasis [bib_ref] Organogenesis in a dish: Modeling development and disease using organoid technologies, Lancaster [/bib_ref]. If provided with the proper microenvironment, ESCs and iPSCs can potentially differentiate into any tissue that arises from the three germ layers, but not the embryo, because they cannot give rise to the placenta and supporting tissues. Most tissues have multipotent stem cells, capable of producing a limited range of differentiated cell lineages appropriate to their location. This capacity to spontaneous differentiation into all three germ layers mimics embryonic development and promotes heterogeneous differentiation. In endoderm organoids generated from ESCs and iPSCs, TGF-β (transforming growth factor) signaling is stimulated to perform the definitive endoderm, differentiating into the corresponding embryonic viscera segment [bib_ref] Efficient differentiation of human embryonic stem cells to definitive endoderm, D'amour [/bib_ref]. In organoids from the ectoderm, ESCs and iPSCs are induced to form "embryoid body"-like aggregates (PSC clusters) that are then driven to a neural or non-neural fate following ectodermal specification [bib_ref] Cerebral organoids model human brain development and microcephaly, Lancaster [/bib_ref] [bib_ref] An in vivo model of functional and vascularized human brain organoids, Mansour [/bib_ref]. In organoids derived from the mesoderm, kidney organoids are generated by modulation of fibroblast growth factor (FGF) and GSK-3β signaling pathways in human iPSCs through a mesodermal intermediate stage [bib_ref] Kidney organoids from human iPS cells contain multiple lineages and model human..., Takasato [/bib_ref].
aSCs are multipotent stem cells, able to differentiate in cell types closely related to the origin tissue. Creating conditions that mimic those of the natural stem niche can give rise to the tissue from which they were isolated. In addition, they play an important role in maintaining tissue homeostasis and tissue repair after injury. The first observation for tissue generation from aSCs was related to the growth of epidermal stem cells and large amounts of epithelium in vitro [bib_ref] Serial cultivation of strains of human epidermal keratinocytes: The formation of keratinizing..., Rheinwald [/bib_ref].
The advent of iPSC technology and the diversity of human aSC culture methods have made it possible, for the first time, to generate laboratory models specific to an individual. Reprogramming other cells into iPSCs has become a routine laboratory procedure, but generating disease models from those cell lines remains challenging. Culture methods of iPSCs have been developed to mimic in vivo organ development in 3D, allowing more complex tissue structures and diverse cell types to be modeled simultaneously. In this methodology, human iPSCs are sequentially exposed to a course of differentiation cues to simulate the stages of a human developmental process. During this process, differentiated iPSCs aggregate to form first an organ bud and later organoids that truly mimic the mature organ structure, including multiple cell types and their interactions [bib_ref] Human organoids: Model systems for human biology and medicine, Kim [/bib_ref].
Human aSC-derived organoids have also emerged as an alternative organoid system that consists of a simpler structure. In contrast to the complicated process of iPSC reprogramming followed by differentiation to the required organ type, these organoids can be generated from biopsies isolated directly from the organ of interest or diseased patient tissue. However, the establishment of human aSC-derived organoids is limited by accessibility to the tissue and prior knowledge of the culture conditions for that tissue. At the same time, an iPSC line, once established from a patient, can be used to generate different tissue models without any time limit repeatedly (that is, beyond the patient's lifespan) [bib_ref] Disease modelling in human organoids, Lancaster [/bib_ref].
During organoid formation, many common factors are used to control spatially and temporally the self-renewal and differentiation of stem cells or assist self-organization. Growth factors or small molecules influence multiple signaling pathways critical in cell survival, proliferation, and self-renewal, often in a tissue-specific manner. The use of ESCs and iPSCs lines to generate organoids involves exposure to factors that promote germ-layer and tissue-specific patterning, incorporation into Matrigel medium to facilitate the development of 3D architecture and treatment with differentiation factors to produce desired organs. Matrigel, which is a heterogeneous and gelatinous protein mixture secreted by Engelbreth-Holm-Swarm (EHS) mouse sarcoma cells, provides a scaffold and additional supplementation of signaling cues via basement membrane ligands to support cell attachment and survival as well as organoid formation [bib_ref] Feeder-free growth of undifferentiated human embryonic stem cells, Xu [/bib_ref]. It comprises mainly adhesive proteins such as collagen, laminin, and heparin sulfate proteoglycans, which resemble the extracellular environment to provide structural support and ECM signals to the cells. Overall, matrigel provides a complex set of ECM signaling inputs and an appropriate mechanical context to organoids in vitro [bib_ref] Matrigel: Basement membrane matrix with biological activity, Kleinman [/bib_ref] [bib_ref] Synthetic alternatives to Matrigel, Aisenbrey [/bib_ref]. However, Matrigel has no defined composition and is animal-derived, representing a limitation to translation in clinical settings. As an alternative, essential signals from native ECMs can be incorporated into synthetic polymer matrices to produce designer ECMs like hydrogel generated from naturally occurring materials, such as fibrin [bib_ref] Growth of Epithelial Organoids in a Defined Hydrogel, Broguiere [/bib_ref] , collagen [bib_ref] Type I collagen as an extracellular matrix for the in vitro growth..., Jabaji [/bib_ref] , or hyaluronic acid [bib_ref] Rapid Induction of Cerebral Organoids From Human Induced Pluripotent Stem Cells Using..., Lindborg [/bib_ref] and synthetic hydrogels [bib_ref] Designer matrices for intestinal stem cell and organoid culture, Gjorevski [/bib_ref]. Many laboratories have used biochemically inert crosslinked hydrogels such as polyethylene glycol (PEG) or alginate to encapsulate cells in 3D [bib_ref] Bio-inspired 3D microenvironments: A new dimension in tissue engineering, Magin [/bib_ref]. These biomimetic scaffolds can be designed for specific organoid applications, and they can be constructed from either synthetic polymers (such as polyacrylamine and polyethylene glycol, PEG) or natural macromolecules (i.e., agarose or collagen) that can be used to make them permissive to biological processes. Indeed, engineering ECM with a material capable of supporting the spectrum of cell behaviors, critical for stem cells surviving, dividing, differentiating, and ultimately self-organizing into organoid-like structures, has involved considerable efforts, but further work is needed to identify materials with a full spectrum of chemical and physical properties necessary to support organoid growth and differentiation from stem cells.
Engineered matrices have been necessary for isolating the effects of mechanical cues on stem cell activity, independent of biochemical signals. Stiffness is a critical parameter influencing stem cell behavior and appears to be a determinant of the differentiation of MSCs towards different lineages [bib_ref] Matrix elasticity directs stem cell lineage specification, Engler [/bib_ref]. Engineered matrices for organoid cultures might be designed with such physical considerations, including stiffness, matrix visco-elasticity, and degradability, which must be optimized for each specific organoid system. The efficiency of stem cell self-renewal and differentiation is correlated to ECM, and alterations in ECM composition are a hallmark of many diseases [bib_ref] Genetic diseases of connective tissues: Cellular and extracellular effects of ECM mutations, Bateman [/bib_ref] [bib_ref] Remodeling and homeostasis of the extracellular matrix: Implications for fibrotic diseases and..., Cox [/bib_ref].
For scaffold-free techniques, cells are cultured in droplets of a defined culture medium hanging from a plate by gravity and surface tension [bib_ref] Generation of multicellular tumor spheroids by the hanging-drop method, Timmins [/bib_ref]. Alternatively, the 3D structure of the organoids can also be established via "air-liquid interface". In this case, cells are cultured on an MEF feeder layer or Matrigel initially submerged in a medium, which gradually evaporates and exposes the upper cell layers to the air to allow polarization and differentiation [bib_ref] Isolation and characterization of mouse and human esophageal epithelial cells in 3D..., Kalabis [/bib_ref] [bib_ref] Organoids and the genetically encoded self-assembly of embryonic stem cells, Turner [/bib_ref]. The initial culture conditions determine whether a cell (or cluster) will form an organoid by self-organization [bib_ref] Engineering Stem Cell Self-organization to Build Better Organoids, Brassard [/bib_ref] [bib_ref] An Efficient Intestinal Organoid System of Direct Sorting to Evaluate Stem Cell..., Fujimichi [/bib_ref] , and the size and number of EBs have also been shown to affect differentiation efficiency [bib_ref] Control of human embryonic stem cell colony and aggregate size heterogeneity influences..., Bauwens [/bib_ref]. The size and shape of the initial cell aggregates are important starting conditions for organoids formation. Microwell structures or microfluidic devices usually achieve controlled cell aggregation; alternatively, the surface of cells, the cell membrane, can be modified to improve or initiate cell clustering [bib_ref] Cell membrane engineering with synthetic materials: Applications in cell spheroids, cellular glues..., Amaral [/bib_ref]. There are different strategies for controlling spatial cell arrangement in vitro, from binding peptides proteins, nanoparticles, polymers, or bio-orthogonal chemical species [bib_ref] Scaffold Free Bio-orthogonal Assembly of 3-Dimensional Cardiac Tissue via, Rogozhnikov [/bib_ref] , to cell surface binding of 3D DNA origami nanostructure that enables the programming of cell-cell adhesion [bib_ref] Engineering Cell Surface Function with DNA Origami, Akbari [/bib_ref]. In addition, it is possible to apply a magnetic field to cells magnetized by membrane-binding nanoparticles [bib_ref] Use of magnetic forces to promote stem cell aggregation during differentiation, and..., Fayol [/bib_ref]. Beyond modifying cellular surfaces, a wealth of genetic engineering strategies is available to control the intrinsic properties of a cell [bib_ref] Synthetic genetic circuits for programmable biological functionalities, Xia [/bib_ref] [bib_ref] Understanding Biological Regulation Through Synthetic Biology, Bashor [/bib_ref].
Cells can be engineered by targeting components of pathways that control stem cell differentiation and key niche signals or by genome editing (i.e., CRISPR-Cas9 technology). The CRISPR system is much more flexible than existing techniques that use proteins such as transcription activator-like effectors (TALEs) and zinc-finger proteins. Although effective for targeting DNA in a sequence-specific manner, these systems utilize proteins that contain a DNA-binding domain (rather than nucleic acids) for their target specificity [bib_ref] Cas-based methods for genome engineering, Gaj [/bib_ref]. Genome editing could modulate the intrinsic response of cells in an organoid to external stimuli or induce specific differentiation drivers that could be employed for the terminal differentiation of cells. In 2013, it was shown that CRISPR could be applied in mouse and human intestinal organoids either to knock out a gene or to correct a disease-causing mutation cystic fibrosis transmembrane conductor receptor (CFTR) [bib_ref] Functional repair of CFTR by CRISPR/Cas9 in intestinal stem cell organoids of..., Schwank [/bib_ref] [bib_ref] CRISPR/Cas 9 genome editing and its applications in organoids, Driehuis [/bib_ref]. Furthermore, genetic bioengineering could be deployed to knock down relevant signaling pathways in specific cells to make them unresponsive to the corresponding stimulus, which could address the lack of spatial signaling control in organoids. Retinal organoids, derived from iPSCs from patients with retinitis pigmentosa, could also be gene-corrected to rescue defects in photoreceptor morphology and function [bib_ref] Gene Correction Reverses Ciliopathy and Photoreceptor Loss in iPSC-Derived Retinal Organoids from..., Deng [/bib_ref]. In 2015, Freedman et al. showed that CRISPR could be used to model disease phenotypes in hPSC-derived organoids, creating an in vitro model for polycystic kidney disease [bib_ref] Modelling kidney disease with CRISPR-mutant kidney organoids derived from human pluripotent epiblast..., Freedman [/bib_ref]. Although the initial stages of differentiation appeared identical between PKD mutant and control organoids, PKDmutant organoids behave differently when differentiated toward more mature kidney cells. Moreover, the tubular structures that develop in wild-type hPSC-derived organoids, the maturation of PKD organoids, results in large cystic structures, a condition present in PKD patients [bib_ref] Modelling kidney disease with CRISPR-mutant kidney organoids derived from human pluripotent epiblast..., Freedman [/bib_ref]. Therefore, gene editing in organoids can be utilized to elucidate signaling pathways responsible for disease development.
Stem cell behavior in vivo is highly regulated through the extrinsic biochemical and biophysical signals from specialized microenvironments. These microenvironments consist of a complex array of signaling mechanisms from niche support cells, the ECM, and mechanical forces, as well as systemic and physiochemical conditions such as oxygen and pH levels. The capacity to self-organize leading to reproducible and tightly regulated tissue architectures in vitro is a considerably variable process. Key chemical, physical, and spatial cues that guide the progress of self-organization in vivo may be lacking after cell reaggregation in vitro [bib_ref] A high-throughput platform for stem cell niche co-cultures and downstream gene expression..., Gracz [/bib_ref]. In these cases, the tools and techniques of engineering could facilitate the more robust formation and analysis of organoids. New engineering technologies such as microwell arrays, droplet-based microfluidics, 3D bioprinting with microscale or nanoscale topography, chemically programmed tissue assembly, and chemically defined ECMs mean that it is now feasible to engineer organoids in such a way as to precisely determine their initial size, composition, and spatial organization [bib_ref] Designer matrices for intestinal stem cell and organoid culture, Gjorevski [/bib_ref] [bib_ref] Bio-inspired 3D microenvironments: A new dimension in tissue engineering, Magin [/bib_ref] [bib_ref] Stem cell niche engineering through droplet microfluidics, Allazetta [/bib_ref] [bib_ref] 3D bioprinting of tissues and organs, Murphy [/bib_ref]. To provide shape-guided morphogenesis in vitro, technologies such as micro-manufacturing, 3D printing, and laser cutting could be used. For example, a poly-dimethylsiloxane (PDMS) stamp can be used to pattern a collagen scaffold, onto which organoid-derived cells can then be seeded, or laser-shaped matrices can be applied [bib_ref] A microengineered collagen scaffold for generating a polarized crypt-villus architecture of human..., Wang [/bib_ref] [bib_ref] Homeostatic mini-intestines through scaffold-guided organoid morphogenesis, Nikolaev [/bib_ref].
Another significant challenge in constructing an artificial scaffold in vitro is precisely replicating the presentation of signals to cells supplied in a precise spatial and temporal order. In traditional 3D cultures, cells are flooded with biochemical signals without any spatio-temporal control. This limitation can be addressed by 3D culture matrices that can release or present biomolecules under spatio-temporal control [bib_ref] Photodegradable hydrogels for dynamic tuning of physical and chemical properties, Kloxin [/bib_ref] [bib_ref] In situ cell manipulation through enzymatic hydrogel photopatterning, Mosiewicz [/bib_ref] [bib_ref] Photoresponsive biomaterials for targeted drug delivery and 4D cell culture, Ruskowitz [/bib_ref]. A possible approach to overcome this limitation is the delivery of morphogen gradients by microfluidic devices that can further induce controlled symmetry breaking. In 2016, Demers and colleagues developed a versatile microfluidic device capable of reproducing the spatial and temporal chemical in vivo environment during neural tube development [bib_ref] Development-on-chip: In vitro neural tube patterning with a microfluidic device, Demers [/bib_ref]. Spatial control can be carried out by independent immobilization spatial-specific of different growth factors, promoting cell migration and differentiation within an agarose hydrogel by employing two-photon photochemistry [bib_ref] Spatially controlled simultaneous patterning of multiple growth factors in three-dimensional hydrogels, Wylie [/bib_ref]. Temporal control can be improved through photochemistry, which allows the design of matrices that locally release soluble chemical factors (or expose masked ones) in response to light stimulation [bib_ref] Morphogenesis Guided by 3D Patterning of Growth Factors in Biological Matrices, Broguiere [/bib_ref] [bib_ref] Hydrogels for Two-Photon Polymerization: A Toolbox for Mimicking the Extracellular Matrix, Torgersen [/bib_ref].
## From cardiac spheroids to cardioids
The heart is a highly specialized organ that possesses a limited capacity for self-repair and regeneration after infarction or disease. Advances in the understanding of stem cell biology as well as advances in tissue engineering have provided an unlimited source of cells, particularly cardiomyocytes (CMs), for the development of functioning cardiac muscle capable of generating force and propagating electrical signals. Cardiac muscle is composed of several cell types. The main population is represented by CMs, but the endothelial cells and fibroblasts are the most abundant cell type when referring to numbers. Cardiac muscle tissue engineered for regeneration requires a significant resource of functional CMs that could be derived either from PSCs and/or cardiac stem cells (CSCs). Most of the current knowledge of cardiovascular biology and disease arises from the use of animal models, predominantly mice, rats, and pigs. Even though animal models have always represented a key informative tool to study heart disease, many aspects, such as the species difference in functional and biological properties, limit their translation to human cardiac disease and drug testing [bib_ref] Human stem cells for modeling heart disease and for drug discovery, Matsa [/bib_ref].
In the development of cardiac organoids, the current limitation on which several research groups are focusing their effort, consist in reproducing the vascular network of cardiac tissue. Current models do not yet achieve proper vascularization and efficient angiogenesis, conditions necessary for the delivery of nutrients and oxygen. The solution produced by techniques such as microfluidics, with which it is possible to control the size and shape of new cellular constructs, has allowed significant progress; however, several bottlenecks in reproducing all the necessary conditions of the heart structure in vitro are still unresolved. The known difficulty in deriving human CMs and their propagation in culture has hindered their use and instigated the use of alternative models. Human pluripotent stem cell-derived CMs represent a valuable resource for modeling the human heart, with significant advantages with respect to the study of toxic effects and efficacy of new drug therapies, as well as in regenerative therapy to treat cardiovascular diseases. One of the main limitations in the use of hPSC-derived CMs is their immature phenotype having contractile and other biological and physiological properties that differ from those of adult human CMs. CMs derived from hPSCs are typically generated through 3D aggregates (EBs) from 2D differentiation cultures using hanging drop methods, suspension cultures, or forced-aggregation method [fig_ref] Table 1: Main methodologies of cardiosphere derivation with their characteristics [/fig_ref]. Usually, EBs are placed in matrix-covered plates for further differentiation, and the presence of contractile CMs will be assessed in the following days. Cardiac differentiation of EBs can be manipulated by the addition of cardiac growth factors, specific morphogenes, or by transgenic modifications to boost PSs cardiac commitment. CMs from cardiospheres revealed enhanced structural maturation with respect to CMs from 2D cultures [bib_ref] Microscale generation of cardiospheres promotes robust enrichment of cardiomyocytes derived from human..., Nguyen [/bib_ref]. Crucial factors that improve the efficiency of cardiac differentiation are the confluency in culture before differentiation, the size of the hPSC undifferentiated aggregates, and the size of EBs. Depending on confluency, the differentiation process is affected by the increase of cell-to-cell interactions and the associated paracrine factors. After 4-7 days of culture in suspension, EBs are plated onto gelatin-or Matrigel-coated dishes and cultured using a cardiomyogenic differentiation media. Despite that the beating foci of cardiomyocytes-derived from PSC occur within the adherent EBs layers, myogenic differentiation can also be observed in EBs cultivated the whole time in the form of floating EBs. Nevertheless, the growth conditions can considerably influence the microenvironment generated in culture and therefore also the differentiation efficiency. Specifically, myogenic differentiation of pluripotent cells can be enhanced by stage-specific application of key growth factors in defined media [bib_ref] Stage-specific optimization of activin/nodal and BMP signaling promotes cardiac differentiation of mouse..., Kattman [/bib_ref] [bib_ref] Aalto-Setälä, K. Cardiac differentiation of pluripotent stem cells, Rajala [/bib_ref] [bib_ref] Controlling expansion and cardiomyogenic differentiation of human pluripotent stem cells in scalable..., Kempf [/bib_ref]. A variety of serum-free defined media for cardiac differentiation in EBs have been tested [bib_ref] Controlling expansion and cardiomyogenic differentiation of human pluripotent stem cells in scalable..., Kempf [/bib_ref] [bib_ref] Human cardiovascular progenitor cells develop from a KDR+ embryonic-stem-cell-derived population, Yang [/bib_ref] [bib_ref] (eGFP/w) hESCs for isolation of human cardiac progenitors and cardiomyocytes, Elliott [/bib_ref]. StemPro-34 is a serum-free media firstly used to culture hematopoietic progenitors. Nowadays, it has been also used for EB cardiac differentiation [bib_ref] Stage-specific optimization of activin/nodal and BMP signaling promotes cardiac differentiation of mouse..., Kattman [/bib_ref] [bib_ref] Human cardiovascular progenitor cells develop from a KDR+ embryonic-stem-cell-derived population, Yang [/bib_ref]. In the absence of serum to induce cardiogenesis, different growth factors implicated in normal cardiac development, including BMP4, activin-A, FGF2, Wnt agonists and antagonists, and vascular endothelial growth factor, were tested. Cardiomyogenesis has been shown to be sensitive to the glycogen synthase kinase-3 (GSK-3) inhibitor CHIR99021 concentration, while aggregate size did not prove to be a prevalent factor among culture platforms [bib_ref] Controlling expansion and cardiomyogenic differentiation of human pluripotent stem cells in scalable..., Kempf [/bib_ref]. Successively, Yan and colleagues have demonstrated that cardiovascular spheroids derived from hPSCs and treated for 48 h with CHIR99021 express higher level of α-actinin and higher ratios of sarcomeric striations and Z-lines sarcomeres compared to 2D cultures [bib_ref] Cell population balance of cardiovascular spheroids derived from human induced pluripotent stem..., Yan [/bib_ref]. 3D aggregates of hiPSC-derived cardiomyocytes (hiPSC-CMs) in comparison with 2D cultures have displayed down-regulation of the genes involved in glycolysis and lipid biosynthesis and increased expression of the genes involved in the mitochondrial oxidative phosphorylation system [bib_ref] 3D aggregate culture improves metabolic maturation of human pluripotent stem cell derived..., Correia [/bib_ref]. Light microscopy inspection of EB derived from hPSCs-derived CMs has revealed that the cells in the border of the EB are elongated with a rod-shaped morphology, with cross-striations of myofibrils in the cytoplasm. On the contrary, the CMs located at the center of EB have shown a round shape, no striations, but numerous lipid vacuoles and glycogen accumulation. Electron microscopy has shown that all EBs contain similar cells with ultrastructural features of CMs, surrounded by a monolayer of epithelial cells. Among the differentiated CMs at the center of EBs, it has been possible to visualize small cellular spaces containing collagen fibrils and cellular debris. The CMs nuclei located in the center of EBs showed a slightly irregular outline, while those located in the periphery were oval [bib_ref] Cardiomyocytes derived from human embryonic and induced pluripotent stem cells: Comparative ultrastructure, Gherghiceanu [/bib_ref]. In addition, immunofluorescence staining has shown many different cardiacspecific markers, including troponin-T, α-actin, atrial myocyte-specific MLC-2a, ventricular myocyte-specific MLC-2v beating cells, and pacing cell-specific HCN4 [bib_ref] In vitro and in vivo differentiation of induced pluripotent stem cells generated..., Jiang [/bib_ref]. CMs induction from PSCs has resulted in mixtures of ventricular-like, atrial-like, nodal-like cells, and pacemaker-like cells defined by intracellular electrophysiological measurements of action potentials [bib_ref] Aalto-Setälä, K. Cardiac differentiation of pluripotent stem cells, Rajala [/bib_ref]. Indeed, the phenotype of CMs-derived PSCs is characterized by a combination of sub-populations but also presenting a different degree of maturity. As reported in several studies, it is possible to guide differentiation towards atrial-like or to ventricular-like CMs by modulating the retinoic acid and Wnt signaling pathways [bib_ref] Atrial-like cardiomyocytes from human pluripotent stem cells are a robust preclinical model..., Devalla [/bib_ref] [bib_ref] Direct differentiation of atrial and ventricular myocytes from human embryonic stem cells..., Zhang [/bib_ref] [bib_ref] Small molecule-mediated directed differentiation of human embryonic stem cells toward ventricular cardiomyocytes, Karakikes [/bib_ref]. In 2003, the identification and characterization of niches of endogenous CSCs in the adult mammalian heart had been associated with the expression of type III receptor tyrosine kinase c-kit (CD117 or SCFR-stem cell factor receptor) [bib_ref] Adult cardiac stem cells are multipotent and support myocardial regeneration, Beltrami [/bib_ref]. However, the adult heart contains a heterogeneous c-kit pos cell population, most of which displays blood and endothelial lineage-commitment [bib_ref] Molecular basis of functional myogenic specification of Bona Fide multipotent adult cardiac..., Cianflone [/bib_ref] [bib_ref] Heterogeneity of Adult Cardiac Stem Cells, Scalise [/bib_ref] [bib_ref] Atrial myxomas arise from multipotent cardiac stem cells, Scalise [/bib_ref] [bib_ref] Combining cell and gene therapy to advance cardiac regeneration, Marotta [/bib_ref] [bib_ref] Unravelling the Biology of Adult Cardiac Stem Cell-Derived Exosomes to Foster Endogenous..., Mancuso [/bib_ref]. Recently, it was demonstrated that only a fraction of c-kit pos cardiac cells, negative for CD45 and CD31, are enriched for multipotent CSCs [bib_ref] Adult cardiac stem cells are multipotent and robustly myogenic: C-kit expression is..., Vicinanza [/bib_ref] [bib_ref] Kit(cre) knock-in mice fail to fate-map cardiac stem cells, Vicinanza [/bib_ref]. These cells are distributed throughout the myocardium with the highest density in the atria and apex [bib_ref] Atrial myxomas arise from multipotent cardiac stem cells, Scalise [/bib_ref] [bib_ref] Myocyte death and renewal: Modern concepts of cardiac cellular homeostasis, Ellison [/bib_ref]. CSCs can be propagated over long-term culture and maintained in an undifferentiated, self-renewing, and stable state, without showing senescence or abnormal karyotype. CSCs are clonogenic in vitro, and when grown in differentiation media for endothelial, smooth muscle, and cardyomyocyte lineages, acquire phenotypic characteristics of these different cell types.
When CSCs grow in suspension, they form spheres of hundreds of cells, similar to the pseudo-embryoid bodies created by the neural stem cells (neurospheres), which by analogy were named "cardiospheres" [bib_ref] Adult cardiac stem cells are multipotent and support myocardial regeneration, Beltrami [/bib_ref]. Cardiospheres are self-assembling, multicellular floating clusters that represent a distinctive feature of multipotent cells. These cardiac-derived multicellular spheroids, are spontaneously formed by cardiac stem/progenitor cells and are cultured on a bacteriological dish or a poly(HEMA)-coated dish [bib_ref] Adult cardiac stem cells are multipotent and robustly myogenic: C-kit expression is..., Vicinanza [/bib_ref] [bib_ref] Isolation and expansion of adult cardiac stem/progenitor cells in the form of..., Chimenti [/bib_ref]. Immunostaining revealed that cardiospheres consist of c-kit pos CSCs and supporting cells bound together by ECM proteins and connexins [bib_ref] Adult cardiac stem cells are multipotent and support myocardial regeneration, Beltrami [/bib_ref] [bib_ref] Stem cell niches in the adult mouse heart, Urbanek [/bib_ref]. The architectural features of cardiospheres resemble those of in vivo stem cell niches, where stem cells are surrounded by supporting cells and linked by interactive ECM molecules [bib_ref] Cardiospheres recapitulate a niche-like microenvironment rich in stemness and cell-matrix interactions, rationalizing..., Li [/bib_ref]. shown that human cardiosphere-derived CSCs display the greatest myogenic differentiation potency in vitro compared with human BM-derived MSCs, adipose tissue-derived MSCs, and BM-derived mononuclear cells [bib_ref] Cardiospheres recapitulate a niche-like microenvironment rich in stemness and cell-matrix interactions, rationalizing..., Li [/bib_ref]. Marban and colleagues isolated a human cardiac progenitor cells (c-kit pos ) population from endomyocardial biopsy composed by a different sub-population expressing endothelial (CD31 and CD34) and mesenchymal markers (CD90, CD105). Cardiospheres derived from this heterogeneous population expressed c-kit in the core and the mesenchymal marker CD105, CD31, CD133, and MDR-1 on their periphery. These cardiospheres also expressed Connexin43, NKX2.5, and desmin whereas in the periphery show cardiomyocyte-specific sarcomeric proteins (cTNI and αMHC). Therefore, such cardiospheres possess a core characterized by the expression of cardiac stem/progenitor markers while the cells located at the periphery seem to represent spontaneous differentiation of precursor cells into endothelial, mesenchymal, or cardiomyogenic lineages [bib_ref] Validation of the cardiosphere method to culture cardiac progenitor cells from myocardial..., Davis [/bib_ref]. However, to boost CSC potential is indispensable to refocus the investigation on cloned Lin neg c-kit pos CSC, since they represent the true pool of multipotent adult cardiac stem/progenitor cells. Indeed, we have demonstrated that the efficiency of cardiomyogenic induction using clonally derived CSCs is remarkably higher when compared to heterogenous freshly isolated CSC-enriched cardiac cells [bib_ref] Adult cardiac stem cells are multipotent and robustly myogenic: C-kit expression is..., Vicinanza [/bib_ref]. Cloned CSC showed robust self-renewing and serial sub-cloning capacity, genome stability, and multi-lineage cardiac cell differentiation potential [bib_ref] Adult cardiac stem cells are multipotent and robustly myogenic: C-kit expression is..., Vicinanza [/bib_ref]. Importantly, they generate cardiospheres at high frequency, which give rise to secondary and tertiary cardiospheres. When placed in laminin-coated plastic dishes with LIF-deprived basic differentiation medium for 14 days, the peripheral cells of the spheres expressed high levels of smooth muscle as well as endothelial and cardiac lineages. Nevertheless, the CM derived from cloned CSCs still reach a biochemical myogenic differentiation that is intermediate between fetal and neonatal cardiomyocytes [bib_ref] In vitro CSC-derived cardiomyocytes exhibit the typical microRNA-mRNA blueprint of endogenous cardiomyocytes, Scalise [/bib_ref]. These data have also been confirmed by RNASeq expression profile analysis. Comparison of transcriptome and miRNome between clonal CSCs, contracting cardiomyocytes obtained from CSC-derived cardiospheres (iCMs), and adult cardiomyocytes (aCMs) has shown that CSCs are cardiomyogenic since they activate the entire gene network and specific cardiomyo-miRs characteristic of the aCMs phenotype. Yet, in concordance with the immature phenotype of iCMs, the expression of these miRNAs is still significantly lower than in aCMs, falling between CSCs and aCMs [bib_ref] In vitro CSC-derived cardiomyocytes exhibit the typical microRNA-mRNA blueprint of endogenous cardiomyocytes, Scalise [/bib_ref]. Cardiosphere attachment on laminin or 0,1% gelatin plates alters cell shape and geometry, leading to cell flattening and changes in the cytoskeleton and nuclear shape. Cardiospheres show upregulation of many stem cell-relevant factors, such as c-kit, SOX2, Nanog, IGF-1, and Tert, which play an essential role in the growth and maintenance of the undifferentiated state [bib_ref] Cardiospheres recapitulate a niche-like microenvironment rich in stemness and cell-matrix interactions, rationalizing..., Li [/bib_ref]. Interestingly, the expression of HDAC2, an important histone deacetylase, is also upregulated in cardiospheres. The mechanism underlying this cultureacquired enrichment in stemness is unclear but may be related to the recapitulation of a stem cell niche microenvironment and HDAC2-mediated epigenetic modification. In addition, ECM and adhesion molecules, including laminin-β1, integrin-α2, and E-selectin, are also upregulated in cardiospheres [bib_ref] Cardiospheres recapitulate a niche-like microenvironment rich in stemness and cell-matrix interactions, rationalizing..., Li [/bib_ref]. As other spheres, also cardiospheres face a limited supply of oxygen and nutrients to the center of the spheroid.
Moreover, the work of Smith et al. has shown the feasibility of generating human and porcine cardiospheres and expanding stem cells from routine endomyocardial biopsy specimens. Cardiosphere-derived cells (CDCs) were isolated from percutaneous endomyocardial biopsies of the adult patients and examined in vitro for biophysical and cytochemical evidence of cardiogenic differentiation [bib_ref] Regenerative potential of cardiosphere-derived cells expanded from percutaneous endomyocardial biopsy specimens, Smith [/bib_ref]. Human and porcine CDCs differentiated into electrically functional myocytes in vitro. Direct injection of human CDCs into the infarct border zone of SCID mice led to functional improvement and myocardial regeneration documented respectively by echocardiography and histology. CDCs transplantation induced both cardiomyogenesis and angiogenesis [bib_ref] Regenerative potential of cardiosphere-derived cells expanded from percutaneous endomyocardial biopsy specimens, Smith [/bib_ref]. Other groups have investigated the ability of cells to secrete cytokines and growth factors (i.e., VEGF, HGF, and IGF) and the potential contribution of paracrine mechanisms to the beneficial and protective effects of cardiosphere-derived cells in vitro and in vivo [bib_ref] Relative roles of direct regeneration versus paracrine effects of human cardiosphere-derived cells..., Chimenti [/bib_ref]. Numerous studies have demonstrated that CDCs can engraft, differentiate, and improve cardiac function post-MI in mice [bib_ref] Cardiospheres recapitulate a niche-like microenvironment rich in stemness and cell-matrix interactions, rationalizing..., Li [/bib_ref] [bib_ref] Relative roles of direct regeneration versus paracrine effects of human cardiosphere-derived cells..., Chimenti [/bib_ref] , rats [bib_ref] Magnetic targeting enhances engraftment and functional benefit of iron-labeled cardiosphere-derived cells in..., Cheng [/bib_ref] [bib_ref] Isolation and expansion of functionally-competent cardiac progenitor cells directly from heart biopsies, Davis [/bib_ref] [bib_ref] Noninvasive quantification and optimization of acute cell retention by in vivo positron..., Terrovitis [/bib_ref] , and pigs [bib_ref] Engraftment, differentiation, and functional benefits of autologous cardiosphere-derived cells in porcine ischemic..., Johnston [/bib_ref] [bib_ref] Intramyocardial injection of autologous cardiospheres or cardiosphere-derived cells preserves function and minimizes..., Lee [/bib_ref]. While using the intracoronary route, CDC-treated pigs had similar values of EF to placebo-treated animals [bib_ref] Engraftment, differentiation, and functional benefits of autologous cardiosphere-derived cells in porcine ischemic..., Johnston [/bib_ref] ; a slight improvement in the EF in post-MI pigs was obtained when CDCs were combined with βFGF [bib_ref] Controlled delivery of basic fibroblast growth factor promotes human cardiosphere-derived cell engraftment..., Takehara [/bib_ref].
## Cardioids
The hiPSC-CMs are useful for traditional two-dimensional (2D) monoculture. Still, this model usually lacks a complete functional maturation [bib_ref] Engineering adolescence: Maturation of human pluripotent stem cell-derived cardiomyocytes, Yang [/bib_ref] , differentiation ability, and gene expression, which hampers their capacity to accurately predict human biology and pathophysiology in some instances [bib_ref] Disease modeling and functional screening using engineered heart tissue, Mills [/bib_ref]. Nevertheless, this gap could be overcome due to the possibility of generating hiPSC-CMs three-dimensional 3D structures [bib_ref] Disease modeling and functional screening using engineered heart tissue, Mills [/bib_ref] [bib_ref] Screening out irrelevant cell-based models of disease, Horvath [/bib_ref] [bib_ref] Screening drug effects in patient-derived cancer cells links organoid responses to genome..., Jabs [/bib_ref] [bib_ref] Opportunities and challenges in phenotypic drug discovery: An industry perspective, Moffat [/bib_ref]. Therefore, cardiac organoids, named as "Cardioids", are generated from hiPSCs that self-organize into complex native-like organ [bib_ref] Cardiac organoid-A promising perspective of preclinical model, Zhao [/bib_ref] structures [bib_ref] Human cardiac organoids for the modelling of myocardial infarction and drug cardiotoxicity, Richards [/bib_ref] , representing faithfully many features of cardiac development stages [fig_ref] Figure 3: Cardioids recapitulate intrinsic self-organization and arrangement to form cardiac architecture and chamber-like... [/fig_ref]. Cardiac organoids are usually prepared by inducing the aggregation of specific cells suspended in a medium or embedded in the 3D gel matrix. Recently, it was shown that hiPSC-derived EBs adherent to collagen-conjugated hydrogels are more successful in forming myocardium-like tissue [bib_ref] Matrix rigidity-modulated cardiovascular organoid formation from embryoid bodies, Shkumatov [/bib_ref]. Collagen-conjugated hydrogel has a stiffness similar to that of myocardial tissue, and probably for this, they can better stimulate cell proliferation and then differentiation into cardiomyocytes. However, there are some limitations related to matrix use. First, using a soft matrix cannot transmit mechanical signals that are insoluble during cardiomyocyte differentiation. Therefore, the difference in gel stiffness results in a difference in the degree of differentiation.
Artificially engineered heart tissues were successfully obtained through many bioengineering approaches using scaffolds, molds, cell hydrogel matrices, and 3D-printed biomaterials [bib_ref] Disease modeling and functional screening using engineered heart tissue, Mills [/bib_ref] [bib_ref] Collagen-based engineered heart muscle, Tiburcy [/bib_ref] [bib_ref] Organs-on-a-Chip: A Fast Track for Engineered Human Tissues in Drug Development, Ronaldson-Bouchard [/bib_ref] [bib_ref] A Platform for Generation of Chamber-Specific Cardiac Tissues and Disease Modeling, Zhao [/bib_ref]. These approaches effectively measure contraction force, perform compound screens, and model structural muscle and arrhythmogenic disorders.
Nevertheless, many previous methods [bib_ref] Human cardiac organoids for the modelling of myocardial infarction and drug cardiotoxicity, Richards [/bib_ref] [bib_ref] Human-iPSC-Derived Cardiac Stromal Cells Enhance Maturation in 3D Cardiac Microtissues and Reveal..., Giacomelli [/bib_ref] [bib_ref] Human heart-forming organoids recapitulate early heart and foregut development, Drakhlis [/bib_ref] [bib_ref] Capturing Cardiogenesis in Gastruloids, Rossi [/bib_ref] do not recapitulate cardiacspecific self-organization to acquire an in vivo-like structure [bib_ref] Cardioids reveal self-organizing principles of human cardiogenesis, Hofbauer [/bib_ref]. To this aim, using a highthroughput approach, it was demonstrated that WNT and BMP, the central regulators of the cardiac specification [bib_ref] The deployment of cell lineages that form the mammalian heart, Meilhac [/bib_ref] , drive chamber-like self-organization [bib_ref] Cardioids reveal self-organizing principles of human cardiogenesis, Hofbauer [/bib_ref]. Moreover, the inhibition of WNT signaling at the cardiac mesoderm stage seems crucial for CM specification but is not necessary for cardioid self-organization. Recent work uses a protocol based on three sequential modulation steps of the WNT pathway (activation/inhibition/activation) at specific time points on suspension EBs to drive the production of significant heart-like structures in terms of organization, functionality, cardiac cell type complexity, ECM composition, and vascularization [bib_ref] Generating Self-Assembling Human Heart Organoids Derived from Pluripotent Stem Cells, Lewis-Israeli [/bib_ref]. The gene expression profile of cardioids showed upregulation of cardiac-specific genes and downregulation of WNT signaling at day 20 differentiation and significantly higher gene expression of transforming growth factor β (TGF-β) signaling (such as TGFβ1, TGFβ2, TGFβ3, TGFβR1, and TGFβR2) and cardiac-specific genes (MYL4, MYH7, and NKX2.5 [bib_ref] Engineering spatial-organized cardiac organoids for developmental toxicity testing, Hoang [/bib_ref]. Comparison of transcriptomes of 2D iPSC-CMs, 3D iPSC-CMs, human cardioids (hCOs) RNA-seq datasets of healthy human myocardium derived from fetal atria, fetal ventricles, and adult ventricles showed that hCOs best summarize cellular diversity and share the highest transcriptomic similarity to fetal myocardium, presenting fibroblast specific ECM organization, endothelial cell vascularization, and early immune cell regulation [bib_ref] Cardioids reveal self-organizing principles of human cardiogenesis, Hofbauer [/bib_ref].
Cardiomyocytes Epicardium Endothelial cells In cardioids, the patterning and morphogenesis of CM and endothelial cell lineages are controlled by the combination of specific cardiac and endothelial growth factors. Specifically, at the early time point of mesodermal differentiation, WNT and ACTIVIN were added to differentiation media, while the next step requires adding VEGF to move on both specification and patterning of the EC layer in cardiac mesoderm [bib_ref] Cardioids reveal self-organizing principles of human cardiogenesis, Hofbauer [/bib_ref]. Therefore, cardioids are a promising system to study the underlying mechanisms of CM and EC patterning and crosstalk in the context of a beating chamber. Moreover, cardiac organoids represent a booster to investigate mechanisms of human cardiogenesis. However, the self-assembly process used in most organoid procedures is still a limiting factor for a consistent generation of cardiovascular tissues. Mostly, this process is a random method resulting in heterogeneous organoids regarding cell composition, size, and shape [bib_ref] Engineering Stem Cell Self-organization to Build Better Organoids, Brassard [/bib_ref]. Nonetheless, the application of hiPSC-derived cardiac organoids to disease modeling shows countless advantages in clinical medicine, contributing to the investigation of a large variety of phenotypes and a robust technology applicable in drug development and screening [bib_ref] Disease modeling and functional screening using engineered heart tissue, Mills [/bib_ref] [bib_ref] Functional screening in human cardiac organoids reveals a metabolic mechanism for cardiomyocyte..., Mills [/bib_ref] [bib_ref] Modelling human cardiac diseases with 3D organoid, Nugraha [/bib_ref] [bib_ref] Drug Screening in Human PSC-Cardiac Organoids Identifies Pro-proliferative Compounds Acting via the..., Mills [/bib_ref] [bib_ref] Human Cardiac Organoids for Disease Modeling, Nugraha [/bib_ref] [bib_ref] Human Cardiac Organoids for Modeling Genetic Cardiomyopathy, Buono [/bib_ref].
## Self-organization of cardioids
Recently, cardiac organoids that incorporate an oxygen-diffusion gradient and that are stimulated by the neurotransmitter norepinephrine resemble the structure of the human heart after myocardial infarction (by mimicking the infarct, border, and remote zones) recapitulating transcriptomic, structural, and functional hallmarks of myocardial infarction [bib_ref] Human cardiac organoids for the modelling of myocardial infarction and drug cardiotoxicity, Richards [/bib_ref]. Furthemore, these organoids can model hypoxia-enhanced doxorubicin cardiotoxicity [bib_ref] Human cardiac organoids for the modelling of myocardial infarction and drug cardiotoxicity, Richards [/bib_ref]. The authors utilized non-adhesive agarose hydrogel molds to make hiPSC-CMs and non-myocyte mixtures used to form human cardioids. The human cardioids were placed into a hypoxic chamber and were treated with 1 µM noradrenaline to create an apoptotic gradient, which simulates the environment of myocardial infarction in vitro. The transcriptomic data and functional analyses showed that infarcted organoids recapitulate key aspects of metabolism in the human infarcted myocardium. However, the absence of inflammatory cells or the level of maturation of hiPSC-CMs obtained in vitro implies the lack of perfect simulation of heart failure in the infarct organoids [bib_ref] The interstitium in cardiac repair: Role of the immune-stromal cell interplay, Forte [/bib_ref]. Changes in transcriptome level explain the extent to which infarcted organoids could model responses of human cardiac tissue after infarction. From this perspective, iPSC-based systems are likely to be very helpful to model diseases across many adult tissues, including the heart [bib_ref] Organogenesis in a dish: Modeling development and disease using organoid technologies, Lancaster [/bib_ref] [bib_ref] Complex Tissue and Disease Modeling using hiPSCs, Passier [/bib_ref].
Another approach has been the generation of 3D cardioids from mouse embryonic stem cell-derived EBs using the substrate laminin-entactin (LN/ET), which includes components of the ECM in connective tissue, and exogenous FGF4 for induction of CM proliferation and cardiac chamber formation. Interestingly, when EBs were generated without the LN/ET complex, they failed to undergo morphological changes suggesting that the LN/ET complex promotes an extracellular environment for heart development. They also investigated the presence of cardiac muscle-specific structures, finding the appearance of intercalated discs, sarcomere structures with Z-bands, mitochondria, and desmosomes, Purkinje cells, and T-tubule [bib_ref] In vitro generation of functional murine heart organoids via FGF4 and extracellular..., Lee [/bib_ref]. Among all, the intercalated disc is involved in the coordination of muscle contraction. Therefore, these cardioids might possess contractile cardiac muscle cell properties. Thus, this heart organoid culture system provides a valid method to significantly improve regenerative medicine, study congenital heart diseases, and screen potentially dangerous drugs that cause cardiac defects.
A human in vitro model of acute cryoinjury has been also developed, which seems to be physiologically representative of the native immature human heart and exhibits an endogenous regenerative response. A different method includes generating a 3D microtissue system composed of the primary cardiac cell line, CMs, cardiac endothelial cells, and cardiac fibroblasts derived entirely from hiPSCs [bib_ref] Human-iPSC-Derived Cardiac Stromal Cells Enhance Maturation in 3D Cardiac Microtissues and Reveal..., Giacomelli [/bib_ref]. This system seems to promote the crosstalk between different cell lines improving cardiomyocytes maturation. Specifically, they identified some key mechanisms in the tri-cellular interactions that enhance CM maturation. One mechanism observed in tri-cellular interactions that enhance CM maturation includes the increased cAMP levels in CMs positively affecting the assembly of CX43 gap junctions [bib_ref] Human-iPSC-Derived Cardiac Stromal Cells Enhance Maturation in 3D Cardiac Microtissues and Reveal..., Giacomelli [/bib_ref]. The silencing of CX43 significantly reduced the structural organization of sarcomeres and the contraction duration [bib_ref] Human-iPSC-Derived Cardiac Stromal Cells Enhance Maturation in 3D Cardiac Microtissues and Reveal..., Giacomelli [/bib_ref]. Other factors such as paracrine effects or cell-extracellular matrix interactions may also contribute to CM specification and maturation.
Taken together, the main purpose of all the experimental procedures is to obtain fully mature hiPSC-CMs to mimic faithfully in vivo heart structures. Combining genomic editing technology and hCOs, it is possible to precisely modify and correct each mutation and realize innovative and personalized therapeutic platforms for disease modeling. Long and colleagues have used hCOs to show that dystrophin mutations impaired cardiac contractility and sensitivity to calcium concentration and that correcting mutations in the X-linked dystrophin gene by myoediting contractile dysfunction was partially-to-fully restored [bib_ref] Correction of diverse muscular dystrophy mutations in human engineered heart muscle by..., Long [/bib_ref]. hCOs can be used to study more electrophysiological phenomena, such as conduction and reentry, in arrhythmogenic syndromes, such as short QT syndrome, exploiting their ability to produce both spontaneous and induced action potentials with a higher conduction velocity [bib_ref] Modeling Reentry in the Short QT Syndrome With Human-Induced Pluripotent Stem Cell-Derived..., Shinnawi [/bib_ref].
The development of 3D stamping and bioprinting techniques also provides scaffolds to generate heart-on-a-chip. Engineered heart tissues and heart-on-a-chip methods allowed modeling of specific cardiac diseases, including Barth syndrome-associated cardiomyopathy [bib_ref] Modeling the mitochondrial cardiomyopathy of Barth syndrome with induced pluripotent stem cell..., Wang [/bib_ref] , Duchenne muscular dystrophy [bib_ref] Correction of diverse muscular dystrophy mutations in human engineered heart muscle by..., Long [/bib_ref] , and primary hypertension-induced left ventricular hypertrophy [bib_ref] A Platform for Generation of Chamber-Specific Cardiac Tissues and Disease Modeling, Zhao [/bib_ref]. Compared to standard 2D culture formats, engineered 3D heart tissues improve CM maturity and exhibit a more physiological 3D muscle environment [bib_ref] Human engineered heart tissue as a model system for drug testing, Eder [/bib_ref] , representing the new frontiers to study human cardiac regeneration in vitro.
Based on the consideration that the adult heart contains a niche of endogenous stem cells [bib_ref] The baby and the bath water: Adult cardiac stem cells revisited, Nadal-Ginard [/bib_ref] [bib_ref] Cardiac stem cell therapy towards the clinic: The way forward re-starts from..., Cianflone [/bib_ref] , these could be considered a potential source for organoid generation. Cardiac organoids can be generated by isolating adult stem cells from small biopsy specimens. CSCs can release their own cardiac morphogens promoting the formation of structures of greatly increased complexity and finely organized architecture. CSCs have the potential to differentiate into many or all the cell types present in the heart and acquire anatomic morphology such as chamber organization and atrioventricular specification. Therefore, cardioids generated from CSCs may provide a basis for the study of congenital and age-related cardiovascular disease [bib_ref] The role of endothelial progenitor and cardiac stem cells in the cardiovascular..., Thijssen [/bib_ref] [bib_ref] Adult Cardiac Stem Cell Aging: A Reversible Stochastic Phenomenon?, Cianflone [/bib_ref] [bib_ref] c-kit Haploinsufficiency impairs adult cardiac stem cell growth, myogenicity and myocardial regeneration, Aquila [/bib_ref] [bib_ref] Statins Stimulate New Myocyte Formation After Myocardial Infarction by Activating Growth and..., Cianflone [/bib_ref] [bib_ref] Targeting Cardiac Stem Cell Senescence to Treat Cardiac Aging and Disease, Cianflone [/bib_ref] [bib_ref] miRNA Regulation of the Hyperproliferative Phenotype of Vascular Smooth Muscle Cells in..., Torella [/bib_ref].
## Future directions
The findings discussed here illustrate that organoid technology is one of the most promising three-dimensional (3D) culture innovations in biological and medical research. Organoids display physiological functions specific to that organ and present cell organization similar to that of the organ itself [bib_ref] Scaffold-free human cardiac tissue patch created from embryonic stem cells, Stevens [/bib_ref]. However, organoids recapitulate some but not all aspects of the tissue of origin. Only a fraction of the in vivo cellular and physical environment is recapitulated. Mimicking the spatial and functional complexity observed in vivo is indeed a defined, yet not readily achievable, target to reach. Different approaches are being followed, from co-culture systems that combine different cell types using engineering approaches to mixing different types of already pre-patterned and differentiated structures to generate multiplex organ tissues. Organoid culture often requires Matrigel or another animal-based matrix extract to support cells to aggregate into 3D structures. The composition of these extracts can change between batches which may affect the reproducibility of experiments. In addition, they may transfer pathogens and are potentially immunogenic when transplanted to humans, limiting the application of organoids in a clinical transplantation protocol. This may be resolved by using clinical-grade collagen or a synthetic polyethylene glycol-based gel [bib_ref] Designer matrices for intestinal stem cell and organoid culture, Gjorevski [/bib_ref] [bib_ref] Functional engraftment of colon epithelium expanded in vitro from a single adult..., Yui [/bib_ref].
Organoids potentially provide alternative cellular sources for cell therapy transplantation, revolutionizing the future treatment of several chronic diseases or extending the time required for an organ transplant. They contribute to reducing the use of animal models and costs in the pharmaceutical industry. Despite that, these approaches tend to be expensive, work-intensive, and not readily scalable. A cardioids model, as a supplement to a preclinical model, can be used to simulate pathological processes and heart development effectively and to detect toxicity and side effects of drugs. Still, it has not yet wholly replaced animal models. Significantly, an important limitation of the hCOs is their inability to include all the cells found in the in vivo heart. hCOs, due to the lack of inflammatory cells and the use of immature hiPSC-CMs, cannot highlight the immune system's role in myocardial infarction, heart failure, or other cardiac diseases. Their ability to recapitulate heart development is still limited compared to other models, such as mice, even though they have the significant advantage of being human in origin rather than a surrogate animal model [bib_ref] A critical look: Challenges in differentiating human pluripotent stem cells into desired..., Fowler [/bib_ref]. There is large room for improvement in the technology, particularly in better recapitulating morphological and anatomical features and inducing the formation of effective vascular networks that can provide nutrients. A fundamental limitation of many organoid systems is a lack of a functional vascular network to facilitate the exchange of nutrients and waste material removal, as they rely solely on diffusion [bib_ref] Growth of engineered human myocardium with mechanical loading and vascular coculture, Tulloch [/bib_ref]. Innovative co-culture methods can better simulate the interaction between multiple cell types, which will be helpful to constructing specific-chambered hCOs containing ventricle-like and atrial-like structures and functions in the future.
The real possibility to manipulate the expression of a specific target gene through the genome editing strategy makes organoids a suitable preclinical tool to approximate the therapeutic efficiency of gene editing.
Moreover, the organoid-on-a-chip platform is an innovative technology to fabricate 3D organ models, which may bridge the gap between monolayer cell cultures and animal models. The artificial assembly of multiple organoids or the combination of organoids with cells from different tissue lineages becomes essential for analyzing other parameters that may change under specific conditions [bib_ref] Cardiac organoid-A promising perspective of preclinical model, Zhao [/bib_ref]. This approach could replicate the definite functions at the multiorgan level and the complex processes of drug metabolism and reaction [bib_ref] HiPSC-derived multi-organoids-on-chip system for safety assessment of antidepressant drugs, Yin [/bib_ref].
High-fidelity hCOs could be valuable in understanding human physiology, cardiac development, specific drug testing, disease modeling, and drug discovery. Therefore, in the last few years, 3D models have become increasingly more sensitive and are used to study many diseases such as neurogenerative diseases, cancer, and cardiomyopathy. Furthermore, 3D models in regenerative medicine are promptly developing and offer an unprecedented approach to personalized medicine. Both spheroids and organoids models are not only closer representations of in vivo tissues than 2D cell cultures but can also efficiently recapitulate human-specific biology in a dish and shed light on biological mechanisms, pathogenesis, and disease treatment. However, the current technologies for spheroid and organoid generation are limited by the inability to replicate the complex cell-cell interactions, cellular diversity, and microenvironment cues of tissues in vivo and lack of reproducibility. Further studies are required to refine the technology to go beyond and definitively abandon animal models to investigate critical cellular events in biology. Bioengineering strategies may provide new directions to overcome this issue.
# Data availability statement:
No new data were created or analyzed in this study, so sharing is not applicable to this article.
## Conflicts of interest:
The authors declare no conflict of interest.
## Abbreviations
The following abbreviations are used in this manuscript:
[fig] Figure 2: (a,b) Representative light microscopy image of monolayer culture and cardiospheres derived from c-kit pos /CD45 neg /CD31 neg hCSCs. Scale bar = 400 µm and 100 µm. (c) Representative confocal microscopy images of cardiospheres derived-c-kit pos /CD45 neg /CD31 neg hCSCs plated in cardiac differentiation media efficiently commit to cardiomyogenic cell lineages (TNNI, green). Nuclei are stained in blue (DAPI). Scale bar = 50 µm. Adapted from Scalise M. et al.[139]. [/fig]
[fig] Figure 3: Cardioids recapitulate intrinsic self-organization and arrangement to form cardiac architecture and chamber-like structures growing from pluripotent and cardiac stem cells. [/fig]
[fig] Author: Contributions: Conceptualization, M.S. and D.T.; methodology, M.S., F.M. and L.S.; visualization, L.S.; writing-original draft preparation, M.S., F.M., L.S., E.C. and C.M.; writing-review and editing, N.S., A.D.A., G.V., K.U. and D.T. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: This work was supported by grants from the Ministry of Education, University and Research (PRIN2017 2017NKB2N4_005), PON03PE00009_2-iCARE. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. [/fig]
[table] Table 1: Main methodologies of cardiosphere derivation with their characteristics. [/table]
|
Exploiting collider bias to apply two-sample summary data Mendelian randomization methods to one-sample individual level data
Over the last decade the availability of SNP-trait associations from genome-wide association studies has led to an array of methods for performing Mendelian randomization studies using only summary statistics. A common feature of these methods, besides their intuitive simplicity, is the ability to combine data from several sources, incorporate multiple variants and account for biases due to weak instruments and pleiotropy. With the advent of large and accessible fully-genotyped cohorts such as UK Biobank, there is now increasing interest in understanding how best to apply these well developed summary data methods to individual level data, and to explore the use of more sophisticated causal methods allowing for non-linearity and effect modification.In this paper we describe a general procedure for optimally applying any two sample summary data method using one sample data. Our procedure first performs a meta-analysis of summary data estimates that are intentionally contaminated by collider bias between the genetic instruments and unmeasured confounders, due to conditioning on the observed exposure. These estimates are then used to correct the standard observational association between an exposure and outcome. Simulations are conducted to demonstrate the method's performance against naive applications of two sample summary data MR. We apply the approach to the UK Biobank cohort to investigate the causal role of sleep disturbance on HbA1c levels, an important determinant of diabetes.Our approach can be viewed as a generalization of Dudbridge et al. (Nat. Comm. 10: 1561), who developed a technique to adjust for index event bias when uncovering genetic predictors of disease progression based on case-only data. Our work serves to clarify that in any one sample MR analysis, it can be advantageous to estimate causal relationships by artificially inducing and then correcting for collider bias.
[formula] a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 [/formula]
# Introduction
Mendelian randomisation (MR) is a technique used to test for, and quantify, the causal relationship between a modifiable exposure and health outcome with observational data, by using genetic variants as instrumental variables. MR circumvents the need to measure and adjust for all variables which confound the exposure-outcome association, and is therefore seen as an attractive additional analysis to perform alongside more traditional epidemiological methods. The following Instrumental Variable assumptions are usually invoked in order justify testing for a causal effect of an exposure X on a health outcome Y using a set of genes, G:
- IV1: G must be associated with X;
- IV2: G must be independent of unmeasured confounding between X and Y;
- IV3: G must be independent of Y conditional on X and all confounders of the X-Y relationship.
These assumptions are encoded in the causal diagram inFurther linearity and homogeneity assumptions are needed in order to consistently estimate the magnitude of the causal effect. When performing an MR-analysis it is best practice to pre-select SNPs for use as instruments using external data, in order to avoid bias due to the winner's curse. Subsequently, if the genetic variants are not as strongly associated with the exposure as in the discovery GWAS, assumption IV1 will only be weakly satisfied, which leads to so-called weak instrument bias. This issue is mitigated as the sample size increases as long as the true association is nonzero. When a genetic variant is in fact associated with the outcome through pathways other than the exposure, a phenomenon known as horizontal pleiotropy, this is a violation of assumptions IV2 and/or IV3. Horizontal pleiotropy is not necessarily mitigated by an increasing sample size and is also harder to detect. Its presence can therefore render very precise MR estimates hopelessly biased. Pleiotropy-robust MR methods have been a major focus of research in recent years for this reason.
## One-sample versus two-sample mr: pros and cons
Obtaining access to a single cohort with measured genotype, exposure and outcome data that is large enough to furnish an MR analysis has been difficult, historically. It has instead been far easier to obtain summary data estimates of gene-exposure and gene-outcome associations from two independent studies, and to perform an analysis within the 'two-sample summary data MR' framework (see .. This has made it an attractive option for the large scale pursuit of MR, through software platforms such as MR-Base. The relative simplicity of these methods (which resemble a standard meta-analysis of study results) and their ability to furnish graphical summaries for the detection and adjustment of pleiotropyhas also acted to increase their popularity. Indeed, the array of pleiotropy robust two sample summary data methods far outstrips those available for one sample individual level data MR analysis. A further advantage of two-sample over one-sample MR is that weak instruments bias causal estimates towards the null (it is often referred to as a 'dilution' bias for this reason) which is conservative. Dilution bias arises precisely because uncertainty in the SNP-exposure association estimates obtained from one cohort is independent of the uncertainty in SNP-outcome association estimates from a non-overlapping cohort. This makes the the SNP-exposure association uncertainty akin to 'classical' measurement errorand enables standard approaches such as Simulation Extrapolationor modified weightingto be used to adjust for its presence. In contrast, weak instruments bias MR estimates obtained from a one sample analysis towards the observational association because uncertainty in the SNP-exposure and SNP-outcome association estimates are correlated. This bias is harder to correct for and is potentially anti-conservative.
There are, however, many disadvantages of using two sample summary data compared to individual level data from a single sample MR, some examples of which are now given: The two-sample approach assumes the two cohorts are perfectly homogeneous. If the distribution of confounders is different between the samples, this can result in severe bias. Alternatively, it may be that the independence assumption is violated due to an unknown number of shared subjects across the two studies, which cannot be easily removed. Even when the homogeneity assumption is satisfied, two sample methods can give misleading results if the two sets of associations are not properly harmonized. Often, summary statistics from a GWAS have been adjusted for factors that might bias MR results, and the unadjusted data are not available. It may not be possible to source summary data on the exact population needed for a particular analysis, for example on either men or women only when looking at sex-specific outcomes). Finally, a richer array of analyses are possible with individual level data. For example, the estimation of non-linear causal effects across the full range of the exposure and the exploration of effect modification via covariates.
It is of course possible to naively apply summary data MR methods to the one-sample context, estimating both the gene-exposure and gene-outcome associations in the same sample, an analysis made increasingly easy by the advent of large open-access cohort studies such as the UK Biobank (UKB). This avoids problems with synthesising and harmonizing data from separate cohorts, but can result potentially anti-conservative weak instrument bias due to correlated error. A preliminary investigation has found that this naive approach is particularly bad for pleiotropy robust approaches such as MR-Egger regression. So far, there is no consensus on how best to implement summary data approaches in the one sample setting.
In this paper we propose a general method which we term 'Collider-Correction' that can reliably apply two-sample summary data MR methods to one-sample data, whilst maintaining the simplicity and appeal of the two-sample approach. Our method builds on the work of Dudbridge et. al., who proposed a method to correct for 'index event' (or collider) bias in genetic studies of disease progression, when all subjects included in the analysis have been diagnosed with the disease. In this setting, the analysis is open to contamination from collider bias. Our work serves to clarify that the procedure can be extended to any MR analysis where the aim is to estimate the causal effect, by artificially inducing collider bias in the observational association between X and Y and then correcting for it. This allows any two sample method to be used in a one sample design, thereby benefiting from the plethora of weak instrument and pleiotropy robust approaches available. We show that this approach is (a) statistically efficient compared to artificially splitting the data in two, and (b) will deliver consistent estimates of the causal effect whenever the assumptions of the underlying two-sample approach are satisfied.
Although our method builds on the work of Dudbridge et al, there are several major differences. Firstly, whilst Dudbridge et al focus on the unbiased estimation of the direct SNP-outcome associations, we treat these as nuisance parameters and focus instead on estimation of the causal effect. Secondly, whilst the underlying method we use is closely related to the approach of Dudbridge et al when the chosen method is MR-Egger regression, our paper shows that the underlying method can actually be applied to any MR method. Thirdly, whereas Dudbridge et al propose a solution to adjust for weak instrument bias within the specific context of an MR-Egger model which relies on the InSIDE assumption, we propose the use of a SIMEX procedure that can be applied to any regression model, including robust regression models that do not rely on InSIDE for identification. Of course, some recent two-sample approaches have weak-instrument robust weighting built into them, for example MR-RAPS. In this case, SIMEX adjustment is unnecessary.
A major reason for the emergence of weak instrument and pleiotropy robust two-sample MR methodsis the avoidance of winner's curse, by using one discovery GWAS for instrument selection and two additional data sources for the two-sample MR analysis (i.e. a 'three-sample' design). Although this removes winner's curse by design, it generally yields far weaker instruments. In practice, it may be hard to obtain data from three independent, homogenous cohorts to enact the three-sample approach, but a nice property of Collider-Correction is that it can be enacted with two-independent data sources rather than three. In Results, we apply Collider-Correction to 1 sample individual level UK Biobank data to investigate the causal role of sleep disturbance on HbA1c levels, using both overlapping and nonoverlapping GWAS data for instrument selection. In the former case winner's curse is seen to induce a dilution in the MR estimates that is not present in the latter case.
We see three scenarios where our Collider-Correction approach is applicable. Firstly, when interest lies in estimation of the causal effect of an exposure X on an outcome, Y and only summary data on 'YadjX' genetic associations are available (for example, waist/hip ratio adjusted for BMI from the GIANT consortium). The second is when researchers have direct access to individual level patient data. This is likely to become much more common over time as further international biobank studies follow the lead of UKB in opening up data access. Extracting the summary statistics for our approach then enables the efficient implementation of any two-sample method to the data. This is attractive because two-sample methods are currently more numerous than one sample methods, more familiar to researchers and more technically advanced (especially in their ability to adjust for weak instrument bias and pleiotropy). Furthermore, if one additional GWAS is available for instrument selection, Collider-Correction enables winner's curse, weak instrument bias and pleiotropy to be accounted for using two independent data sets rather than three. The third is when data custodians prefer not to grant direct access to individual level data, but are willing to provide the requisite summary statistics for implementing the Collider-Correction approach, safe in the knowledge that the individual-level data analysis can be performed whilst maintaining data security. Allowing large scale, rapid access to confidential data has obvious benefits to the research community and wider society, as demonstrated through initiatives such as OpenSA-FELY.
# Methods
To motivate ideas, we assume the following individual level data model for the exposure X and continuous outcome Y for subject i:
[formula] X i jG i ; U i ¼ X k j¼1 b XG j G ij þ b UX U i þ ε X ið1ÞY i jX i ; G i ; U i ¼ bX i þ X k j¼1 ¼ X k j¼1 ða j þ bb XGj ÞG ij þ ðbb UX þ b UY ÞU i þ bε X i þ ε Y i ¼ X k j¼1 b YGj G ij þ ε � Y ið3Þ [/formula]
Here, G i = (G i1 , . . ., G ik ) 0 represents a set of k variants that predict X i , β represents the target estimand, reflecting the causal effect of inducing a 1-unit change in the exposure on the outcome, and U represents unmeasured confounding predicting both X and Y. The variables ε X i , ε Y i represent independent residual error terms. Since the unmeasured confounder U is common to both X and Y, the total residual errors around X|G, Y|X, G and Y|G in Eqs (1)-(3) are correlated. This linear model tacitly assumes that the causal effect is the same for all individuals (that is, regardless of their observed exposure level). This is referred to as 'Homogeneity': it is an example of a fourth IV assumption that is needed to 'point identify' β (assumptions IV1-IV3 are sufficient to test for causality only). We re-write modelin 'reduced form' as modelto clarify that the underlying SNP-outcome association β YGj is equal to α j + ββ XGj . When the exposure is binary, so that X = 0, and X = 1 refer to being unexposed and exposed respectively, we can again identify β by assuming Homogeneity. This would mean that the effect of intervening and changing X from 1 to 0 is equal and opposite to the effect of intervening and changing X from 0 to 1.
The standard approach to estimating β with individual level data is Two Stage Least Squares (TSLS). This assumes that all instruments are valid (not pleiotropic), so that α j = 0 for all j. TSLS firstly regresses the exposure on all k genotypes simultaneously to derive an estimate for subject i's genetically predicted exposure:X i ¼ P k j¼1bXG j G ij , whereb XG j is the estimated association between SNP j and X. The outcome Y is then regressed onX i and its regression coefficient is taken as the causal estimateb. As explained inwhen the set of k SNPs which predict X are mutually independent (i.e. not in linkage disequilibrium), the TSLS estimate is asymptotically equivalent to the IVW estimateobtained by:
- Calculating the causal estimateb j by dividing the SNP-outcome associationb YGj (obtained from a regression of Y on G j ) by the SNP-exposure estimateb XGj for each SNP and;
- Performing an inverse variance weighted meta-analysis of the k individual causal estimates,
[formula] b 1 ; . . . ;b k . [/formula]
The inverse variance weights traditionally used make the simplifying assumption that the SNP-exposure associationb XGj is sufficiently precise that its uncertainty can be ignored. This is referred to as the No Measurement Error (NOME) assumption. This procedure is equivalent to fitting the following weighted regression model
[formula] b YGj ¼ bb XG j þ � YGjð4Þ [/formula]
where � YGj is the mean zero residual error with Varð� YGj Þ ¼ s 2 YGj ¼ Varðb YGj Þ and the intercept is constrained to zero. We will refer to this as the 'standard' IVW approach. It is commonly used in two sample summary data MR out of necessity because only summary statistics are available, but not typically in the one sample setting.
## Inducing collider bias into snp-outcome associations
Consider a regression of the outcome Y on G and X together (but not U). Under our assumed data generating model:
[formula] E½Y i jX i ; G i � ¼ b � X i þ X k j¼1 a � j G ij ;ð5Þ [/formula]
yielding estimated coefficientsb � andâ � 1 ; . . . ;â � k . Since X is a function of both G and U, conditioning on X induces a correlation between them. This is commonly referred to as 'collider bias'. Its presence contaminates the G j -Y association estimate with a contribution through U so thatâ � j is not a consistent estimate for α j . For the same reason,b � is not a consistent estimate for β. It instead reflects the causal effect, plus a contribution from X to Y via U. Such 'collider biased' analyses are usually avoided for this reason. However, it is in a special sense advantageous to fit model (5) because under models (1) and (2), a � j , α j , β � and β are linked through the following linear relation:
[formula] a � j ¼ a j þ ðb À b � Þb XG j þ y j ;ð6Þ [/formula]
where θ j is mean zero residual error with Varðy j Þ ¼ s 2
[formula] a � j ¼ Varðâ � j Þ (see S1(A) [/formula]
Text for a detailed derivation). This suggests the following algorithm for estimating the causal effect:
1. Regress Y on X and G to obtain the collider biased parameter estimatesb � andâ � 1 ; . . . ;â � k .
## Regress
[formula] X on G to obtain estimatesb XG1 ; . . . ;b XGk , wherê b XGj ¼ b XGj þ d j ;ð7Þ [/formula]
for independent residual error term δ j with mean zero and variance s 2
[formula] XGj ¼ Varðb XGj Þ; [/formula]
3. Fit the linear model:
[formula] E½â � j jb XGj � ¼ a 0 þ ðb À b � Þb XG jð8Þ [/formula]
under a user-specified loss function and pleiotropy-identifying assumption in order to obtain an estimate for the Collider-Correction term (β − β � ).
4. Adjust the observational estimate to obtain an estimate for the causal effect β via:
[formula] b ¼b � þ d b À b �ð9Þ [/formula]
The above procedure, which we call 'Collider-Correction' is a modification and generalisation of the Dudbridge approach. In step 3 and 4 we instead focus on estimation of the Collider-Correction term and the causal parameter β rather than, as Dudbridge et al do, the pleiotropic effects. Crucially, we clarify that, as long as the model for Y given X and G in step 1 is correctly specified, the correlation between the residual error in modeland residual error in the first-stage model (7) will have a mean of zero. To illustrate this we simulated 500 independent sets of data from models (1)-(2), each containing individual level data on 10,000 subjects. We fixed the number of SNPs to k = 50: each SNP was bi-allelic (taking the values 0,1 or 2), mutually uncorrelated with other SNPs, had a minor allele frequency of 0.3, and collectively explained 1.5% of the variance in the exposure. The correlation between the residual errors in model (1) and (2) was approximately 0.5 to reflect moderate confounding. SNP-exposure and SNP outcome association parameters β XGj and α j were generated from dependent distributions, so that their average correlation was approximately 0.45. This is a clear violation of the InSIDE assumption that the sample covariance d Covða j ; b XG Þ is zero. We then applied
Step 1 and 2 of the Collider-Correction algorithm to estimate theâ � j andb XGj terms., for a single simulated data set, the extent of correlation between the 50 β XGj and α j .shows across all 500 independent data sets, the sample correlation between the first stage residual d j ¼b XGj À b XGj and both:
- The Collider-Correction residual: y j ¼â � j À a j À ðb À b � Þb XGj (shown in black);
- The 'standard' SNP-outcome residual: � YGj ¼b YGj À a j À bb XGj (shown in red).
We see that the mean correlation of the Collider-Correction residual with 1st stage residual is zero whereas the mean correlation of the standard SNP-outcome residual with the 1st stage residual is 0.5. This residual error independence property is advantageous because it means that step 3 of the Collider-Correction algorithm can be implemented using any pleiotropy robust two-sample summary data MR method, where the estimand of interest is β − β � rather than the causal effect β directly. Crucially, the residual error independence property means that weak instrument bias will induce a dilution in the slope estimate d b À b � towards zero, because it can be viewed as a consequence of 'classical' measurement error. This makes it easy to quantify and correct for using standard methods, as we will subsequently discuss.
In the toy example above we purposefully generated the data so that the InSIDE assumption was violated across the entire set of SNPs to demonstrate that residual error independence does not rely on InSIDE. However, the success of any subsequently applied Collider-Corrected two sample approach in consistently estimating the causal effect β (i.e. so that it is asymptotically unbiased) will of course depend on the pleiotropy identifying assumption being met, just as if it were being applied in a standard two-sample setting. Although the Collider-Correction algorithm is generalisable in theory to any MR analysis method, we now describe several canonical implementations, which require that the InSIDE assumption is satisfied across either the entire set of SNPs or a subset of SNPs.
## Implementing collider-correction
Collider-Corrected IVW implementation. To implement the Collider-Corrected IVW approach we set the parameter α 0 to zero in Eqand estimate the slope (β − β � ) using weighted least squares via the model:
[formula] a � j ¼ ðb À b � Þb XG j þ y � jð10Þ [/formula]
where y � j is mean zero residual error with an assumed variance s 2
[formula] a � j ¼ Varðâ � j Þ. [/formula]
Note that under data-generating model (6) y � j is actually equal to θ j + α j . Under the assumption that the mean pleiotropic effect is zero and the InSIDE assumption is satisfied, the residual error independence property of Collider-Correction will mean that y � j is also independent of uncertainty in b XG j so that (β − β � ) can be consistently estimated. The IVW approach then quantifies additional uncertainty in the estimate for (β − β � ) due to the presence of pleiotropy, by increasing its variance by a factor ϕ proportional to the variance of the estimated residual Varðŷ j Þ whenever this variance is greater than 1. This is equivalent to fitting a multiplicative random effects model.
The IVW estimate uses '1st order weights' that ignore uncertainty in the SNP-exposure association estimate by assuming that its variance s 2 XGj � 0. This is referred to as the NO Measurement Error (NOME) assumption. When this is violated the estimate d b À b � from model (10) will be diluted towards zero by a factor of ð � F À 1Þ= � F, where:
[formula] � F ¼ X k j¼1b 2 XGj s 2 XGjð11Þ [/formula]
See Section 3.2 infor a more detailed explanation. Note that, whilst the Collider-Correction slope is diluted towards zero in the presence of weak instrument bias, the causal estimate itself is still biased toward the observational association estimateb � , because the causal effect calculated in Step 4 of the Collider-Correction algorithm is the sum ofb � and d b À b � . A simple and general method for weak instrument bias adjustment that can be applied directly to the IVW estimate from model (10) is Simulation Extrapolation (SIMEX). Under SIMEX, a parametric bootstrap is used to generate 'pseudo' SNP-exposure associations, each one centred on the observed estimate, but with an increasing amount of uncertainty (i.e. with larger and larger values of s 2 XGj ). This subsequently induces an increasing dilution in the IVW estimate for (β − β � ). A global model is then fitted to the entire set of simulated data in order to extrapolate back to the estimate for (β − β � ) that would have been obtained if there were no uncertainty in the SNP-exposure associations (i.e. s 2 XGj ¼ 0, NOME satisfied). SIMEX is attractive because it can be applied to any regression model (and hence many MR methods), and reliable software is available in standard packages, such as R and Stata.
Connecting IVW to LIML and MR-RAPS. An alternative to SIMEX in the special case of the IVW approach is to find the values of (β − β � ) and s 2 a � ¼ Varða � j Þ that minimises the weighted sum of squared residuals in the extended model (12):
[formula] a � j ¼ ðb À b � Þb XG j þ [/formula]
ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi
[formula] ðb À b � Þs 2 XGj þ s a � j þ zs 2 a � q � jð12Þ [/formula]
When z = 0 in (12), the pleiotropy variance s 2 a � ¼ Varða � j Þ is fixed to zero and the above procedure is equivalent to performing Limited Information Maximum Likelihood (LIML) with summary data (see Section 3.1 in. Furthermore, the weighted sum of squared residuals from (12) follows a w 2 LÀ 1 distribution when the assumption that Varða � j Þ ¼ 0 is satisfied, thus providing a simple weak instrument bias robust test for the presence of pleiotropy. This is referred to as the 'exact' Q statisticwhich is similar to the simulation-based MR-PRESSO test for 'global' pleiotropy.
Unfortunately, when pleiotropy is present so that Varða � j Þ 6 ¼ 0, then the LIML estimate will be biased. In order to account for both weak instrument bias and non-zero pleiotropy, z can be set to 1 so that the squared residual minimisation is over both (β − β � ) and s 2 a � . This is equivalent to applying 'MR-RAPS'when applied to the Collider-Correction summary statistics. MR-RAPS actually uses an approximation to the least-squares method because the maximum likelihood estimates are inherently unstable, this entails the use of a score function to proxy for the likelihood and a penalization term to dampen the effect of large residuals.
## Collider-corrected mr-egger implementation
In order to account for pleiotropy with a non-zero mean but under the InSIDE assumption, we could instead allow the intercept α 0 and slope (β − β � ) to be freely estimated via weighted least squares by fitting a Collider-Correction MR-Egger modela
[formula] � j ¼ a 0 þ ðb À b � Þb XG j þ y � j ;ð13Þ [/formula]
where y � j is mean zero residual error with an assumed variance s 2
[formula] a � j ¼ Varðâ � j Þ. [/formula]
Note that under data-generating model (7) y � j is actually equal to θ j + α j − α 0 . Using the same argument as for the IVW model, when InSIDE is satisfied this will consistently estimate the Collider-Correction slope (adjusted for α 0 ) and from there, the causal effect. Additional uncertainty due to pleiotropy can again be handled using a multiplicative random effects model. To assess the vulnerability of the MR-Egger regression estimates to weak instrument bias due to violation of the NOME assumption, we use the I 2 GX statistic:
[formula] I 2 GX ¼ Q GX À ðk À 1Þ Q GX ; where Q GX ¼ X k j¼1 ðb XGj À � b XGj Þ 2 s 2 XGjð14Þ [/formula]
The expected dilution in the Collider-Correction d b À b � due to weak instruments is equal to ðb À b � ÞI 2 GX . This can easily be adjusted for by applying SIMEX to model, just as for the IVW approach.
## Collider-corrected robust regression
IVW MR-Egger and MR-RAPS rely on the InSIDE assumption to consistently estimate the causal effect. This may be violated in practice, hence the rationale for the development of alternative, robust methods such as the Weighted Median. In the two-sample summary data context it can consistently estimate the causal effect if the majority of the 'weight' in the MR analysis stems from genetic variants that are not pleiotropic. That is, the existence of a SNP subset S is assumed for which d Cov j2S ða j ; b XGj Þ ¼ d Cov S ð0; b XGj Þ ¼ 0, but InSIDE is allowed to be violated for SNPs not in subset S. The downside of weighted median approach is that it is not directly equivalent to a regression model, which in turn means that we can not benefit from a procedure like SIMEX to perform a weak instrument bias adjustment. However, there is a close connection between the median and minimisation using a Least Absolute Deviation (LAD), or L1-norm. We therefore propose the use of LAD regressioninstead of least squares, at Step 3 of the Collider-Correction algorithm, with α 0 set to zero. This is close in spirit to the Weighted Median, and is amenable to SIMEX-adjustment too. The exact 'breakdown point' of LAD regression (or the proportion of pleiotropic SNPs above which LAD regression will not deliver a consistent estimate) depends on the data generating model, but is bounded between 1/k (k being the number of SNPs) and 1/2.
## Simulation studies
In order to confirm our theoretical results and assess the performance of the Collider-Correction algorithm, data sets of between 5000 and 50,000 individuals were generated under models (1) and (2) as described previously. Across all simulations:
- The causal effect of inducing a one-unit change in the exposure on the outcome, β, was set to 0.5 for all individuals;
- The correlation between the residual errors in model (1) and (2) was set to approximately 0.9 to reflect strong confounding;
- The observational estimate forb � and the true Collider-Correction term β − β � were approximately 1.12 and -0.62 respectively.
To showcase the ability of IVW-based approaches, MR-Egger regression and LAD regression, pleiotropy parameters and SNP-exposure associations were generated under three distinct models:
- For IVW simulations, pleiotropic effect parameters α 1 , . . ., α 50 were generated with a zero mean independently of the SNP-exposure associations β XG1 , . . ., β XG50 (InSIDE satisfied);
- For MR-Egger simulations, pleiotropic effect parameters α 1 , . . ., α 50 were generated with a non-zero mean independently of the SNP-exposure associations β XG1 , . . ., β XG50 (InSIDE satisfied);
- For LAD regression simulations, pleiotropic effect parameters α 1 , . . ., α 15 were generated with a non-zero mean dependent on the SNP-exposure associations β XG1 , . . ., β XG15 (with an average correlation of 0.5) whilst α 16 , . . ., α 50 were set to 0. InSIDE was therefore strongly violated across SNPs 1:15, satisfied across SNPs 16:50 and violated across all SNPs, respectively., for a range of sample sizes the average value across 1000 independent data sets of: (a) The standard IVW estimate (black line); (b) the SIMEX adjusted standard IVW estimate (blue line); (c) the Collider-Corrected IVW estimate (red line); (d) the Collider-Corrected IVW estimate with SIMEX correction (green line); (e) the TSLS estimate (orange line) and (f) the Collider-Corrected MR-RAPS estimate (implemented using the 'Tukey' penalization option). We see that methods (a), (c) and (e) give approximately the same answer, and are therefore hard to individually distinguish in the figure. The approximate equivalence of the TSLS and IVW approaches with uncorrelated SNPs is well known, but it is also reassuring that our two step approach is also equivalent. We also see that applying a direct SIMEX correction to method (a) (i.e. method (b)) dramatically increases the bias of the causal estimate beyond even that of the observational estimate for small sample sizes. This bias is slow to diminish as the sample size grows. This poor performance is because uncertainty in the SNP-exposure association estimates can not be viewed as classical measurement error within a standard IVW model. Conversely, we see that applying a SIMEX correction to the Collider-Corrected IVW estimate (c) (i.e method (d)) yields a steadily decreasing bias which is essentially zero when the mean F statistic across the instruments is larger than 5. The Collider-Corrected MR-RAPS estimate performs very well too, and is essentially unbiased for mean F statistics greater than 3.5.
# Ivw simulation results
## Plos genetics
## Exploiting collider bias in mendelian randomization
shows the true Collider-Correction multiplied by the expected dilution factor � F À 1 � F , which varies as a function of the sample size. The fact that the two lines are in good agreement indicates that the dilution in d b À b � can be perfectly predicted by the F-statistic formula and underlines why SIMEX can be used to correct for it.the performance of the IVW estimate implemented using the (one sample) Collider-Correction algorithm, versus that obtained from artificially splitting the data in two, and applying the 'standard' IVW approach. That is, calculating SNP-exposure associations in one half, SNP-outcome associations in the other half and combining in the usual manner. This ensures that the residual error independence property is satisfied, as it is for the one sample Collider-Correction approach. Results for each method are shown with and without SIMEX correction. We see that the absolute bias of the Collider-Correction implementations is less than that of the two-sample implementation. However, the two estimation strategies differ more substantially in terms of precision, as shown inCollider-correction of one sample data is shown to be far more efficient than sampling splitting.show the corresponding standard deviation and mean-squared error (a measure of accuracy that equals an estimate's variance plus the squared bias) for all IVWbased methods across the same set of simulations. They show that whilst the MR-RAPS estimate is less biased for small sample sizes than the Collider-Corrected IVW method with SIMEX adjustment, it is more variable and less accurate.
MR-Egger simulation results.for a range of sample sizes the average value across 1000 independent data sets of: (a) The standard (one-sample) MR-Egger estimate (black line); (b) the SIMEX adjusted standard MR-Egger estimate (blue line); (c) the Collider-Corrected MR-Egger estimate (red line) and (d) the Collider-Corrected MR-Egger estimate with SIMEX correction (green line). As the sample size increases the I 2 GX statistic increases from 0.1 to 0.5. This signals that the 50 SNPs get collectively stronger as a set of instruments within MR-Egger as the sample size increases, but even at the largest sample size we expect a dilution of 50% in the MR-Egger slope. Again, we see that standard and Collider-Corrected MR-Egger methods give the same results, but the two approaches differ greatly under SIMEX correction, with the SIMEX adjusted Collider-Corrected estimate being least biased. Inwe show how dilution in the Collider-Corrected slope estimate d b À b � for MR-Egger can be accurately quantified using the I 2 GX statistic, just as the F-statistic predicts the dilution for IVW. This explains why SIMEX adjustment works. . For comparison we also show (e) the standard IVW estimate: its bias does not approach zero as the sample size increases because of the presence of non-zero mean pleiotropy violating InSIDE, which is the very motivation for LAD regression. As in the previous simulations, standard and Collider-Corrected LAD regression give identical point estimates on average, but when SIMEX adjustment is applied the two estimates diverge substantially. Collider-Corrected LAD regression with SIMEX adjustment results in the least biased estimates of all.the mean dilution in the Collider-Corrected LAD regression estimate, versus that predicted by the IVW dilution factor � F À 1 � F . The fact that the observed dilution is below the expected IVW dilution illustrates that LAD regression is more vulnerable to weak instrument bias, because it is a less efficient but more robust technique. This emphasises the importance of being able to address its weak instrument bias.
Figs 8C and 9C show the corresponding standard deviation and mean-squared error for all LAD regression-based methods across the same set of simulations. The same pattern of higher variance but lower mean-squared error is seen for the SIMEX adjusted Collider-Corrected LAD regression approach as in the MR-Egger case.
The MR-RAPS approach can, in theory, consistently estimate the causal effect when a small proportion of SNPs are pleiotropic and violate the InSIDE assumption, as long as their contribution is strongly penalized by its robust loss function. In order to test this we also calculated the MR-RAPS estimate when applied to the simulated data for LAD regression. MR-RAPS was seen to work well for a proportion of simulated data sets, but its estimates were unstable: in many cases they were an order of magnitude larger than the true value of 0.5. To illustrate this,the distribution of its estimates at the largest sample size of 50,000 subjects, where it was most stable. Even in this case substantial instability is observed.
R code for reproducing the simulation study results is available in S1 Code.
## Results: assessing the causal role of insomnia on hba1c
Observationally, sub-optimal sleep (i.e., low sleep quantity and quality) has been found to be associated with hyperglycaemiaand increased diabetes risk. Insomnia, defined as difficulty initiating or maintaining sleep, is one of the most important indices of sleep quality. It has been associated with type 2 diabetes in observational studiesand in a previous Mendelian randomization study. However, it is unclear whether associations with insomnia are mediated through HbA1c in the general population, whose glucose levels may not meet the threshold criteria for a formal diabetes diagnosis. As such, we focus on a potentially causal role of insomnia on HbA1c, a well-established clinical assessment of long-term glycaemic regulation that is central to the diagnosis of diabetes. To address this question we use individual level data on approximately 320,000 individuals in UK Biobank to furnish a one sample Mendelian randomization study. Two hundred and forty-eight independent genetic variants at 202 loci were associated with self-reported insomnia at or below the standard genome-wide significance threshold (p-value<5 × 10 −8 ) in a recent GWAS of over 1.33 million UK Biobank and 23andMe individuals reported by Jansenwhich collectively explained 2.6% of the total trait variance. SNP-exposure associations were measured on the log-odds scale using logistic regression. Among this set of variants, 240 SNPs were in principle available for use as instruments in UK Biobank. In this cohort, participants were asked: "Do you have trouble falling asleep at night or do you wake up in the middle of the night?" with responses "Never/rarely", "Sometimes", "Usually", or "Prefer not to answer". Those who responded "Prefer not to answer" were set to missing. To reflect the Jansen analysis, the remaining entries were treated as a binary variable for insomnia symptoms, with "Never/rarely", "Sometimes", and "Usually" coded as 0, 0, and 1, respectively and a logistic regression performed. HbA1c measurements were obtained from a panel of biomarkers assayed from blood samples collected at baseline from UK Biobank participants. HbA1c (mmol/mol) was measured in red blood cells by HPLC analysis using Bio-Rad VARIANT II Turbo and log-transformed.
## Instrument selection and winner's curse
The mean F statistic for the 240 genetic instruments in the original GWAS was 41, but in order to avoid winner's curse we did not want to incorporate these estimates directly into our MR analysis. In UK Biobank the same SNPs had an � F of approximately 8.3 and an I 2 GX statistic of approximately 40%, meaning that the MR analysis was susceptible to bias due to both weak instrument and pleiotropy. This motivates the use of our Collider-Correction method for causal estimation. However, the original Jansen GWAS combined data from the UK Biobank (n = 386,533) and 23andMe (n = 944,477) using METAL. As such, there was an approximate 23% overlap between data used for SNP discovery and for estimation in our MR model. To additionally assess the impact of winner's curse for this reason we performed our subsequent analysis using (a) all 240 SNPs and (b) a subset of 112 SNPs that were only genome-wide significant using only the 23andMe portion of the Jansen data. Analysis (b) is completely protected from winner's curse whereas (a) is not. The downside of analysis (b) is that, with an � F of 6.8, it is even more susceptible to weak-instrument bias.
# Methods used
We applied the TSLS, IVW, MR-Egger, LAD regression and MR-RAPS approaches to the data. The IVW, MR-Egger and LAD regression approaches were implemented in three ways (1) The 'Standard' 1-sample approach (i.e. using all the data to estimate SNP-exposure and SNP-outcome associations); (2) the Collider-Correction algorithm and (3) Collider-Correction with SIMEX adjustment. Note that MR-RAPS incorporates an internal weak instrument bias adjustment and there is no need to additionally apply a SIMEX algorithm to it. Along with MR-RAPS, we refer to approach (3) as the 'gold-standard' methods.
## Causal estimates
SNP exposure associationsb XGj were obtained from a logistic regression of insomnia on the set of SNPs as well age at recruitment, sex, assessment centre, 10 genetic principal components, and genotyping chip. Estimates for collider biased SNP outcome associationsâ � j were obtained from a multivariable regression of HbA1c on observed insomnia severity, all genetic variants and the same additional covariates. This second regression additionally yielded an estimate for the collider biased observational association between insomnia severity and HbA1c of b � ¼ 0:012 (S.E. = 0.00057).the collider biased SNP-outcome associations versus the SNP-exposure associations for analysis (a). Overlaid on the plot are the weak-instrument and pleiotropy adjusted Collider-Correction slopes d b À b � estimated by the four gold standard methods. The Q statistic is 809 (df = 239) providing overwhelming evidence of heterogeneity due to pleiotropy. The 13 SNPs circled in black contribute a component to this global statistic with a bonferroni corrected p-value below (5/240)% and could therefore be classed as outliers. Adjusted causal effect estimates can be found in. Across all methods, we see a consistent picture: a unit increase in the log-odds of insomnia leads to an increase of between 0.17 and 0.24 units of log mmol/mol HbA1c. All estimates are further from the null than the collider biased observational association,b � . However the results highlight that, without weak-instrument adjustment, all summary data MR-methods are biased in the direction ofb � .(rows 6:10) andthe MR results for analysis (b) using only the 112 SNPs identified in Jansen from 23andMe data, which are immune to the dilution bias caused by winner's curse. These SNPs have a weaker mean F statistic of 6.88 but a higher I 2 GX statistic of 52%. All causal estimates are seen to increase when compared to analysis (a). This is because the winner's curse which is present in (a) leads to an over-estimation of the SNP-exposure association (which forms the denominator of the standard ratio estimate for β) and thus an underestimation of the causal effect. Again, across all methods, we see consistent evidence that the insomnia causally increases HbA1c.
In total there were 14 outlier SNPs (13 SNPs in analysis (a) andin analysis (b), respectively), which were investigated using the GWAS Catalog (https://www.ebi.ac.uk/gwas/), a full list of which can be found in S1(B) Text and S1 Data. Most of these SNPs are only associated with insomnia except rs10758593 (type 1 and type 2 diabetes), rs12917449 (type 2 diabetes), rs1861412 (BMI) and rs429358 (70+ traits). This provides some biological evidence for the existence of pleiotropy, which further underlines the utility of using robust methods that account for its presence.
# Discussion
In this paper we clarify how the principle of Collider-Correction offers a vehicle for applying any two-sample summary data MR method to one sample data, making it easy to account for both pleiotropy and weak instrument bias. Our method is closely related to the approach of Dudbridge et alfor genetic studies of disease progression, and primarily serves to emphasise that this procedure is in fact applicable to any MR analysis. We used our new method to provide important insights into the role of insomnia on glycated haemoglobin and, by extension, on incident diabetes.
A nice feature of our approach is that the Collider-Correction term β − β � will be large (and therefore the Collider-Corrected estimate will be clearly distinct from the observational association) precisely when there is strong confounding. Conversely, when there is weak confounding, or the confounding has been sufficiently adjusted for, β − β � will be zero and Collider-Correction estimate will equal the observational association. In this case, the observational association then becomes a consistent and likely very efficient estimate of the true causal effect. Collider-Correction therefore naturally promotes the triangulation and synthesis of observational and MR estimates, which can estimate the true causal effect with distinct but complementary assumptions. We showcased the Collider-Correction approach using four univariate MR approaches that estimate a single causal effect parameter. At the cutting-edge of MR methods research, new approaches are attempting to: estimate causal effects identified by different clusters of SNPs; simultaneously estimate causal effects via multiple exposures, or quantify non-linear effects of an exposure. The Collider-Correction algorithm can in principle be adapted to fit all of these multi-parameter approaches and this is an important topic of future research.
The insomnia data was affected by a small amount of winner's curse, which we removed by design in a sensitivity analysis by restricting our SNP set to those obtained from a purely independent data source. More sophisticated approaches to adjusting for winner's curse are possible by incorporating the original Discovery data. For example, Bowden and Dudbridgedescribe the most statistically efficient way to combine SNP discovery and validation data from two non-overlapping GWAS studies and remove winner's curse. As further work, we plan to extend this approach and combine it with Collider-Correction. Often in MR analyses the outcome of interest is binary and a logistic regression model is used in place of the linear model to estimate the causal effect on the odds ratio scale. In this case, the interpretation of causal estimates from a resulting Collider-Correction analysis will be more nuanced for the following reason. Even if we replaced the assumed linear outcome model in Eq (2) with a logistic model, so that β reflected the true causal log-odds ratio for a unit increase in the exposure experienced by each individual, the causal effect estimate (which is a population average) will be diluted by a factor that is proportional to the variance of the residual error in the model not explained by the genetically predicted exposure. This is due to the fact that the odds ratio is a non-collapsible measure. Although this dilution is a very general phenomenon that affects all logistic regression based analysis, three obvious options exist to the applied researcher if implementing Collider-Correction in the binary outcome case. The first would be to simply accept the interpretation of the causal estimate as a population average effect. The second would be to attempt to better approximate the individual causal effect by additionally adjusting for the first stage residual, (that is the observed exposure minus its genetically predicted value) in the second stage logistic model. This is referred to as the Control Function or adjusted IV approach. The third option would be to estimate the causal effect on a risk difference scale. Since the risk difference is a collapsible measure, individual and population average effects are the same. Risk difference estimates can be estimated either by fitting a linear probability model or by extracting the risk difference contrast from the logistic model. This latter approach can be implemented using the margins() package in R. A thorough investigation of the performance of Collider-Correction in the binary outcome setting is an interesting avenue for future research. Supporting information S1 Text. A: A formal proof of the Collider-Correction formulae. B: A list of outlying SNPs detected in analysis (a) and analysis (b) of the data example. (PDF) S1 Data. Additional functional information on the outlying SNPs detected in analysis (a) and analysis (b) of the data example. (ZIP) S1 Code. R scripts for re-creating the simulation study results in the paper. (R)
# Author contributions
Conceptualization: Ciarrah Barry, Frank Dudbridge, Jack Bowden.
|
Bilateral Sterile Pyogranulomatous Keratitis in a Dog
Purpose. To describe the clinicopathologic features of bilateral sterile pyogranulomatous keratitis in a 16-year-old spayed female rat terrier dog. Methods. The dog presented one year prior due to ulceration of the right and left corneas. The ulcers healed but plaques developed on both eyes which progressed, during the course of one year, to cover both the left and the right corneas. Due to the animal's loss of sight and its painful condition, bilateral enucleation was performed with submission of the eyes for histopathology. Results. Microscopic examination revealed bilateral pyogranulomatous keratitis absent of etiological organisms. Conclusions. To the authors' knowledge, this is the first documented case of bilateral sterile pyogranulomatous keratitis in a dog.
# Introduction
Pyogranulomatous inflammation is a chronic inflammatory response of both macrophages and neutrophils which can result from a variety of causes, including failure of the acute inflammatory response, unique biochemical characteristics (foreign material), and/or virulence factors in infectious agents such as B. dermatitidis, Nocardia, Mycobacterium spp., or Leishmania. Generally, this inflammation is induced by an exogenous agent which has entered the body, is recognized as foreign, and is subsequently surrounded and destroyed by reactive white blood cells. However, an etiologic agent is not always necessary, or found, to incite pyogranuloma formation. There are numerous cases of granulomatous or pyogranulomatous inflammation which are thought to be caused by immunoreactivity to an allergen or an autoantigen, including sperm granulomas, eosinophilic granuloma complex, and sterile pyogranuloma/granuloma syndrome (SPGS). In the eye, nodular granulomatous episcleritis (NGE) and necrotic scleritis are two conditions which are known to induce histiocytic to granulomatous inflammation in the absence of an etiologic agent. However, these conditions are primarily restricted to the sclera of domestic animals.
Here, we present the clinicopathologic characteristics of a 16-year-old spayed female rat terrier with a history of bilateral corneal plaques. Following therapeutic failure and progression of the plaques, bilateral enucleation and histopathology revealed the lesions to consist of sterile pyogranulomatous nodules which were restricted to the cornea. To the authors' knowledge, this is the first reported case of bilateral sterile pyogranulomatous keratitis in a dog.
## Case presentation
A 16-year-old spayed female rat terrier that had never traveled out of the state of Iowa presented to the Humboldt Veterinary Clinic in June of 2017 for squinting of the right eye (OD). There was neovascularization and congestion of the sclera OD and fluorescein stain identified two small corneal ulcers in the right (1 mm) and left (1.5 mm) eyes. Intraocular pressure was 10 mmHg OD and 9 mmHg on the left eye (OS). Topical gentocin drops were started at that time.
The dog represented four months later, in October of 2017, for continued squinting. Body temperature (99.5 ∘ F) and intraocular pressure were still normal, OD 16 mmHg and OS 19 mmHg. The corneal ulcers were no longer present, but the sclera still showed neovascularization and congestion of vessels. Gentocin eye drops were continued and oral nonsteroidal anti-inflammatories (NSAIDs), carprofen, were started. At recheck, two weeks later, the sclera of the right eye was still inflamed so a neomycin and polymyxin B sulfates with dexamethasone (.1%) eye drop were added to the treatment regimen OD. The dog was scheduled to be rechecked in two weeks but only presented seven weeks later, exhibiting bilateral white to red corneal plaques. The right cornea was completely covered ) while only 1/3 of the left cornea was effaced . Differentials for the plaques included neoplasia, fungal, protozoal, or bacterial infection, or autoimmune keratitis. To rule out cultivable infectious agents, both corneas were anesthetized, the corneal plaques were gently debrided, and culture swabs of both corneas were taken for bacterial and fungal culture. Neither culture resulted in any growth. The dog was prescribed erythromycin eye drops for both eyes and continued on the oral NSAID.
The animal was rechecked five days later. Complete blood cell count (CBC) and blood chemistry (chem) were performed. The CBC showed a leukopenia, 5.06X10 9 /L (normal = 6 -17 X10 9 /L) and mild lymphopenia, 0.89X10 9 /L (normal = 1 -4.8 X10 9 /L). The only significant finding in the blood chemistry was a high blood urea nitrogen (BUN) 48 mg/dL (normal = 7 -27 mg/dL); though, creatinine was normal. Prednisolone drops (1%) were prescribed for the eyes, as well as oral tramadol at a dose of 3.125 mg/kg every eight hours for pain management.
After five months of prednisolone therapy, the corneal plaques had continued to progress resulting in blindness in both eyes. As a result of this, the owners elected for bilateral enucleation. Following surgery, both eyes were submitted for histopathologic evaluation to the Surgical Pathology Service of the Department of Veterinary Pathology, Iowa State. A follow-up CBC/chem showed no abnormal findings in the CBC, while the serum chemistry panel revealed the abnormally high BUN, 47 mg/dL (normal = 7-27 mg/dL), with normal creatinine similar to previous results.
Histopathologic evaluation of both eyes revealed markedly expanded corneas due to pronounced hyperplasia of the surface epithelium, along with neovascularization, fibrosis of the stroma, and a marked infiltrate of numerous distinct pyogranulomas surrounded by a diffuse infiltrate of lymphocytes and plasma cells . The pyogranulomas were characterized by a peripheral zone of epithelioid macrophages, fewer lymphocytes and plasma cells, and a central zone of polymorphonuclear cells . Descemet's membrane was intact but there was a layer of plasma cells and lymphocytes subjacent to the endothelium. There was a moderate perivascular infiltrate of lymphocytes and plasma cells in the iris. Also, there was a substantial infiltrate of lymphocytes and plasma cells in the bulbar conjunctiva and around the scleral vessels at the limbus. Posterior segments of the eyes were unremarkable. There was no evidence of neoplasia.
Special stains were ordered to highlight potential infectious organisms within the pyogranulomas. However, no agents were identified with Haemotoxylin and Eosin (H&E), periodic acid Schiff (PAS), Grocott's methenamine silver (GMS), Giemsa, Gram and acid-fast stains. The animal responded well to the bilateral enucleation and is doing great post-op.
# Discussion
Previous reported cases of pyogranulomatous keratitis in the dog identified several protozoal organisms (ex. Toxoplasma gondii and Leishmania) as well as Acanthamoeba as causative agents with visualization of organisms followed by confirmatory ancillary testing. The Beckwith-Cohen et al. protozoal keratitis cases were all identified in dogs that had received long term, topical or systemic, immunosuppressive therapy for keratoconjunctivitis sicca before masses were noted to be progressing over the corneas. In the presented case, the lesions started progressing prior to steroid treatment and continued in spite of prescribed therapy. In fact, the left eye had not received treatment of any kind prior to its development of plaques. Had the animal been infected with a protozoan, Toxoplasma or Acanthamoeba organisms likely would have been visible in one of our many sections or stains. While Leishmania would be more difficult to identify with histopathology, the dog has never left the state of Iowa and lives in a nonendemic area. Furthermore, Giemsa stains, which have previously been shown to highlight Leishmania organisms in histopathology sections, were negative for amastigotes.
It is important to note that, while there was extensive pyogranulomatous inflammation in the cornea, there were also substantial populations of lymphocytes and plasma cells throughout the cornea with more mild aggregates in adjacent conjunctiva, iris, and subjacent to Descemet's membrane. This lymphoplasmacytic infiltrate was often peripheral to pyogranulomatous inflammation and could have developed in response to a number of stimuli including cytokines released from reactive macrophages, damaged keratocytes, or corneal epithelium. However, the lymphocytes may also be the primary reactors as can be seen in autoimmune disease and are thus responsible for pulling in additional inflammatory cells, or they may be regulatory T cells attempting to mitigate inflammation. An incomplete understanding of the relationship between canine T cell coreceptor phenotype and autoimmune function/importance makes further investigation of the described lymphocytic populations difficult and likely unrewarding.
The seemingly sterile nature of these lesions would suggest that this animal developed an immune response against one of the components of the corneal stroma. This is strongly supported by the absence of pyogranulomatous inflammation in any structure of the eye besides the cornea. In general, autoimmune keratitis is rare, as the cornea is an immune privileged site. Under homeostatic conditions there are numerous factors, such as anterior chamber-associated immune deviation (ACAID), that prevent naïve effector lymphocytes from responding to antigens which are normally expressed there. However, following injury to the eye and subsequent inflammation, this immune privilege can break down and result in an autoimmune response to proteins normally found in the anterior chamber or cornea. The result, in humans, is a sometimes markedly delayed T cell-mediated granulomatous inflammatory response of the uvea in both the injured eye (known as the exciting eye) and the uninjured contralateral eye (known as the sympathetic eye). This is known as sympathetic ophthalmia and while there are numerous publications on this phenomenon in humans, this autoimmune disease has never been described in a companion animal species.
Nodular granulomatous episcleritis (NGE) and necrotizing scleritis are two idiopathic inflammatory conditions of the eye characterized by numerous infiltrates of histiocytes and lymphocytes. Classically, these lesions are primarily confined to the sclera. However, recent work has presented three cases of corneocentric variants of NGE. It is important to note that the microscopic features of corneocentric NGE do not fit the described microscopic findings of this case. In necrotizing scleritis, lesions consist of coalescing granulomas centered on remnants of collagen with peripheral infiltrates of lymphocytes. This presentation is similar to the described case; though, we observed a large neutrophilic component to the inflammatory infiltrate and an absence of necrotic collagen within the center of granulomas. While it is possible that the presented case is a corneocentric variant of necrotizing scleritis, the disparity in microscopic characteristics warrants caution and likely requires the description of additional cases with similar features before this can be termed a true variant. As a result of this, the presented case is, to the authors' knowledge, the first reported example of bilateral pyogranulomatous keratitis in a dog without infectious stimulus or foreign substance.
## Conflicts of interest
The authors declare that they have no conflicts of interest.
|
Secreted mucins in pseudomyxoma peritonei: pathophysiological significance and potential therapeutic prospects
Pseudomyxoma peritonei (PMP, ORPHA26790) is a clinical syndrome characterized by progressive dissemination of mucinous tumors and mucinous ascites in the abdomen and pelvis. PMP is a rare disease with an estimated incidence of 1-2 out of a million. Clinically, PMP usually presents with a variety of unspecific signs and symptoms, including abdominal pain and distention, ascites or even bowel obstruction. It is also diagnosed incidentally at surgical or non-surgical investigations of the abdominopelvic viscera. PMP is a neoplastic disease originating from a primary mucinous tumor of the appendix with a distinctive pattern of the peritoneal spread. Computed tomography and histopathology are the most reliable diagnostic modalities. The differential diagnosis of the disease includes secondary peritoneal carcinomatoses and some rare peritoneal conditions. Optimal elimination of mucin and the mucin-secreting tumor comprises the current standard of care for PMP offered in specialized centers as visceral resections and peritonectomy combined with intraperitoneal chemotherapy. This multidisciplinary approach has reportedly provided a median survival rate of 16.3 years, a median progression-free survival rate of 8.2 years and 10-and 15-year survival rates of 63% and 59%, respectively. Despite its indolent, bland nature as a neoplasm, PMP is a debilitating condition that severely impacts quality of life. It tends to be diagnosed at advanced stages and frequently recurs after treatment. Being ignored in research, however, PMP remains a challenging, enigmatic entity. Clinicopathological features of the PMP syndrome and its morbid complications closely correspond with the multifocal distribution of the secreted mucin collections and mucin-secreting implants. Novel strategies are thus required to facilitate macroscopic, as well as microscopic, elimination of mucin and its source as the key components of the disease. In this regard, MUC2, MUC5AC and MUC5B have been found as the secreted mucins of relevance in PMP. Development of mucin-targeted therapies could be a promising avenue for future research which is addressed in this article.
# Introduction
With an estimated incidence of 1-2 out of a million, pseudomyxoma peritonei (PMP, ORPHA26790)-also known as adenomucinosis or gelatinous ascites-is listed as a rare disease by the NIH Office of Rare Diseases Research (ORDR) and National Organization for Rare Disorders (NORD). As an indolent neoplasm with unspecific manifestations, PMP tends to be misdiagnosed, or discovered at advanced stages. Moreover, it is a challenging entity with debilitating, even fatal complications. Despite a multidisciplinary approach composed of an extensive surgical procedure and chemotherapy, PMP frequently recurs and increasingly jeopardizes quality of life. Being ignored in research, however, PMP remains poorly understood and enigmatic. This is despite the fact that "rare diseases are rare, but rare disease patients are numerous"and those with progressive, life-threatening courses deserve to be further explored. To enhance outcomes of the conventional therapy, novel approaches based on in-depth understanding of the pathological processes and biological events in the pathogenesis of the disease are warranted. Since PMP and mucin are inextricably linked, any therapeutic intervention needs to properly target the mucin ectopy. In this article, the current knowledge on the crucial role of mucin in the pathogenesis of the disease is reviewed and a number of potential therapeutic strategies for mucin elimination in PMP are addressed.
## Definition
PMP is characterized by dissemination of mucinous tumor implants on peritoneal surfaces and progressive accumulation of mucinous ascites throughout the peritoneal cavity resulting in the so-called "jelly belly".
## Etiology
Since initial descriptions of PMP as a syndrome in association with an ovarian tumor [bib_ref] Klinische und anatomische untersuchungen zur lehre von den bauchgeschwülsten und der laparatomie, Werth [/bib_ref] or an appendiceal mucocele [bib_ref] Ueber das sogennante pseudomyxoma perotonei, Fraenkel [/bib_ref] , a pre-existing intraperitoneal mucinous neoplasm has been implicated as the primary cause of PMP. As follows, emerging evidence supports the appendiceal rather than ovarian origin of the disease.
## Classification
PMP has been broadly applied to a heterogeneous group of pathological conditions with a similar clinical presentation, with the site of the primary tumor, neoplastic phenotype of the peritoneal tumor cells and classification of the disease being a matter of controversy. However, attempts have been made to better define and classify the condition based on the clinicopathological characteristics. Ronnett et al. [bib_ref] Disseminated peritoneal adenomucinosis and peritoneal mucinous carcinomatosis. A clinicopathologic analysis of 109..., Ronnett [/bib_ref] suggested a classification of multifocal peritoneal mucinous tumors into three groups. These include disseminated peritoneal adenomucinosis (DPAM), peritoneal mucinous carcinomatosis (PMCA) and a third hybrid group called peritoneal mucinous carcinomatosis with intermediate or discordant features (PMCA I/D), also known as intermediate features group (IFG). According to this clinicopathological classification, DPAM includes histologically benign peritoneal lesions associated with ruptured appendiceal mucinous adenomas as well as those with similar pathology but lacking a demonstrable appendiceal adenoma. Subsequently, Sugarbaker defined PMP as a grade I mucinous adenocarcinoma that arises from an appendiceal adenoma. Later, they stipulated that the term PMP should be exclusively used to describe the clinical syndrome of mucinous ascites accompanied by a characteristic distribution of peritoneal mucinous tumors with the pathologic features of DPAM [bib_ref] Patients with pseudomyxoma peritonei associated with disseminated peritoneal adenomucinosis have a significantly..., Ronnett [/bib_ref]. Further studies strongly supported the notion that PMP is a generally low-grade, indolent neoplasm of appendiceal origin with rare distant metastases and unlikely involvement of solid organs [bib_ref] Pseudomyxoma peritonei is a disease of MUC2-expressing goblet cells, O'connell [/bib_ref] [bib_ref] Appendiceal mucinous neoplasms: controversial issues, Misdraji [/bib_ref]. Since neither adenoma nor adenocarcinoma precisely represents the neoplastic nature of PMP, "mucinous neoplasm of low malignant potential" and "low-grade appendiceal mucinous neoplasm" have been proposed in the literature as the alternative terms for pathological description of the PMP tumor [bib_ref] Pathologic diagnosis, origin, and natural history of pseudomyxoma peritonei, Buell-Gutbrod [/bib_ref].
## Pathogenesis
The pathological process starts with neoplastic transformation of the appendiceal goblet cells and subsequent formation of a primary mucinous tumor. While proliferating, tumor cells maintain their constitutive level of mucin expression. As a result, the overall secretion of mucin dramatically rises [bib_ref] Pseudomyxoma peritonei is a disease of MUC2-expressing goblet cells, O'connell [/bib_ref]. This is followed by intraluminal accumulation of mucin and eventual development of an appendiceal mucocele. A small perforation or rupture of the mucocele is the key event towards the development of PMP through which tumor cells gain access into the peritoneal cavity. Lacking cell surface adhesion molecules, the exfoliated tumor cells passively circulate with the peritoneal fluid and redistribute throughout the peritoneal cavity. As a result, tumor implants and mucin collections form at the peritoneal fluid reabsorption sites as well as within the dependent portions of the peritoneal cavity to create PMP's characteristic pattern of the peritoneal dissemination [fig_ref] Figure 1: Schematic representation of the events resulting in the development of PMP [/fig_ref]. Accumulating mucin increases intraabdominal pressure and compresses visceral organs. Furthermore, extensive involvement of the peritoneal surface promotes variable inflammatory and fibrotic responses in the peritoneal environment and hence the development of bowel obstruction as a fatal complication of the disease [bib_ref] New standard of care for appendiceal epithelial neoplasms and pseudomyxoma peritonei syndrome?, Sugarbaker [/bib_ref] [bib_ref] Pseudomyxoma Peritonei: Inflammatory Responses in the Peritoneal Microenvironment, Lohani [/bib_ref]. The detailed role of mucin in the pathogenesis of PMP will be discussed later.
## Clinical presentation
The disease is usually diagnosed after the age of 40, with an average age at diagnosis of 53 [bib_ref] Early-and long-term outcome data of patients with pseudomyxoma peritonei from appendiceal origin..., Chua [/bib_ref]. PMP clinically presents with a variety of unspecific and sometimes uncommon signs and symptoms. PMP's clinical manifestations can be roughly classified based on the disease progression [fig_ref] Table 1: Common presentations or incidental discovery of PMP on the basis of the... [/fig_ref] [bib_ref] Pseudomyxoma peritonei, Smeenk [/bib_ref]. In advanced disease, increased abdominal girth and complaints of abdominal pain related to intestinal obstruction are the most presenting symptom seen in 30-50% of the PMP patients as a result of disseminated mucinous tumor and ascites classically presenting at laparotomy with jelly belly. In less advanced disease, local symptoms are seen in 50-80% of PMP patients without jelly belly ascites and might correspond to the site of the primary tumor, such as appendicitis-like symptoms in 25% of cases, or the location of the peritoneal implants, including lower abdominal pain, pelvic pressure and gynecological complaints in females due to the ovarian deposits of the mucinous tumor in 20-30% of the patients [bib_ref] Pseudomyxoma peritonei syndrome, Sugarbaker [/bib_ref] [bib_ref] Clinical presentation of the Pseudomyxoma peritonei syndrome, Esquivel [/bib_ref]. Finally, PMP is coincidentally found in up to 20% of patients undergoing such procedures as laparotomy, laparoscopy or imaging for other medical conditions, e.g. hernia [bib_ref] Pseudomyxoma peritonei syndrome, Sugarbaker [/bib_ref] [bib_ref] Clinical presentation of the Pseudomyxoma peritonei syndrome, Esquivel [/bib_ref] [bib_ref] Recurrent mucinous adenocarcinoma of the ovary presenting as an inguino-labial hernia, Ben-Hur [/bib_ref] [bib_ref] Mucin deposits within inguinal hernia sacs: a presenting finding of low-grade mucinous..., Young [/bib_ref] [bib_ref] The use of laparoscopy in a case of appendiceal cystadenoma presenting as..., Edwards [/bib_ref] [bib_ref] Pseudomyxoma peritonei in a hernia sac: analysis of 20 patients in whom..., Esquivel [/bib_ref] [bib_ref] Pseudomyxoma peritonei: a case of mucinous adenocarcinoma of the appendix presenting as..., Rezkalla [/bib_ref] [bib_ref] Pseudomyxoma peritonei due to mucinous cystadenocarcinoma in situ of the urachus presenting..., Shinohara [/bib_ref] [bib_ref] Pseudomyxoma peritonei presenting as inguinal hernia, Campbell [/bib_ref] [bib_ref] Pseudomyxoma peritonei presenting with inguinal hernia, Ghidirim [/bib_ref] [bib_ref] Unusual contents of a scrotal swelling, Morris-Stiff [/bib_ref] , bladder tumor [bib_ref] Computed tomography of mucinproducing adenocarcinoma of the appendix presenting as a bladder..., Skaane [/bib_ref] , jelly like material in urine [bib_ref] Pseudomyxoma peritonea: uncommon presentation, Gandhi [/bib_ref] , total uterovaginal prolapse [bib_ref] Mucinous cystadenocarcinoma of the appendix with pseudomyxoma peritonei presenting as total uterine..., Snyder [/bib_ref] , recurrent rectal cancer [bib_ref] Pseudomyxoma peritonei presenting as recurrent rectal cancer: a preventable condition?, Newman [/bib_ref] , pregnancy [bib_ref] Pseudomyxoma peritonei originating from colorectal cancer during pregnancy, Koyama [/bib_ref] and cesarean section [bib_ref] Incidental finding of pseudomyxoma peritonei at primary cesarean section, Abdu [/bib_ref]. PMP cases presenting with an ulcerated skin fistula on the right flank [bib_ref] Pseudomyxoma retroperitonei presenting with a skin fistula, Cakmak [/bib_ref] or a subcutaneous nontender umbilical nodule [bib_ref] Pseudomyxoma peritonei: a rare presentation as an umbilical nodule, Srinivasaiah [/bib_ref] have also been reported.
Diagnosis Imaging -Ultrasound and parallel fine needle biopsy. Ultrasound is accessible and inexpensive. However, conclusions cannot be drawn from ultrasound alone since the mucinous ascites resembles free intraperitoneal fluid [bib_ref] Pseudomyxoma peritonei, Smeenk [/bib_ref]. Although cytological investigation of the mucin collections seems to be a useful accompanying procedure, sampling errors, dry taps and false negative results due to low amount of mucin and/or low cellular density are considered as the disadvantages of parallel fine needle biopsy [bib_ref] Pseudomyxoma peritonei, Smeenk [/bib_ref]. Non-surgical procedures Infertility investigation [bib_ref] Clinical presentation of the Pseudomyxoma peritonei syndrome, Esquivel [/bib_ref] Postmenopausal bleeding [bib_ref] Clinical presentation of the Pseudomyxoma peritonei syndrome, Esquivel [/bib_ref] [bib_ref] Incidental ultrasound diagnosis of pseudomyxoma peritonei in an asymptomatic woman, Khan [/bib_ref] Abnormal Pap test [bib_ref] Clinical presentation of the Pseudomyxoma peritonei syndrome, Esquivel [/bib_ref] Others [bib_ref] Clinical presentation of the Pseudomyxoma peritonei syndrome, Esquivel [/bib_ref] Deep vein thrombosis, rectal bleeding, anaemia Laparoscopy or laparotomy [bib_ref] Clinical presentation of the Pseudomyxoma peritonei syndrome, Esquivel [/bib_ref] Hernia repair, fibroids, colon cancer, tubal ligation, nephrectomy, abdominal aortic aneurysm repair -Computed tomography. Computed tomography (CT) remains the most widely used imaging modality in PMP. Higher densities of mucinous ascites compared to the nonmucinous collections [bib_ref] Computed tomography of fluid collections within the abdomen, Kreel [/bib_ref] , characteristic pattern of the mucinous accumulation [bib_ref] CT in pseudomyxoma peritonei: a review of 17 cases, Sulkin [/bib_ref] and the extent of the disease for preoperative planning and prognostic purposes [bib_ref] Pseudomyxoma peritonei, Smeenk [/bib_ref] can be evaluated by CT.
-Other imaging methods. Magnetic resonance imaging (MRI) has been described to show the location of mucocele and its morphologic criteria identically to CT. T1and T2-weighted MRI are more sensitive in distinguishing between mucin and fluid ascites [bib_ref] Magnetic resonance imaging of pseudomyxoma peritonei, Buy [/bib_ref]. Although positron emission tomography (PET) has been suggested for predicting the peritoneal dissemination [bib_ref] The diagnostic value of PET-CT for peritoneal dissemination of abdominal malignancies, Yang [/bib_ref] and preoperative evaluation of pathological grade and potential for complete cytoreduction [bib_ref] Pseudomyxoma peritonei: role of 18 F-FDG PET in preoperative evaluation of pathological..., Passot [/bib_ref] , its value in PMP remains controversial [bib_ref] Intraperitoneal hyperthermic chemotherapy for peritoneal surface malignancy: current status and future directions, Stewart [/bib_ref] [bib_ref] Use of FDG-PET imaging for patients with disseminated cancer of the appendix, Rohani [/bib_ref].
## Circulating tumor markers
Although relatively non-specific, the following tumor markers have been reported to be of value in PMP:
-Carcinoembryonic antigen (CEA). This tumor marker has been shown to serve as a valuable diagnostic [bib_ref] Utility of CEA and CA 19-9 tumor markers in diagnosis and prognostic..., Carmignani [/bib_ref] and prognostic tool [bib_ref] Utility of CEA and CA 19-9 tumor markers in diagnosis and prognostic..., Carmignani [/bib_ref] [bib_ref] Elevated tumour markers prior to complete tumour removal in patients with pseudomyxoma..., Alexander-Sefre [/bib_ref] in the management of PMP.
-Carbohydrate antigen 19.9 (CA19.9). Practical value of CA19.9 in diagnostic [bib_ref] Utility of CEA and CA 19-9 tumor markers in diagnosis and prognostic..., Carmignani [/bib_ref] and prognostic [bib_ref] Utility of CEA and CA 19-9 tumor markers in diagnosis and prognostic..., Carmignani [/bib_ref] [bib_ref] Prognostic value of baseline and serial carcinoembryonic antigen and carbohydrate antigen 19.9..., Van Ruth [/bib_ref] [bib_ref] Prognostic value of circulating tumor markers in patients with pseudomyxoma peritonei treated..., Baratti [/bib_ref] [bib_ref] Inflammatory markers in blood and serum tumor markers predict survival in patients..., Chua [/bib_ref] [bib_ref] Carbohydrate antigen 19-9 (CA 19-9) is an independent prognostic indicator in pseudomyxoma..., Koh [/bib_ref] evaluation of PMP has been reported.
-Carbohydrate antigen 125 (CA125) also known as MUC16. Although suggested as a marker with diagnostic sensitivity for PMP [bib_ref] Prognostic value of circulating tumor markers in patients with pseudomyxoma peritonei treated..., Baratti [/bib_ref] , CA125 is not widely used as a tumor marker for PMP. Instead, as a gynecological tumor marker, it is recommended for exclusion of an ovarian neoplasm [bib_ref] Pseudomyxoma peritonei, Smeenk [/bib_ref].
These tumor markers are also used as a baseline value for postoperative follow-up. Moreover, they have been reported as predictors of the completeness of cytoreductive surgery (CRS) [bib_ref] Elevated tumour markers prior to complete tumour removal in patients with pseudomyxoma..., Alexander-Sefre [/bib_ref] [bib_ref] Prognostic value of circulating tumor markers in patients with pseudomyxoma peritonei treated..., Baratti [/bib_ref] [bib_ref] Inflammatory markers in blood and serum tumor markers predict survival in patients..., Chua [/bib_ref] [bib_ref] Circulating tumor markers: predictors of incomplete cytoreduction and powerful determinants of outcome..., Kusamura [/bib_ref] , a significant prognostic factor for PMP.
# Histopathological analysis
Apart from the characteristic feature of PMP as acellular to paucicellular mucin pools with variable amounts of neoplastic mucinous epithelium, immunohistochemical markers can help to identify the organ of origin. These include positive cytokeratin 20 (CK20), CEA, caudal-type homeobox protein 2 (CDX-2), and MUC2 as well as negative cytokeratin 7 (CK7) and CA125 [bib_ref] Pseudomyxoma peritonei, Smeenk [/bib_ref]. Of particular interest is the secreted mucin MUC2 that is extensively positive in the patients. Although MUC2 has been suggested as a biological marker of PMP [bib_ref] MUC2 is a molecular marker for pseudomyxoma peritonei, O'connell [/bib_ref] [bib_ref] A specific cadherin phenotype may characterise the disseminating yet non-metastatic behaviour of..., Bibi [/bib_ref] [bib_ref] Pseudomyxoma peritonei: is disease progression related to microbial agents? A study of..., Semino-Mora [/bib_ref] [bib_ref] Exploring the peritoneal surface malignancy phenotype-a pilot immunohistochemical study of human pseudomyxoma..., Flatmark [/bib_ref] [bib_ref] Histological origin of pseudomyxoma peritonei in Chinese women: clinicopathology and immunohistochemistry, Guo [/bib_ref] , its significance as a prognostic factor is a matter of controversy [bib_ref] Pseudomyxoma peritonei: biological features are the dominant prognostic determinants after complete cytoreduction..., Baratti [/bib_ref].
## Differential diagnosis
The main entity to be considered in the differential diagnosis of PMP is peritoneal mucinous carcinomatosis arising from a primary mucinous carcinoma. Other differential diagnoses reported in the literature include endometriosis with myxoid change [bib_ref] Endometriosis with myxoid change. A case simulating pseudomyxoma peritonei, Clement [/bib_ref] , melioidosis (a lethal infectious disease caused by Burkholderia pseudomallei) [bib_ref] Melioidosis presenting as pseudomyxoma peritonei: yet another pretense of the great mimicker:..., Subramaniam [/bib_ref] and those with abdominal CT resemblance, e.g. extensive abdominal plexiform neurofibromatosis [bib_ref] Abdominal plexiform neurofibromatosis simulating pseudomyxoma peritonei on computed tomography, Mirich [/bib_ref].
## Treatment
Traditional treatment consists of repetitive surgical debulking. Due to the presence of tumor deposits after the first debulking surgery, this approach could results in short term palliation with imminent recurrence or progression, hence redo procedures and a shorter 5-to 10-year overall survival (OS) rate of approximately 50% [bib_ref] Pseudomyxoma peritonei. Long-term patient survival with an aggressive regional approach, Gough [/bib_ref] [bib_ref] Long-term survival following treatment of pseudomyxoma peritonei: an analysis of surgical therapy, Miner [/bib_ref] [bib_ref] Survival of patients with pseudomyxoma peritonei treated by serial debulking, Jarvinen [/bib_ref]. A more aggressive approach by Sugarbaker [bib_ref] Cytoreductive surgery and intraperitoneal chemotherapy with peritoneal spread of cystadenocarcinoma, Sugarbaker [/bib_ref] [bib_ref] Cytoreductive surgery and peri-operative intraperitoneal chemotherapy as a curative approach to pseudomyxoma..., Sugarbaker [/bib_ref] utilizes peritonectomy and visceral resections, called cytoreductive surgery (CRS), in combination with hyperthermic intraperitoneal chemotherapy (HIPEC) that is featured by direct targeting of the microscopic disease, locoregional drug availability, minimal systemic exposure and improved drug penetration through hyperthermia. This strategy comprises the current standard of care for PMP with known benefits well-documented by us [bib_ref] Early-and long-term outcome data of patients with pseudomyxoma peritonei from appendiceal origin..., Chua [/bib_ref] [bib_ref] Inflammatory markers in blood and serum tumor markers predict survival in patients..., Chua [/bib_ref] [bib_ref] Carbohydrate antigen 19-9 (CA 19-9) is an independent prognostic indicator in pseudomyxoma..., Koh [/bib_ref] [bib_ref] Review of patients with peritoneal malignancy treated with peritonectomy and heated intraperitoneal..., Hadi [/bib_ref] [bib_ref] Cytoreductive surgery and perioperative intraperitoneal chemotherapy for pseudomyxoma peritonei from appendiceal mucinous..., Yan [/bib_ref] [bib_ref] Longterm survival in patients with pseudomyxoma peritonei treated with cytoreductive surgery and..., Chua [/bib_ref] [bib_ref] The St George Hospital peritoneal surface malignancy program-where are we now?, Chua [/bib_ref] [bib_ref] Critical assessment of risk factors for complications after cytoreductive surgery and perioperative..., Saxena [/bib_ref] [bib_ref] Palliative effects of an incomplete cytoreduction combined with perioperative intraperitoneal chemotherapy, Chua [/bib_ref] [bib_ref] Secondary cytoreduction and perioperative intraperitoneal chemotherapy after initial debulking of pseudomyxoma peritonei:..., Chua [/bib_ref] [bib_ref] Evaluation of the cost-effectiveness of cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (peritonectomy)..., Chua [/bib_ref] [bib_ref] Determining the association between preoperative computed tomography findings and postoperative outcomes after..., Chua [/bib_ref] [bib_ref] Upfront compared to delayed cytoreductive surgery and perioperative intraperitoneal chemotherapy for pseudomyxoma..., Chua [/bib_ref] [bib_ref] Early recurrence of pseudomyxoma peritonei following treatment failure of cytoreductive surgery and..., Chua [/bib_ref] [bib_ref] Quality of life study following cytoreductive surgery and intraperitoneal chemotherapy for pseudomyxoma..., Kirby [/bib_ref] and others [bib_ref] Results of treatment of 385 patients with peritoneal surface spread of appendiceal..., Sugarbaker [/bib_ref] [bib_ref] Peritonectomy and intraperitoneal hyperthermic perfusion (IPHP): a strategy that has confirmed its..., Deraco [/bib_ref] [bib_ref] Cytoreductive surgery with intraperitoneal chemohyperthermia for the treatment of pseudomyxoma peritonei: a..., Loungnarath [/bib_ref] [bib_ref] Early results of surgery in 123 patients with pseudomyxoma peritonei from a..., Murphy [/bib_ref] [bib_ref] Survival analysis of pseudomyxoma peritonei patients treated by cytoreductive surgery and hyperthermic..., Smeenk [/bib_ref] [bib_ref] Consensus statement on the loco-regional treatment of appendiceal mucinous neoplasms with peritoneal..., Moran [/bib_ref] [bib_ref] Pseudomyxoma peritonei: a French multicentric study of 301 patients treated with cytoreductive..., Elias [/bib_ref] [bib_ref] Outcome differences between debulking surgery and cytoreductive surgery in patients with Pseudomyxoma..., Andreasson [/bib_ref] [bib_ref] Complete cytoreduction for pseudomyxoma peritonei is optimal but maximal tumor debulking may..., Dayal [/bib_ref] [bib_ref] Cytoreduction and HIPEC in the Netherlands: nationwide long-term outcome following the Dutch..., Kuijpers [/bib_ref] [bib_ref] Improved survival of patients with pseudomyxoma peritonei receiving intraperitoneal chemotherapy with cytoreductive..., Mcbride [/bib_ref]. Other proposed modalities include laparoscopy for less advanced disease [bib_ref] Laparoscopic management of pseudomyxoma peritonei secondary to adenocarcinoma of the appendix, Raj [/bib_ref] [bib_ref] Risk-reducing laparoscopic cytoreductive surgery and hyperthermic intraperitoneal chemotherapy for low-grade appendiceal mucinous..., Fish [/bib_ref] , whole abdominopelvic radiotherapy (WAPRT) as a palliative treatmentand use of mucolytic [bib_ref] Pseudomyxoma peritoneinonoperative management and biochemical findings. A case report, Green [/bib_ref] [bib_ref] Pseudomyxoma peritonei: possible prevention of mucinous ascites by peritoneal lavage, Piver [/bib_ref] [bib_ref] Mucolysis by ascorbic acid and hydrogen peroxide on compact mucin secreted in..., Pillai [/bib_ref] [bib_ref] Potential mucolytic agents for mucinous ascites from pseudomyxoma peritonei, Pillai [/bib_ref] , antibacterial [bib_ref] Pseudomyxoma peritonei: is disease progression related to microbial agents? A study of..., Semino-Mora [/bib_ref] [bib_ref] A core microbiome associated with the peritoneal tumors of pseudomyxoma peritonei, Gilbreath [/bib_ref] [bib_ref] Antibiotic treatment decreases microbial burden associated with pseudomyxoma peritonei and affects beta-catenin..., Semino-Mora [/bib_ref] and anti-inflammatoryagents with potential values as complementary/adjuvant therapies.
## Prognosis
Although PMP as a neoplastic disease runs a chronic, indolent course with late invasion and only rare metastasis outside the peritoneum, it is a morbid, recurrent condition with life-threatening complications. Biological features of the tumor [bib_ref] Pseudomyxoma peritonei: biological features are the dominant prognostic determinants after complete cytoreduction..., Baratti [/bib_ref] [bib_ref] Proposed classification of pseudomyxoma peritonei: influence of signet ring cells on survival, Shetty [/bib_ref] and access to the current standard of care at specialized oncology centers with a peritoneal surface malignancy program [bib_ref] Pseudomyxoma extraperitonei occurring 35 years after appendicectomy: a case report and review..., Solkar [/bib_ref] [bib_ref] Mucinous appendiceal adenocarcinoma presenting 5 years after appendectomy, Thompson [/bib_ref] [bib_ref] Mucocele of the appendix, Ruiz-Tovar [/bib_ref] [bib_ref] Pseudomyxoma peritonei occurring after an uneventful 23 years interval from appendectomy, Taii [/bib_ref] [bib_ref] The influence of tumor cell entrapment phenomenon on the natural history of..., Spiliotis [/bib_ref] Mucins are a diverse family of high molecular weight, heavily glycosylated proteins [fig_ref] Table 2: Classification, designation and distribution of mucin family [/fig_ref]. Also known as MUC glycoproteins, mucins are differentially expressed by specialized epithelial cells of mucosal surfaces throughout the body in a relatively organ-and cell type-specific manner [bib_ref] Mucin genes and the proteins they encode: structure, diversity, and regulation, Gum [/bib_ref]. Mucins are categorized into membrane-associated and secreted types, with the latter being divided to gel-forming and non-gel-forming subtypes [bib_ref] Respiratory tract mucin genes and mucin glycoproteins in health and disease, Rose [/bib_ref] [bib_ref] Airway mucus: from production to secretion, Williams [/bib_ref]. Membrane-associated mucins communicate information about extracellular conditions, mediate intracellular signal transduction, and contribute to morphological and behavioral characteristics of the epithelial cells [bib_ref] Mucins in cancer: protection and control of the cell surface, Hollingsworth [/bib_ref] [bib_ref] Current status of mucins in the diagnosis and therapy of cancer, Rachagani [/bib_ref].
Secreted mucins provide a physical barrier for epithelial cells lining the respiratory and gastrointestinal tracts and form the ductal surfaces of such organs as liver, breast, pancreas and kidney [bib_ref] Mucins in cancer: function, prognosis and therapy, Kufe [/bib_ref]. Moreover, they are part of a defensive system at the mucosal surfaces, including intestinal mucosa.
## Mucin in intestinal physiology
While facilitating the transit of intestinal contents [bib_ref] Barrier properties of mucus, Cone [/bib_ref] , secreted mucins participate in the front line of the enteric host defense generated by the alliance of the epithelial cells, immune cells and resident microbiota [bib_ref] The front line of enteric host defense against unwelcome intrusion of harmful..., Lievin-Le Moal [/bib_ref]. This interactive ecosystem is essential for the maintenance of intestinal homeostasis and the normal function and activity of digestive system [bib_ref] The gastrointestinal ecosystem: a precarious alliance among epithelium, immunity and microbiota, Mccracken [/bib_ref]. The gastrointestinal epithelium and the overlying mucus layer also function as a barrier against intestinal luminal hazards [bib_ref] Role of intestinal mucins in innate host defense mechanisms against pathogens, Dharmani [/bib_ref]. Colonic mucus is composed of two layers. The outer, loose layer is the habitat of the microbial flora. The inner, dense layer is bacteria-free and firmly attached to the epithelium. This organization keeps the flora well separated from the mucosal surface. The gel-forming mucin MUC2, which is specifically secreted in the small intestine and colon, comprises the substantial component of this double-layered mucus compartment [fig_ref] Figure 2: MUC2 in colonic mucosa [/fig_ref]. MUC2 of the inner layer is uncleaved. To form the outer layer, however, MUC2 undergoes proteolytic cleavage to allow expansion of the polymeric structure [bib_ref] The two mucus layers of colon are organized by the MUC2 mucin,..., Johansson [/bib_ref].
## Mucin in pmp
Under normal conditions, metabolic turnover of intestinal mucin is maintained by the constitutive expression against enzymatic degradation, and, elimination. In PMP, however, mucin is ectopically secreted and increasingly deposited in the peritoneal cavity where it is unable to degrade or drain away. Accumulating mucin causes a major part of the morbidity in PMP. The typical syndrome develops after secreted mucin forms voluminous gels over months and years. Mucin also plays a key role in the biology of the PMP tumor. Most of the tumor cells are surrounded by a mucin coat that allows them to freely move, disseminate and "redistribute" within the peritoneal cavity to create the distinctive feature of PMP. This coating also seems to act as a protective shield against immune recognition and chemotherapy. MUC2, MUC5AC and MUC5B are the gel-forming mucins reportedly found in the PMP secretions. The intestinal mucin MUC2 is known as the PMP-specific mucin. According to O'Connell et al. [bib_ref] Pseudomyxoma peritonei is a disease of MUC2-expressing goblet cells, O'connell [/bib_ref] [bib_ref] MUC2 is a molecular marker for pseudomyxoma peritonei, O'connell [/bib_ref] , primary ovarian mucinous tumors essentially express MUC5AC whereas solitary appendiceal mucinous tumors and different categories of PMP express MUC2 along with MUC5AC. This finding also supports the notion that PMP is a neoplasm of appendiceal origin. In their studies, O'Connell et al. also showed that MUC2 is behind the high degree of gelation formed in PMP. Since MUC2 is more extensively glycosylated, it is more voluminous than MUC5AC on an equimolar basis, hence formation of abundant mucinous collections where the average mucin:cell ratio is higher than 10:1. Taken together, the investigators concluded that PMP is a disease of the MUC2-secreting goblet cells and that MUC2 could serve as a molecular marker for PMP [bib_ref] Pseudomyxoma peritonei is a disease of MUC2-expressing goblet cells, O'connell [/bib_ref] [bib_ref] MUC2 is a molecular marker for pseudomyxoma peritonei, O'connell [/bib_ref]. Since the expression level of MUC2 in PMP is seemingly independent of the degree of the malignant transformation, prognostic significance of MUC2 is controversial [bib_ref] Pseudomyxoma peritonei: biological features are the dominant prognostic determinants after complete cytoreduction..., Baratti [/bib_ref].
In two case studies, Mall et al. reported the presence of MUC5B, in addition to MUC2 and MUC5AC, in the PMP material [bib_ref] MUC2, MUC5AC and MUC5B in the mucus of a patient with pseudomyxoma..., Mall [/bib_ref] [bib_ref] Immunohistochemical and biochemical characterization of mucin in pseudomyxoma peritonei: a case study, Mall [/bib_ref]. Based on the investigations by Sheehan et al. implicating a low-charge glycoform of MUC5B in the production of a tenacious respiratory mucus plug [bib_ref] Analysis of respiratory mucus glycoproteins in asthma: a detailed study from a..., Sheehan [/bib_ref] [bib_ref] Physical characterization of a low-charge glycoform of the MUC5B mucin comprising the..., Sheehan [/bib_ref] , Mall et al. speculated that it may be MUC5B that is responsible for the semisolid material found in some PMP patients. Given the high protein content of the PMP secretions, they also raised the possibility that interactions between mucin and non-mucin proteins could contribute to the viscous nature of the PMP exudates [bib_ref] MUC2, MUC5AC and MUC5B in the mucus of a patient with pseudomyxoma..., Mall [/bib_ref] [bib_ref] Immunohistochemical and biochemical characterization of mucin in pseudomyxoma peritonei: a case study, Mall [/bib_ref]. [fig_ref] Table 3: Some studies exploring the expression of MUC2 against other mucins in PMP [/fig_ref] summarizes a number of studies in which diseasespecific expression pattern of MUC2 and other mucins in PMP has been explored.
## Mucin elimination in the management of pmp
Mucin comprises the cornerstone of the PMP pathogenesis. However, optimal removal of mucin and mucinous implants via conventional therapies remains challenging, with the residual disease accounting for the recurrence of the condition. In order to enhance the current standard of care, novel strategies are required to more effectively eliminate mucin and its source. As such, some efforts have been made to disrupt PMP production of MUC2. In this regard, targeting biological mechanisms regulating the mucin synthesis at transcriptional and post-transcriptional levels could be of potential value. Since MAPK pathway has been implicated in the pathological induction of MUC2 by Pseudomonas aeruginosa infection [bib_ref] Activation of NF-kappaB via a Src-dependent Ras-MAPK-pp90rsk pathway is required for Pseudomonas..., Li [/bib_ref] and mucoepidermoid carcinogenesis [bib_ref] Van Seuningen I: Induction of MUC2 and MUC5AC mucins by factors of..., Perrais [/bib_ref] , targeted inhibition of MAPK might be of therapeutic benefit in PMP. On the basis of the cross-talk between MAPK and other inflammation-associated signaling pathways [bib_ref] NF-kappaB as a therapeutic target in cancer, Orlowski [/bib_ref] , multi-targeted agents may be more effective in PMP where the disease develops within an inflammatory milieu [bib_ref] Mucin as a therapeutic target in pseudomyxoma peritonei, Choudry [/bib_ref].
In addition, PMP inflammatory microenvironment with a unique profile of cytokines [bib_ref] Pseudomyxoma Peritonei: Inflammatory Responses in the Peritoneal Microenvironment, Lohani [/bib_ref] contributes to the mucin overproduction. Cytokines reportedly upregulate the expression of MUC2 and enhance the mucin secretion [bib_ref] Proinflammatory cytokines trigger MUC gene expression and mucin release in the intestinal..., Enss [/bib_ref] [bib_ref] Regulation of IL-1beta-mediated MUC2 gene in NCI-H292 human airway epithelial cells, Kim [/bib_ref] [bib_ref] mRNA of MUC2 is stimulated by IL-4, IL-13 or TNF-alpha through a..., Iwashita [/bib_ref]. Glucocorticoids, on the other hand, have been shown to downregulate MUC2 via direct inhibition of glucocorticoid response elements (GREs) and indirect transrepression of inflammation-associated transcription factors [bib_ref] Dexamethasone suppresses mucus production and MUC-2 and MUC-5 AC gene expression by..., Kai [/bib_ref] [bib_ref] Mechanisms of glucocorticoid signalling, Schoneveld [/bib_ref]. Choudry et al. recently demonstrated dexamethasone-and celecoxib-induced inhibition of the mucin production in a mucin-secreting cancer cell line with goblet cell phenotype as well as in a murine model of PMP. Methylation of the MUC2 promoter has also been suggested as a potential tool for manipulating the expression of MUC2. Okudaira et al. showed that colorectal cancer (CRC) cell lines of mucinous type exhibit low-level methylation at the MUC2 promoter as compared to nonmucinous cell lines. They concluded that low methylation status of the MUC2 gene plays a predominant role in the high level MUC2 expression in mucinous CRC [bib_ref] MUC2 gene promoter methylation in mucinous and non-mucinous colorectal cancer tissues, Okudaira [/bib_ref]. Through inhibiting the MUC2 promoter methylation, the Sp-family of transcription factors augments MUC2 expression. Thus, mithramycin, a Sp1 binding site inhibitor, effectively blocks the MUC2 expression in colorectal cancer [bib_ref] Van Seuningen I: Induction of MUC2 and MUC5AC mucins by factors of..., Perrais [/bib_ref] [bib_ref] The Sp family of transcription factors in the regulation of the human..., Aslam [/bib_ref].
Breakdown of mucin as well as enhancement of cytoreduction through locoregional therapies outlines our intended approach to this challenge. In this regard, we aim to develop a novel treatment for facilitated, enhanced removal of the mucin-tumor burden at both macroscopic and microscopic levels. For this purpose, a panel of mucin-secreting cancer cells of gastrointestinal and peritoneal origin and animal models of PMP [bib_ref] In vivo model of pseudomyxoma peritonei for novel candidate drug discovery, Chua [/bib_ref] and peritoneal carcinomatosis [bib_ref] A new peritoneal carcinomatosis model in cyclosporine immunosuppressed rats, Akhter [/bib_ref] are employed in our preclinical investigations. Among a number of mucolytic agents studied [bib_ref] Mucolysis by ascorbic acid and hydrogen peroxide on compact mucin secreted in..., Pillai [/bib_ref] [bib_ref] Potential mucolytic agents for mucinous ascites from pseudomyxoma peritonei, Pillai [/bib_ref] , bromelain and N-acetyl cysteine (NAC), two generally-safe mucolytics of plant sources, have shown promise. Our preliminary results indicate bromelain- [bib_ref] Cytotoxic effects of bromelain in human gastrointestinal carcinoma cell lines (MKN45, KATO-III,..., Amini [/bib_ref] [bib_ref] Anticancer property of bromelain with therapeutic potential in malignant peritoneal mesothelioma, Pillai [/bib_ref] [bib_ref] Anticancer effect of bromelain with and without cisplatin or 5-FU on malignant..., Pillai [/bib_ref] and NAC-induced inhibition of the growth and proliferation of the mucinproducing cancer cells in vitro, with the cytotoxicity being augmented when the two are used in combination (unpublished data). We have also observed the ability of these mucolytics in dissolving both the mucin samples from PMP patients ex vivo and the mucin implants in a xenograft model of PMP in vivo with no treatmentrelated toxicity in rats [bib_ref] A formulation for in situ lysis of mucin secreted in pseudomyxoma peritonei, Pillai [/bib_ref]. We are currently performing final preclinical testing of our mucolytic compound before proceeding to the clinical phase.
# Conclusion
Despite its morbid, debilitating nature with severe impact on quality of life, PMP remains orphan and enigmatic. Essential for physiological function of the gastrointestinal tract, mucin is the major contributor to the pathophysiology of PMP. Peritoneal implantation of the tumor cells originating from an appendiceal mucinous tumor results in the progressive accumulation of ectopic mucin. Given the clinicopathological profile of the disease compatible with an indolent, generally low grade malignancy, multifocal collections of voluminous mucin and the ensuing complications are the main determinants of the disease prognosis. As the predominant, gel-forming mucin secreted in PMP, MUC2 is responsible for the high degree of gelation and the characteristic feature of the clinical syndrome. Despite the current standard of care as extensive surgical resection combined with chemotherapy, PMP frequently recurs; with treatmen options being limited at recurrence. On this basis, in-depth investigations are warranted to illuminate unknown aspects of the disease and to seek novel therapeutic approaches for an enhanced treatment. Owing to the substantial role of mucin in the pathogenesis of PMP, development of strategies for targeting mucin and its biology seems to be of particular significance that needs to be further explored in future studies.
[fig] Figure 1: Schematic representation of the events resulting in the development of PMP. The pathologic process starts with a neoplastic transformation of the appendiceal goblet cells and development of a primary mucinous tumor (1). Overproduction of mucin and obstruction of the appendiceal lumen lead to the development, and subsequent rupture, of a mucocele (2). Shredded tumor cells gain access to the peritoneal cavity and circulate with the peritoneal fluid (3). Accordingly, tumor cells redistribute and accumulate within the dependent portions of the peritoneal cavity (3*, downward arrows) as well as at the peritoneal fluid reabsorption sites (3**, upward arrows). [/fig]
[fig] Figure 2: MUC2 in colonic mucosa. A Synthesis, secretion and organization of colonic mucosa. The first stage in the biosynthesis of MUC2 is the formation of MUC2 monomer as an N-glycosylated apoprotein in the endoplasmic reticulum. Subsequently, MUC2 dimers are formed when intermolecular disulfide bonds bridge between the C-terminal cysteine knot domains. During transit through the Golgi apparatus, MUC2 dimers become heavily O-glycosylated. Complete glycosylation of the dimers occurs in Golgi where trimerization through disulfide bonds at the N-terminus forms protease-resistant trimers. The fully glycosylated and processed MUC2 is densely packed and stored in secretory granules/ vesicles and released through constitutive or stimulated secretory mechanisms. Once released, MUC2 is organized into the firmly adherent inner layer. At a certain distance from the epithelium, this layer is converted into the loose outer layer through proteolytic cleavage and expansion. Mucus also contains immunoglobulins and other proteins. B MUC2 structure. The protein core consists of five different regions. Segment (a) and (b) are two central repetitive regions rich in potential O-glycosylation sites, to which branched carbohydrate chains of 4-12 sugars are O-glycosidically linked to form a closely packed sheath around the central protein core. Segment (a), also known as VNTR domain, is a large domain that contains 50-100 "variable number of tandem repeats (VNTRs)" of 23 amino acids, in particular, threonine. Segment (b), also called PTS domain, is a 347 amino acid domain, containing irregular repeats rich in proline, threonine and serine (PTS). These two segments are linked together by segment (c) which is a 148 amino acid, cysteine containing region. Segments (d) and (e) are extensive peptide chains rich in cysteine located at the C and N terminal ends, respectively, containing D domains with sequence homology to von Willebrand factor. These regions are the presumed sites for end to end polymerization of mucin subunits. [/fig]
[table] Table 1: Common presentations or incidental discovery of PMP on the basis of the disease progression [/table]
[table] Table 2: Classification, designation and distribution of mucin family [/table]
[table] Table 3: Some studies exploring the expression of MUC2 against other mucins in PMP (2002-2012) *Full-text article, in Czech, not accessible. Results of IHC study not available. **This study reports the expression of MUC2 and MUC5A as the volumetric density of apomucin (Vvi/10 4 μm) in such compartments of DPAM and PMCA tissues as epithelium, lymphoid aggregates, stroma vessels and free mucin, respectively, as follows: ‡MUC2: in DPAM: 264 ± 60, 47 ± 16, 31 ± 14 and 261 ± 51; in PMCA: 356 ± 90, 170 ± 26, 117 ± 25 and 1043 ± 282. ‡ ‡MUC5AC: in DPAM: 90 ± 13, 345 ± 20, 65 ± 17, 37 ± 6; in PMCA: 56 ± 12, 246 ± 17, 50 ± 15 and 48 ± 9. ***The percentages shown for this study are numerical estimations of data originally presented in a column graph. ****In this study, among a total of 14 patients with mucinous adenocarcinoma, 4 cases have reportedly exhibited PMP syndrome. Results of the expression of MUC2 and MUC5AC, however, are reported in total, with no data individually available regarding the PMP cases. †, † †Data shown is the percentage of MUC2/MUC5AC expression in all patients with mucinous adenocarcinoma, including PMP ones. [/table]
|
Hepatitis B Virus X Protein Inhibits the Expression of Barrier To Autointegration factor1 via Upregulating miR-203 Expression in Hepatic Cells
Hepatitis B virus (HBV) infection targets host restriction factors that inhibit its replication and survival. Previous studies have shown that barriers to autointegration factor1 (BANF1) inhibited the replication of herpes simplex virus and vaccinia virus by binding to phosphate backbone of dsDNA. To date, no reports are available for the interplay between BANF1 and HBV. In this study, we elucidated the mechanisms by which HBV inhibit BANF1. First, the effect of HBV on BANF1 was observed in Huh-7, Hep G2, and Hep G2.2.15 cells. Huh-7 cells were transfected with pHBV 1.3 or HBx plasmids. The results showed that there was a decreased expression of BANF1 in Hep G2.2.15 cells (P # 0.005) or in HBV/HBx expressing Huh-7 cells (P # 0.005), whereas BANF1 overexpression decreased viral replication (P # 0.05). To study whether phosphorylation/dephosphorylation of BANF1 was responsible for antiviral activity, mutants were created, and it was found that inhibition due to mutants was less significant compared to BANF1 wild type. Previous studies have shown that HBV, at least in part, could regulate the expression of host miRNAs via HBx. It was found that miR-203 expression was high in Hep G2.2.15 cells (P # 0.005) compared to Hep G2 cells. Next, the effect of HBx on miR-203 expression was studied and result showed that HBx upregulated miR-203 expression (P # 0.005). Overexpression of miR-203 downregulated BANF1 expression (P # 0.05) and viral titer was upregulated (P # 0.05), while inhibition of miR-203, reversed these changes. In conclusion, BANF1 downregulated HBV, whereas HBV inhibited BANF1, at least in part, via HBx-mediated miR-203 upregulation in hepatic cells.IMPORTANCE In this study, for the first time, we found that BANF1 inhibited HBV replication and restricted the viral load. However, as previously reported for other viruses, the results in this study showed that BAF1 phosphorylation/dephosphorylation is not involved in its antiviral activity against HBV. HBV infection inhibited the intracellular expression of BANF1, via HBx-mediated upregulation of miR-203 expression. Overexpression of miR-203 downregulated BANF1 and increased the viral titer, while inhibition of miR-203 reversed these changes. This study helped us to understand the molecular mechanisms by which HBV survives and replicates in the host cells.
HBV replication [bib_ref] Hepatitis B virus protein X induces degradation of talin-1, Van De Klundert [/bib_ref]. In both the studies, HBV encoded HBx protein was reported to be requisite to rescue HBV inhibition. HBV encodes HBx protein, surface antigen HBsAg, HBV core protein HBcAg and HBV polymerase (Pol) which is basically reverse transcriptase. Several studies have shown that HBx is indispensable for HBV replication. HBx acts directly on viral promoters and enhancers to promote viral replication. HBx indirectly also contributes to viral replication by modulating cellular pathways to benefit HBV infection [bib_ref] The hepatitis B virus (HBV) HBx protein activates AKT to simultaneously regulate..., Rawat [/bib_ref]. Apart from host proteins, various miRNAs have been shown to control HBV replication, including the epigenetic modulation of HBV genome. Recent studies show that HBV itself encodes a miRNA (HBV-miR-3) which restricts HBV replication by activating the innate immune system [bib_ref] Hepatitis B virus-encoded MicroRNA controls viral replication, Yang [/bib_ref]. HBV accessory protein HBx also modulates the expression of various miRNAs that are involved in innate and adaptive immune pathways, and thus, helps in HBV replication and survival [bib_ref] The multiple roles of hepatitis B virus X protein (HBx) dysregulated microRNA..., Sartorius [/bib_ref]. BANF1 is a small (89 residues; 10 kD) protein present in the nucleus of cells and is capable of forming stable homodimers that can bind two dsDNA molecules [bib_ref] Barrierto-autointegration factor (BAF) bridges DNA in a discrete, higher-order nucleoprotein complex, Zheng [/bib_ref]. Each monomer of BANF1 consists of two copies of the helix-hairpin-helix (HhH) motif that can bind dsDNA in a sequence independent manner [bib_ref] Structural basis for DNA bridging by barrier-to-autointegration factor, Bradley [/bib_ref]. BANF1 binds the phosphate backbone of dsDNA, and thus, lacks or does not require DNA sequence specificity [bib_ref] Barrierto-autointegration factor (BAF) bridges DNA in a discrete, higher-order nucleoprotein complex, Zheng [/bib_ref] [bib_ref] Structural basis for DNA bridging by barrier-to-autointegration factor, Bradley [/bib_ref]. Originally BANF1 was identified in mammalian cells by virtue of its association with the preintegration complexes (PICs) of retroviruses [bib_ref] A previously unidentified host protein protects retroviral DNA from autointegration, Lee [/bib_ref]. BANF1 acts as a potent antiviral effector against the poxvirus vaccinia virus in the cytoplasm due to its DNA-binding/compaction activity. Poxviruses express a Ser/Thr protein kinase named B1 kinase which phosphorylates BANF-1 at its N-terminal; this in turn inactivates its DNA binding capability [bib_ref] Poxviral B1 kinase overcomes barrier to autointegration factor, a host defense against..., Wiebe [/bib_ref]. BANF1 is a threat not only to cytoplasmic poxviruses but for nuclear DNA viruses as well. BANF1 has been reported to interfere with herpes simplex virus 1 (HSV-1) viral DNA replication and gene expression leading to viral inhibition [bib_ref] Barrier to auto integration factor becomes dephosphorylated during HSV-1 infection and can..., Jamin [/bib_ref]. Over the past decade, many efforts have been made into identification of HBV interacting host cell proteins. To date, several host dependency factors also known as proviral factors have been identified that are beneficial for HBV and act positively to support HBV replication. This puts host cells under tremendous selective pressure to prevent HBV infection, thus leading toward evolution of various restriction factors that act negatively to inhibit or block HBV. Because BANF1 can crossbridge DNA and induces DNA condensation by cross-linking different regions of DNA, we hypothesized that BANF1 might restrict the HBV, and hence, HBV might inhibit the expression of BANF1.
# Results
HBV infection downregulates BANF1 expression. Effects of HBV infection on BANF1 expression was analyzed by real-time PCR (RT-PCR) in Huh-7 cells Hep G2 cells and HBV expressing Hep G2.2.15 cells. The results showed that BANF1 expression was decreased by 2-fold in HBV expressing Hep G2.2.15 cells in comparison to Huh-7 cells and Hep G2 cells [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref]. Further, Huh-7 cells were transfected with HBV plasmid for 48 h and total RNA was isolated. RT-PCR results revealed that there was a significant increase in HBx expression [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref] , while BANF1 expression was decreased by 60 percent in HBV-transfected Huh-7 cells [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref]. HBx is requisite to initiate and maintain HBV replication. To study whether HBx played a role in the BANF1 expression, Huh-7 cells were transfected with HBx plasmid. After 48 h of transfection, the cells were harvested for RNA and total protein isolation. RT-PCR results showed significant increase in HBx expression in the HBx-transfected cells [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref]. BANF1 expression was significantly downregulated in HBx-transfected Huh-7 cells [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref]. Western blot results showed that there was significant decrease in BANF1 expression in HBx transfected Huh-7 cells [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref]. The BANF1 band intensity quantification showed that BANF1 was downregulated by 65% in HBx-transfected cells [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref]. To further validate these results, Huh7 cells were transfected with HBsAg, HBcAg, and HBV polymerase (three additional proteins encoded by HBV apart from HBx) expressing plasmids, and performed RT-PCR to evaluate the change in expression of BANF1. Results showed that HBsAg, HBcAg and HBV-pol were upregulated [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref] but there was no change in expression of BANF1 in HBsAg, HBcAg, and HBV pol transfected cells [fig_ref] FIG 1: HBx downregulates BANF1 expression in hepatoma cells [/fig_ref]. Hence, it was concluded that the decrease in BANF1 expression was due to HBx, but not due to HBsAg, HBcAg and HBV pol. These results suggest that HBV infection downregulated BANF1 expression via HBx protein.
BANF1 overexpression inhibits HBx, HBsAg, and the viral load. In order to investigate the effect of BANF1 on HBV life cycle, BANF1 gene was amplified from Huh-7 cells, cloned into pcDNA3.1(-) vector and confirmed by sequencing. BANF1 plasmid was transfected either alone or cotransfected with HBV 1.3 plasmid (pHBV 1.3 ) in Huh-7 cells. Western blot results showed that BANF1 overexpression downregulated HBx expression significantly [fig_ref] FIG 2: BANF1 overexpression inhibits expression of HBx, HBsAg and HBV replication [/fig_ref]. BANF1 and HBx band intensity quantification showed that HBx expression was downregulated in BANF1-overexpressing cells [fig_ref] FIG 2: BANF1 overexpression inhibits expression of HBx, HBsAg and HBV replication [/fig_ref]. Furthermore, Hep G2.2.15 cells were transfected with BANF1 plasmid and cultured for 48 h. Then, the spent media and the cells were collected for HBsAg analysis by ELISA and HBV DNA analysis by real-time PCR, respectively. HBsAg titer was significantly low in BANF1-transfected cells compared to the control or empty vector transfected cells [fig_ref] FIG 2: BANF1 overexpression inhibits expression of HBx, HBsAg and HBV replication [/fig_ref]. Similarly, the total HBV load was decreased in BANF1-overexpressing Hep G2.2.15 cells [fig_ref] FIG 2: BANF1 overexpression inhibits expression of HBx, HBsAg and HBV replication [/fig_ref]. These results suggest that BANF1 possess antiviral activity against HBV.
The expression of HBV-encoded proteins and total HBV viral load is downregulated significantly by BANF1-wild type compared to unphosphorylatable BANF1 mutants. Previously it was shown that when thr-2, thr-3, and ser-4 residues were replaced with alanine, BANF1 was no longer phosphorylated, and hence, BANF1 could bind with dsDNA strongly and further decrease the total herpes simplex viral load [bib_ref] Barrier to auto integration factor becomes dephosphorylated during HSV-1 infection and can..., Jamin [/bib_ref] [bib_ref] The vaccinia-related kinases phosphorylate the N' terminus of BAF, regulating its interaction..., Nichols [/bib_ref]. Literature suggests that wild type (WT)-BANF1 can be phosphorylated and upon phosphorylation, BANF1 loses its ability to interact with DNA. Also, phosphorylation/dephosphorylation of BANF1 regulates the subcellular localization of BANF1. Unphosphorylatable BANF1 mutants are primarily expressed in the nucleus, while phosphorylated form of BANF1 translocates to cytoplasm [bib_ref] Barrier to auto integration factor becomes dephosphorylated during HSV-1 infection and can..., Jamin [/bib_ref]. Hence, three BANF1 mutants, namely, Mutant1 (M1) (TTS-ATS), Mutant2 (M2) (TTS-AAS), and Mutant3 (M3) (TTS-AAA) were created by site directed mutagenesis (SDM). WT-BANF1 contains phosphorylation site, and upon phosphorylation loses its ability to bind dsDNA and translocates to cytoplasm, while unphosphorylated BANF1 mutants were hypothesized to remain bound with HBV cccDNA for longer duration and thus inhibit replication and transcription.
BANF1 WT plasmid and BANF1 mutants (M1, M2, M3) were cotransfected with pHBV 1.3 in Huh-7 cells. Western blot analysis revealed that in BANF1 WT and pHBV 1.3 -cotransfected cells there was a 6.4-fold decrease in the expression of HBx [fig_ref] FIG 3: Mutation at the N-terminus of BANF1 modulates the expression of HBx, HBsAg... [/fig_ref]. In the cells 15 cells were cultured, and the total RNA was isolated followed by RT-PCR for BANF1 expression (n = 3; **, P , 0.01) (A). Huh-7 cells were transfected with pHBV 1.3 plasmid and the cells were collected for the isolation of RNA, followed by RT-PCR for HBx (B) or BANF1 (C) expression (n = 3; ***, P , 0.001; **, P , 0.01). Huh-7 cells were transfected with HBsAg, HBcAg and HBV-pol plasmid, total RNA was isolated and RT-PCR was performed for HBsAg, HBcAg and HBV-pol (n = 3; *, P , 0.05; **, P , 0.01) (H) and BANF1 (I). After 48 h, both the RNA and protein were isolated and RT-PCR or Western blots were performed. The expression of HBx (D) or BANF1 (E) was determined (n = 3; ***, P , 0.001; **, P , 0.01). The total cellular protein was run on SDS-PAGE gels and Western blots were performed and a representative picture is shown. Lane 1, control; lane 2, Empty vector transfected cells; and lane 3, HBx-transfected cells (F). The band intensities were quantified using Image J software and presented as fold change (BANF1/b-actin) n = 3; **, P , 0.01 (G).
## Hbx-induced mir-203 inhibits banf1
Microbiology Spectrum cotransfected with pHBV 1.3 and M1 or M2 or M3, there was a 1.5-fold, 2.5-fold, and 2.5-fold decrease in the expression of HBx, respectively [fig_ref] FIG 3: Mutation at the N-terminus of BANF1 modulates the expression of HBx, HBsAg... [/fig_ref] and B). HBx and BANF1 band intensity quantification further confirmed that HBx expression was downregulated more significantly in BANF1-WT and HBx cotransfected samples compared to BANF1 mutants (M1, M2, M3) and HBx cotransfected samples [fig_ref] FIG 3: Mutation at the N-terminus of BANF1 modulates the expression of HBx, HBsAg... [/fig_ref]. To further validate these results, HBV expressing Hep G2.2.15 cells were transfected with BANF1 WT, M1, M2, or M3 plasmid. After 48 h, spent media was collected for HBsAg analysis by ELISA and cells were collected to determine the total viral load by RT-PCR. Result showed that decrease in HBsAg levels due to BANF1 WT and BANF1 mutants were similar [fig_ref] FIG 3: Mutation at the N-terminus of BANF1 modulates the expression of HBx, HBsAg... [/fig_ref]. Decrease in total HBV viral load was also similar in all the transfected cells (WT and mutants) and BANF1 mutants did not modulate the HBV expression further as hypothesized [fig_ref] FIG 3: Mutation at the N-terminus of BANF1 modulates the expression of HBx, HBsAg... [/fig_ref]. These results indicate that BANF1-mediated HBV inhibition does not involve phosphorylation/dephosphorylation of BANF1 because unphosphorylatable BANF1 mutants, which are predominantly present in nucleus and also remain bound to DNA for a longer period, could not induce more intense inhibition of HBV compared to WT-BANF1. HBx upregulates miR-203 expression in hepatic cells. HBV infection deregulates various miRNAs, and these miRNAs in turn help in successful HBV infection establishment [bib_ref] Dysregulated microRNAs in hepatitis B virusrelated hepatocellular carcinoma: potential as biomarkers and..., Xu [/bib_ref]. To test whether miRNAs target BANF1 expression, miRNA databases (TargetScan and miRBase) were searched and it was found that both miR-203 and miR-150 are possible regulators [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref]. Total RNA was isolated from Huh-7, Hep G2 cells, and Hep G2.2.15 cells and RT-PCR was performed to compare the expression of miR-150 and miR-203. There was no significant change in the expression of miR-150 [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref]. MiR-203 expression was upregulated by 2-fold in Hep G2.2.15 cells compared to Huh-7 cells and Hep G2 cells [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref]. Next, Huh-7 cells were transfected with pHBV 1.3 and total RNA was isolated and RT-PCR was performed for HBx and miR-203 expression. There was a significant expression of HBx [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref] and miR-203 expression [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref] in HBV transfected cells. To study whether the increased expression of miR-203 is due to HBx protein, Huh-7 cells were transfected with HBx plasmid and the total RNA was isolated. The RT-PCR results showed that there was a significant increase in the expression of HBx [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref] and miR-203 [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref] in HBx-transfected cells. To further validate these results, Huh7 cells were transfected with HBsAg, HBcAg and HBV polymerase (three additional proteins encoded by HBV apart from HBx) expressing plasmids and performed RT-PCR to evaluate the change in expression of miR-203. Results showed that HBsAg, HBcAg, and HBV-pol were upregulated [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref] but there was no change in expression of miR-203 in HBsAg, HBcAg, and HBV pol transfected cells [fig_ref] FIG 4: HBx upregulates miR-203 expression in hepatic cells [/fig_ref]. These results suggest that HBV accessory protein HBx upregulates the expression of miR-203 in HBV infected cells, at least in part.
miR-203 targets the BANF1 expression and increases the HBV viral titer. Hep G2.2.15 cells were transfected with miR-203 pre-miR oligos and the expression of miR-203 and BANF1 were determined. The RT-PCR results showed that there was a significant increase in the miR-203 levels [fig_ref] FIG 5: miR-203 downregulates BANF1 expression in HBV expressing cells [/fig_ref] and the BANF1 expression was significantly downregulated [fig_ref] FIG 5: miR-203 downregulates BANF1 expression in HBV expressing cells [/fig_ref] in miR-203 overexpressing cells. Next, BANF1 protein expression was determined using Western blots in these cells. The results showed that there was a significant decrease of BANF1 in miR-203 transfected Hep G2.2.15 cells [fig_ref] FIG 5: miR-203 downregulates BANF1 expression in HBV expressing cells [/fig_ref]. To confirm these findings, anti-miR-203 was transfected in Hep G2.2.15 cells and the expression of miR-203 and BANF1 were determined using RT-PCR. The results showed that there was a 40% decrease in the expression of miR-203 [fig_ref] FIG 5: miR-203 downregulates BANF1 expression in HBV expressing cells [/fig_ref] , while the expression of BANF1 was upregulated [fig_ref] FIG 5: miR-203 downregulates BANF1 expression in HBV expressing cells [/fig_ref] in anti-miR-203 transfected cells. Furthermore, Hep G2.2.15 cells were transfected with miR-203, and anti-miR-203, and after 48 h, spent media was collected for HBsAg analysis by ELISA. HBV viral titer was determined by estimating HBsAg level. ELISA results showed that HBsAg expression was upregulated in miR-203 transfected Hep G2.2.15 cells significantly [fig_ref] FIG 5: miR-203 downregulates BANF1 expression in HBV expressing cells [/fig_ref] while HBsAg levels were significantly declined in ant-miR-203 transfected Hep G2.2.15 cells [fig_ref] FIG 5: miR-203 downregulates BANF1 expression in HBV expressing cells [/fig_ref]. These results thus established the link between miR-203, BANF1, and HBV interplay and suggested that HBx mediated miR-203 upregulation in HBV infected cells downregulates BANF1 expression and increases HBV titer.
# Discussion
Chronic HBV infection is one of the leading causes for hepatocellular carcinoma. Once HBV infect the hepatocytes, it regulates the expression of various host proteins and signaling pathways to its favor for survival and replication. Previously, it was shown that Farnesoid X receptor-a and Polo-like-kinase 1 act as proviral factors during HBV infection [bib_ref] Polo-like-kinase 1 is a proviral host factor for hepatitis B virus replication, Diab [/bib_ref] [bib_ref] Farnesoid X receptor-a is a proviral host factor for hepatitis B virus..., Mouzannar [/bib_ref] , while Talin-1 and SMC5/6 were reported to act as host restriction factors [bib_ref] Hepatitis B virus protein X induces degradation of talin-1, Van De Klundert [/bib_ref]. Identification of such HBV interacting host proteins may provide a deeper insight into HBV life cycle, which might be used for future targeted therapy. Hence, in this study, we determined the role of BANF1 in HBV replication and the mechanisms by which HBV inhibit BANF1 expression.
The DNA binding protein BANF1 is highly conserved and there are six BANF1 homologous genes in mammals [bib_ref] Barrier to autointegration factor interacts with the cone-rod homeobox and represses its..., Wang [/bib_ref]. BANF1 exits in cytoplasm as well as nucleus of the cell. These nuclear and cytoplasmic pools are dynamic in nature, but studies showed that they cannot replenish each other [bib_ref] Dynamic interaction between BAF and emerin revealed by FRAP, FLIP, and FRET..., Shimi [/bib_ref]. BANF1 interacts with LEM domain proteins and plays an important role in the nuclear envelope assembly during the late phase of mitosis [bib_ref] Barrier-to-autointegration factor is required to segregate and enclose chromosomes within the nuclear..., Margalit [/bib_ref] [bib_ref] Caenorhabditis elegans BAF-1 and its kinase VRK-1 participate directly in post-mitotic nuclear..., Gorjánácz [/bib_ref]. Apart from its role in mitosis, BANF1 is involved in other cellular processes like gene expression [bib_ref] Barrier to autointegration factor interacts with the cone-rod homeobox and represses its..., Wang [/bib_ref] , developmental process (23), cell cycle progression [bib_ref] Barrier-to-autointegration factor is required to segregate and enclose chromosomes within the nuclear..., Margalit [/bib_ref] , and DNA damage response in the nucleus [bib_ref] Barrierto-autointegration factor proteome reveals chromatin-regulatory partners, Montes De Oca [/bib_ref]. BANF1 can sense and bind to foreign DNAs in the cytoplasm. By virtue of this, BANF1 associates with retroviral DNA preintegration complexes and protects it from undergoing suicidal autointegration [bib_ref] Solution structure of the cellular factor BAF responsible for protecting retroviral DNA..., Cai [/bib_ref]. The ability of BANF1 to interact with dsDNA in a sequence independent manner allows it to act as host defense against HSV1 and vaccinia virus [bib_ref] Barrier to auto integration factor becomes dephosphorylated during HSV-1 infection and can..., Jamin [/bib_ref] [bib_ref] Vaccinia virus B1 kinase: phenotypic analysis of temperature-sensitive mutants and enzymatic characterization..., Rempel [/bib_ref]. To date, there is no data available to show the relationship between BANF1 and HBV replication, and the mechanism by which HBV inhibits BANF1 expression in hepatic cells.
First, the effect of HBV on BANF1 expression was studied. It was found that BANF1 expression was significantly low in Hep G2.2.15 cells (HBV stably expressing Hep G2 cells) compared to Hep G2 cells. The expression of BANF1 was also tested in another hepatic cell line, Huh-7 cells, which showed the BANF1 protein levels were similar to Hep G2 cells, and not significantly different. These results showed that different biological features of cells probably do not affect the BANF1 expression, rather the change is due to HBV infection. To confirm these findings, Huh-7 cells were transfected with HBV plasmid (pHBV 1.3 ), HBx plasmid, HBsAg plasmid, HBcAg plasmid or HBV-pol expressing plasmid and found that HBV inhibited the intracellular expression of BANF1 via HBx, at least in part. These results showed that the HBV infection inhibited BANF1 expression irrespective of cell lines. Next, the role of BANF1 in modulating HBV life cycle was established. The overexpression of BANF1 in Hep G2.2.15 cells resulted in decreased expression of HBsAg and HBV DNA (viral load). When Huh-7 cells were cotransfected with BANF1 and HBV plasmid, the HBx protein levels were decreased. These findings correlated well with the previous findings where the authors showed that BANF1 inhibited both HSV1 and vaccinia virus in L929 cells [bib_ref] Barrier to auto integration factor becomes dephosphorylated during HSV-1 infection and can..., Jamin [/bib_ref] [bib_ref] Vaccinia virus B1 kinase: phenotypic analysis of temperature-sensitive mutants and enzymatic characterization..., Rempel [/bib_ref].
Poxviruses express a Ser/Thr protein kinase B1 which phosphorylates the N-terminal serine threonine residues of BANF1 and renders it incapable of binding with poxviral DNA [bib_ref] Poxviral B1 kinase overcomes barrier to autointegration factor, a host defense against..., Wiebe [/bib_ref]. In the case of HSV-1, it was found that BANF1 mutants that cannot be phosphorylated, impaired viral DNA replication [bib_ref] Barrier to auto integration factor becomes dephosphorylated during HSV-1 infection and can..., Jamin [/bib_ref]. BANF1 showed a well-documented antiviral activity against cytoplasmic and nuclear viruses, but still its role in HBV life cycle is not known. Because HBV belongs to hepadnaviridae family and replicates its DNA genome in the nucleus, the phosphorylation/dephosphorylation of BANF1 was hypothesized to be responsible for its anti-HBV activity. WT-BANF1 (Met-Thr-Thr-Ser) contains ser/thr amino acids at its N-terminus which can be phosphorylated either by cellular vaccinia related kinase or by B1 kinase in case of Pox virus infection [bib_ref] Poxviral B1 kinase overcomes barrier to autointegration factor, a host defense against..., Wiebe [/bib_ref] [bib_ref] The vaccinia-related kinases phosphorylate the N' terminus of BAF, regulating its interaction..., Nichols [/bib_ref]. BANF1 phosphorylation renders it incapable of interacting with DNA and it can no longer induce cross bridging and condensation [bib_ref] Barrier to auto integration factor becomes dephosphorylated during HSV-1 infection and can..., Jamin [/bib_ref]. Previous reports suggest that phosphorylation/dephosphorylation of BANF1 regulates the subcellular localization of BANF1 [bib_ref] Barrier to auto integration factor becomes dephosphorylated during HSV-1 infection and can..., Jamin [/bib_ref]. Because WT-BANF1 contains phosphorylation site and upon phosphorylation loses its ability to bind dsDNA and translocates to cytoplasm, it was hypothesized that mutants will be retained in nucleus only and will remain bound to HBV DNA genome for longer period of time, thus leading to more inhibition compared to WT-BANF1. To determine whether mutation resulted in enhanced inhibition of HBV levels, three different BANF1 mutants (M1, M2, M3), M1 (TTS-ATS), M2 (TTS-AAS), and M3 (TTS-AAA) were created by SDM and sequenced. The sequence results confirmed the changes in the correct positions. Overexpression of these mutants showed that these unphosphorylatabale BANF1 mutants still inhibited the replication, thereby decreasing total viral load, HBx and HBsAg, but as hypothesized, the inhibition induced by BANF1 mutants was not more than WT BANF1. These results led us to conclude that BANF1-mediated HBV inhibition does not involve cross bridging and condensation of HBV genome as reported for HSV-1, and inhibitory mechanism still remains elusive.
Several studies have shown that HBx is indispensable for HBV replication. HBx acts directly on viral promoters and enhancers to promote viral replication. HBx indirectly contributes to viral replication by modulating cellular pathways to benefit HBV infection [bib_ref] The hepatitis B virus (HBV) HBx protein activates AKT to simultaneously regulate..., Rawat [/bib_ref]. Previously, we also showed that HBx induced proliferation via miR-21 in hepatic cells [bib_ref] Hepatitis B virus induces cell proliferation via HBx-induced micro-RNA-21 in hepatocellular carcinoma..., Damania [/bib_ref]. Thus, HBx is indispensable for successful HBV infection and promotes replication from extrachromosomal templates [bib_ref] Hepatitis B virus X protein stimulates gene expression selectively from extrachromosomal DNA..., Van Breugel [/bib_ref]. Though other proteins of HBV might play a role in modulating BANF1 expression, in this manuscript the role of HBx in HBV pathogenesis was studied.
HBV infection dysregulates expression of various miRNAs which in turn help in HBV infection establishment [bib_ref] Dysregulated microRNAs in hepatitis B virusrelated hepatocellular carcinoma: potential as biomarkers and..., Xu [/bib_ref]. To study if HBx modulates BANF1 expression via miRNA expression, miRNA databases (TargetScan and miRBase) were searched and based on the literature both miR-203 and miR-150 were selected, because they might target the 39-UTR of BANF1 expression. It was found that expression of miR-203 was significantly upregulated in Hep G2.2.15 cells, while there was no change in expression of miR-150. Further the expression of HBx induced the expression of miR-203 in HBV infected cells; however, there was no change observed in expression of miR-203 in HBsAg, HBcAg, and HBV-pol transfected cells. Previously, it was shown that HBx-mediated miR-203 is involved in inducing inflammation [bib_ref] miR-203a is involved in HBx-induced inflammation by targeting Rap1a, Wu [/bib_ref]. HBV infection leads to hepatocellular carcinoma progression via HBV-induced immune imbalance [bib_ref] HBV-induced immune imbalance in the development of HCC, Chen [/bib_ref]. Previous studies have shown that the overexpression of miR-203 reduced HCC prevalence, and could inversely affect the HCC outcome [bib_ref] MicroRNA-203 impacts on the growth, aggressiveness and prognosis of hepatocellular carcinoma by..., Simile [/bib_ref]. Based on the available data with the progression of HCC and the HBV-induced HCC, it might be possible that different mechanisms exist. Inhibition of miR-203 using anti-miR-203 upregulated BANF1 expression. To establish the link between miR-203, BANF1, and HBV interplay, Hep G2.2.15 cells were transfected with miR-203 or anti-miR-203. The results confirmed that miR-203 upregulated the viral titer, while anti-miR-203 inhibited the viral titer. In summary, for the first time we showed that BANF1 inhibited the expression of HBx, HBsAg, and HBV DNA (viral load) in HBV-infected cells, and HBV inhibited BANF1 expression, at least in part, via HBx-mediated upregulation of miR-203 in hepatic cells. Cloning and SDM of BANF1 mutants. BANF1(WT) from Huh-7 cells was amplified by PCR by using the primers listed in [fig_ref] TABLE 1: List of primers used for RT-PCR [/fig_ref]. The PCR product was digested with restriction enzymes EcoRI (NEB, Ipswich, MA, USA) and BAMH1 (NEB) and cloned into pcDNA3.1(-) vector. Using the specific primers with one or two or three mutations [fig_ref] TABLE 1: List of primers used for RT-PCR [/fig_ref] , mutants (M1, M2, M3) were generated by using Quick-change II XL Site Directed Mutagenesis kit (Agilent Technologies, Santa Clara, CA, USA). SDM PCR was run according to the manufacturer's instructions and cloned into pcDNA3.1(-). All the cloned and SDM generated vectors were confirmed by sequencing.
# Materials and methods
Western blotting. The cells were harvested after 48 h of transfection using the mammalian protein extraction reagent (mPER; Thermo Fisher Scientific) supplemented with protease inhibitor cocktail (PIC, Thermo Fisher Scientific). Bicinchoninic acid (BCA) assay was performed for the estimation of protein concentration. Total cellular protein (40 mg per well) was run in the SDS-PAGE gels, transferred onto a PVDF membrane, blocked using bovine serum albumin fraction-V (HiMedia) and were incubated overnight with BANF1 (Abcam, Cambridge, UK), HBx (Santa Cruz Biotechnology, Dallas, TX, USA), or bactin (Santa Cruz). PVDF membranes were washed 6 Â 5 min each and incubated with HRP-conjugated corresponding secondary antibodies and developed using ECL kit (Thermo Fisher Scientific) on photosensitive films.
RNA isolation, cDNA synthesis, and RT-PCR. The total RNA was isolated using TRIzol reagent as per the manufacturer's instructions (Thermo Fisher Scientific; Cat. No. 15596026). The cDNA was synthesized from the total RNA using, Universal cDNA synthesis kit II (Exiqon, Vedbaek, Denmark) for miRNA expression analysis or TaKaRa cDNA synthesis kit for mRNA expression analysis. Primers for miRNA-203 (Exiqon), 5S rRNA (Exiqon), BANF1, HBx, HBsAg, HBcAg, HBV-pol and GAPDH (Integrated DNA Technologies, Skokie, IL, USA) [fig_ref] TABLE 1: List of primers used for RT-PCR [/fig_ref] were used for real-time PCR using SYBR green in a ViiA 7 real-time PCR system (Applied Biosystems, Thermo Fisher Scientific). Expression analysis of all the genes was done using 2 -DDCt method as described previously [bib_ref] Butyrate induces ROS-mediated apoptosis by modulating miR-22/SIRT-1 pathway in hepatic cancer cells, Pant [/bib_ref].
HBV DNA isolation and quantification. HBV DNA isolation and quantification protocol was followed as described previously [bib_ref] Butyrate inhibits HBV replication and HBV-induced hepatoma cell proliferation via modulating SIRT-1/Ac-p53..., Pant [/bib_ref]. Briefly, Hep G2.2.15 cells transfected with BANF1 were harvested after 48 h and cell pellet was lysed by addition of autoclaved milli Q water and TE saturated phenol. The tubes were incubated at 65°C for 2 h and then centrifuged at 12,000 rpm for 10 min followed by addition of 200 mL chloroform. The supernatant precipitation was carried out with 7.5 M ammonium acetate and 100% ethanol. DNA precipitate was washed with 70% ethanol, air dried for approximately 20 min and stored at 220°C. HBV DNA quantification was performed by qPCR using primers [fig_ref] TABLE 1: List of primers used for RT-PCR [/fig_ref] and a TaqMan probe (HBV1570-59FAM CGGACCGTGTGCACTTCGCTT-BHQ) of X region of HBV. Standard curve was derived by using serial dilutions (2 to 8 log 10 ) of HBV-DNA High control (Acrometrix, Thermo Fisher Scientific) which also served as a control. HBV DNA panel (Acrometrix, Thermo Fisher Scientific) compared with the reference (WHO HBV control) was used to standardize HBV DNA quantity. HBV DNA quantity was analyzed in international units per mL (IU/mL).
Enzyme linked fluorescent assay. Hepatitis B surface antigen (HBsAg) in cell culture supernatant was detected by enzyme linked fluorescent assay (ELFA) in automated benchtop immunoanalyzer VIDAS (BIOMÈRIEUX). HBsAg detection in 100 mL of cell culture supernatant was carried out by using VIDAS HBS kit as per manufacturer's protocol. HBsAg titer was derived from sample/cut off standard ratio using relative fluorescent value (RFV). Samples with ratio values , 0.13 were considered negative and with ratio values > 0.13 were considered positive and sample titer mentioned accordingly. The percentage inhibition was determined on the basis of sample: cut off RFV.
ELISA. Hepatitis B surface antigen (HBsAg) in cell culture supernatant was detected by enzyme linked immunosorbent assay (ELISA). HBsAg detection in 100 mL of cell culture supernatant was carried out by using Monolisa HBs Ag ULTRA (Bio-Rad) kit as per manufacturer's protocol. Briefly, 100 mL of cell culture supernatant along with positive-and negative-control samples (provided in kit) were loaded in triplicates into enzyme coated strips followed by addition of enzyme conjugate solution. The plate was incubated at 37°C for one and half hour. Strips were washed six times with provided washing solution. Chromogenic substrate was added in dark and incubated for half an hour followed by OD measurement at 450 nm in enzyme plate reader (Bio-Rad). The cutoff value was calculated using the following formula. Cutoff value = Mean absorbance of negative control 1 0.1. The samples with OD more than cutoff value were considered positive. Statistical analysis. All the experiments shown were performed at least three times (n = 3) and in duplicates. The statistical significance (P value) was calculated using unpaired Student's t test while comparing between two groups. For the Western blot analyses, a blot representative of one of the three experiments is shown. Data were considered significant if P , 0.05.
[fig] FIG 1: HBx downregulates BANF1 expression in hepatoma cells. Huh-7, Hep G2, and Hep G2.2. [/fig]
[fig] FIG 2: BANF1 overexpression inhibits expression of HBx, HBsAg and HBV replication. Huh-7 cells were transfected either alone with BANF1 plasmid, pHBV 1.3 plasmid or cotransfected together. After 48 h, cells were harvested for total cellular protein isolation and Western blots were performed. A representative picture is shown. Lane 1, control; lane 2, empty vector transfected cells; lane 3, BANF1 transfected cells; lane 4, pHBV 1.3 transfected cells; lane 5, BANF1 and pHBV 1.3 cotransfected cells (A). The band intensities were quantified using Image J software and presented as fold change (n = 3; *, P , 0.05) (B). Hep G2.2.15 cells were transfected with BANF1 (WT) plasmid and after 48 h media was collected for HBsAg analysis by Enzyme Linked Fluorescent Assay (ELFA) (n = 3; *, P , 0.05) (C) and cells were harvested for HBV DNA isolation and quantification by qPCR (n = 3; ***, P , 0.001) (D). [/fig]
[fig] FIG 3: Mutation at the N-terminus of BANF1 modulates the expression of HBx, HBsAg and HBV replication. Huh-7 cells were transfected either alone with pHBV 1.3 plasmid or cotransfected with BANF1 (WT), M1, M2, or M3 plasmid. After 48 h cells were harvested for total cellular protein isolation and Western blots were performed. A representative picture is shown. Lane 1, control; lane 2, empty vector transfected cells; lane 3, BANF1 transfected cells; lane 4, pHBV 1.3 transfected cells; lane 5, BANF1(WT) and pHBV 1.3 cotransfected cells; lane 6, M1 and pHBV 1.3 cotransfected cells; lane 7, M2 and pHBV 1.3 cotransfected cells; lane 8, M3 and pHBV 1.3 cotransfected cells (A). The band intensities were quantified using Image J software and presented as fold change (HBx/b-actin, BANF1/b-actin) (n = 3; *, P , 0.05) (B). Hep G2.2.15 cells were transfected with BANF1 (WT), M1, M2, and M3 plasmid and after 48 h media was collected for HBsAg analysis by Enzyme Linked Fluorescent Assay (ELFA) (n = 3; *, P , 0.05) (C) and cells were harvested for HBV DNA isolation and quantification by qPCR (n = 3; ***, P , 0.001) (D). [/fig]
[fig] FIG 4: HBx upregulates miR-203 expression in hepatic cells. BANF1 targeting miRNAs (miR-150, miR-203) were forecasted using miRNA target prediction tools (TargetScan and miRBase) (A). Total RNA was isolated from Huh-7, Hep G2 and Hep G2.2.15 cells followed by RT-PCR to compare the expression of miR-150 (B) and miR-203 (n = 3; **, P , 0.01) (C). Huh-7 cells were transfected with pHBV 1.3 plasmid and the cells were collected for the isolation of RNA, followed by RT-PCR for HBx (D) or miR-203 (E) expression (n = 3; ***, P , 0.001; **, P , 0.01). Huh-7 cells were transfected with HBsAg, HBcAg and HBV-pol plasmid, total RNA was isolated and RT-PCR was performed for HBsAg, HBcAg and HBV-pol (n = 3; *, P , 0.05; **, P , 0.01) (H) and miR-203 (I). After 48 h, RNA was isolated and RT-PCR was performed for the expression of HBx (F) or miR-203 (G) (n = 3; ***, P , 0.001; **, P , 0.01). [/fig]
[fig] FIG 5: miR-203 downregulates BANF1 expression in HBV expressing cells. Hep G2.2.15 cells were transfected with miR-203 pre-miR oligos and after 48 h total cellular protein was isolated, and RNA was isolated after 72 h. RT-PCR was performed for miR-203 expression (A) and BANF1 expression (B) (n = 3; ***, P , 0.001; **, P , 0.01). The total cellular protein was run on SDS-PAGE gels and Western blots were performed and a representative picture is shown. Lane 1, control; lane 2, NS transfected cells; and lane 3, miR-203 transfected cells (C). The band intensities were quantified using Image J software and presented as fold change (BANF1/b-actin) (n = 3; **, P , 0.01) (D). Hep G2.2.15 cells were transfected with anti-miR-203 oligos and after 72 h RNA was isolated followed by RT-PCR for miR-203 expression (E) and BANF1 expression (F) (n = 3; **, P , 0.01; **, P , 0.01). Hep G2.2.15 cells were transfected with either miR-203 oligos and anti-miR-203 oligos, respectively. After 48 h spent media was collected for HBsAg analysis (G), (H) (n = 3; *, P , 0.05; *, P , 0.05). [/fig]
[table] TABLE 1: List of primers used for RT-PCR [/table]
|
Platelet inhibition during ticagrelor monotherapy versus ticagrelor plus aspirin in patients with coronary artery disease (TEMPLATE study): study protocol for a randomised controlled trial
Background: Dual antiplatelet therapy (DAPT) with aspirin (ASP) and a P2Y 12 blocker is currently standard care after percutaneous coronary intervention (PCI) with stent insertion, and aims to inhibit platelet function in order to prevent stent thrombosis. The P2Y 12 blocker ticagrelor (TIC) has greater antiplatelet effect than the previously used members of this class, such as clopidogrel. In healthy volunteers, TIC is sufficient to cause strong platelet inhibition, with little additional effect from ASP. Omission of ASP may improve the safety of antiplatelet regimes by reducing bleeding. However, the effect of single antiplatelet treatment with TIC, compared to DAPT with TIC + ASP, has not been studied in detail in patients with coronary artery disease.Methods: To compare TIC with TIC + ASP, we have initiated a single centre, open-label randomised controlled trial (TEMPLATE study) in adults receiving DAPT following PCI with a sample size of 110 patients. Patients are invited to join the study when, as part of standard care, they are due to switch from DAPT (ASP + any P2Y 12 blocker) to single antiplatelet treatment with ASP alone after 6-12 months. Patients are randomised to receive either TIC or TIC + ASP for 4 weeks. All patients then revert to standard care with ASP alone. Blood samples and clinical data are collected at three study visits: at baseline during treatment with ASP + any P2Y 12 blocker (visit 1); approximately 4 weeks after visit 1 during treatment with either TIC or TIC + ASP (visit 2); and approximately 8 weeks after visit 1 when treatment has reverted to ASP alone (visit 3). The primary outcome is the extent of platelet inhibition, measured by light transmission aggregation, flow cytometry, flow chamber and plasma biomarker tests. The primary analysis will compare the extent of platelet inhibition between the TIC and TIC + ASP groups at visit 2, adjusted for baseline platelet reactivity. Secondary analyses will compare the extent of platelet inhibition at visit 2 with that at visit 3. Discussion: This is the first study to compare in detail the extent of platelet inhibition in patients who are receiving TIC compared with TIC + ASP. The study findings will complement larger-scale trials of the clinical efficacy and safety of TIC compared to TIC + ASP. Trial registration: ISRCTN registry, identifier ISRCTN84335288. Registered on 23 June 2014.
# Background
Dual antiplatelet therapy (DAPT) with aspirin (ASP) and a P2Y purinoreceptor (P2Y 12 ) blocker is standard care for patients with coronary artery disease in several settings, including after acute coronary syndrome (ACS) and percutaneous coronary intervention (PCI) with or without coronary stent insertion [bib_ref] Effects of pretreatment with clopidogrel and aspirin followed by long-term therapy in..., Mehta [/bib_ref] [bib_ref] ACCF/AHA focused update incorporated into the ACC/AHA 2007 guidelines for the management..., Anderson [/bib_ref]. The rationale for this treatment is that most adverse cardiac events that occur in these settings result from the formation of abnormal platelet aggregates causing coronary artery or stent thrombosis. Inhibition of platelets by simultaneously targeting both the thromboxane A 2 pathway (with ASP) and the P2Y 12 receptor pathway (with a P2Y 12 blocker) results in strong platelet inhibition and helps prevent aggregate formation [bib_ref] Effects of pretreatment with clopidogrel and aspirin followed by long-term therapy in..., Mehta [/bib_ref]. The intensity of antiplatelet treatment is typically reduced at 6-12 months following ACS or PCI by switching patients from DAPT to single-agent treatment with ASP alone. This ensures strong platelet inhibition in the highest risk period after ACS or PCI and minimises the long-term bleeding risk associated with DAPT.
When used in combination with ASP, clopidogrel (CLOP; the first widely used P2Y 12 blocker) reduced the incidence of major adverse cardiac events compared to single antiplatelet treatment with ASP [bib_ref] Aspirin has little additional anti-platelet effect in healthy volunteers receiving prasugrel, Leadbeater [/bib_ref]. However, approximately 30% of individuals display incomplete inhibition of the platelet P2Y 12 receptor pathway with CLOP, resulting in reduced efficacy [bib_ref] Cytochrome 2C19*17 allelic variant, platelet aggregation, bleeding events, and stent thrombosis in..., Sibbing [/bib_ref]. These limitations have been circumvented by the development of strong P2Y 12 blockers such as ticagrelor (TIC), which give greater and more consistent inhibition of the P2Y [bib_ref] Shorter (</=6 months) versus longer (>/=12 months) duration dual antiplatelet therapy after..., Pandit [/bib_ref] receptor pathway compared with CLOP. Accordingly, TIC in combination with ASP is associated with fewer major adverse cardiac events than CLOP in combination with ASP [bib_ref] Ticagrelor versus clopidogrel in patients with acute coronary syndromes, Wallentin [/bib_ref].
More recent observations made in experimental models of platelet inhibition [bib_ref] Reduction of platelet thromboxane A(2) production ex vivo and in vivo by..., Armstrong [/bib_ref] and in observational studies of healthy volunteers receiving antiplatelet treatments [bib_ref] Aspirin has little additional anti-platelet effect in healthy volunteers receiving prasugrel, Leadbeater [/bib_ref] suggest that strong P2Y 12 blockers such as TIC used alone result in a consistent and full platelet inhibition, and that little further platelet inhibition is achieved by the addition of ASP. In response to these findings, it has been proposed that medication with strong P2Y 12 blockers without additional ASP may be sufficient to achieve full therapeutic antiplatelet effect in clinical settings in which DAPT is currently standard care [bib_ref] Dual antiplatelet therapy in cardiovascular disease: does aspirin increase clinical risk in..., Warner [/bib_ref]. Avoidance of ASP may also confer safety benefits because ASP is associated with gastrointestinal bleeding due to its local effect on gastric mucosa, unrelated to the antiplatelet effect [bib_ref] Endogenous biosynthesis of prostacyclin and thromboxane and platelet function during chronic administration..., Fitzgerald [/bib_ref].
Single antiplatelet treatment with TIC is currently being evaluated in the GLOBAL LEADERS study, which is a large, international, multicentre, randomised controlled trial of TIC versus TIC + ASP following PCI and coronary stent insertion [bib_ref] Long-term ticagrelor monotherapy versus standard dual antiplatelet therapy followed by aspirin monotherapy..., Vranckx [/bib_ref]. However, the main endpoints of this trial are all-cause mortality or non-fatal myocardial infarction and bleeding. The GLOBAL LEADERS study does not include a detailed laboratory evaluation of the pharmacodynamic effect of the two different treatment regimes on platelet inhibition.
## Aims and objectives
The main aim of the TEMPLATE study is to determine the extent of platelet inhibition in patients with coronary artery disease achieved with single antiplatelet treatment with TIC compared with DAPT with TIC + ASP. The secondary aim is to compare the effect of TIC or TIC + ASP to single antiplatelet treatment with ASP alone.
The study objectives are to perform laboratory measurements of platelet inhibition in patients at the following points of treatment with antiplatelet drugs after PCI: (i) at the end of DAPT with ASP + any P2Y 12 blocker treatment; (ii) after 4 weeks of treatment with either TIC or TIC + ASP; and, (iii) after all patients are switched to ASP alone after the study intervention.
# Methods
## Study design and setting
The TEMPLATE study is a single centre, open-label, randomised controlled trial of the effects of different antiplatelet treatments on platelet laboratory test results in patients who have undergone PCI and coronary stent insertion at the Bristol Heart Institute (BHI).
## Study population
The study population is patients at the BHI who have been prescribed DAPT as part of standard clinical care following PCI and coronary stent insertion.
A patient will be eligible for inclusion if: (i) prescribed DAPT with ASP + any P2Y 12 blocker for a minimum interval of 4 weeks; and (ii) intending to transfer to ASP after cessation of treatment with ASP + any P2Y 12 blocker; and, (iii) aged over 18 years at the point of randomisation.
Patients will be ineligible if any of the following apply: (i) contraindication to DAPT; (ii) treatment with ASP + any P2Y 12 blocker after PCI was interrupted or terminated because of bleeding or increased bleeding risk; (iii) contraindication to TIC; (iv) pregnancy and or breast feeding; (v) the patient is a woman of childbearing potential who is unwilling to use contraception; (vi) the patient is a man with a spouse or partner with childbearing potential unless the patient is sterilised or has agreed to use barrier contraceptives.
Concomitant medications which may modify the therapeutic effect of TIC or which may have their therapeutic effect modified by TIC are not absolute exclusions to the study, but will be considered as potential reasons for exclusion by the enrolling clinician. These include: (i) strong CYP3A4 inducers (e.g. rifampicin, dexamethasone, phenytoin, carbamazepine and phenobarbital), (ii) CYP3A4 substrates with narrow therapeutic indices (e.g. cisapride, ergot alkaloids, simvastatin or lovastatin, (iii) digoxin, and (iv) potent P-glycoprotein inhibitors (e.g. verapamil, quinidine, cyclosporin).
Concomitant medications with a potential adverse interaction when co-prescribed with ASP will be considered as potential reasons for exclusion by the enrolling clinician. These include: some diuretics, probenecid, heparin, phenindione, antiepileptics such as phenytoin, warfarin, insulin and sulphonylurea hypoglycaemic agents, non-steroidal anti-inflammatory drugs, methotrexate, corticosteroids, antacids, iron salts, carbonates, alkali hydroxides and ibuprofen.
## Patient approach and study visits
Potentially eligible patients will be identified using hospital databases which record the demographic details of patients who undergo interventional cardiology procedures at the BHI and the antiplatelet treatment prescribed after the procedure. The first approach will be an invitation letter and patient information leaflet. Since most patients in the study population were receiving DAPT for 6-12 months, this was usually sent by post approximately 6 weeks before the planned completion of DAPT. Potentially eligible patients who reply expressing an interest in taking part will be contacted by telephone and invited to attend three outpatient study visits at the BHI. They will be instructed to continue with their existing DAPT until the first study visit. A study flow diagram is shown in .
At each of three study visits, clinical data will be collected and an 18 mL venous blood sample will be obtained by peripheral venepuncture for laboratory analysis. The schedule of data collection is shown in [fig_ref] Figure 2: Schedule of data collection [/fig_ref]. All data will be collected on a bespoke database and stored on a secure server.
## Study visit 1
Study visit 1 will occur when a patient would be switching from DAPT with ASP + any P2Y 12 blocker to ASP alone in the standard clinical care pathway. A member of the research team will describe the study, answer questions and confirm eligibility. Eligible patients will be invited to complete and sign the study consent form.
## Randomisation
Allocation to the TIC + ASP group or the TIC group in a 1:1 ratio will be performed by an authorised member of the research team using a secure internet-based randomisation system (http://www.sealedenvelope.com/), using block randomisation with blocks of varying sizes. Clinicians, research nurses, trial coordinators, pharmacists and patients will be aware of the allocation. As the trial is open label, no code-breaking procedure is required. Since the costs associated with production of a placebo medication were prohibitively high, we adopted a pragmatic open-label design in which the study patients, their clinicians, research nurse and trials coordinator will not be blinded to treatment allocation. We do not anticipate this to have a significant impact on trial results. The laboratory staff will be blinded to group allocation.
## Trial interventions
Patients randomised to the TIC + ASP group will transfer from ASP + any P2Y 12 blocker to ASP 75 mg once per day plus a loading dose of TIC 180 mg followed by TIC 90 mg twice per day. Patients randomised to the TIC group will transfer from ASP + any P2Y 12 blocker to a loading dose of TIC 180 mg followed by TIC 90 mg twice per day. The study medication will continue to be taken until study visit 2 at 28 days (acceptable range 21-35 days) after enrolment and randomisation. The study medication will be stored and dispensed by the trial site's pharmacy in accordance with Good Clinical Practice. Patients will be asked to record any missed doses, and to return any unused medication at their second study visit to ascertain adherence. Patients will be instructed to take the next scheduled dose as usual, if they miss a dose.
## Study visit 2
Patients will be interviewed by the study team to collect safety data and to check compliance with the study medication [fig_ref] Figure 2: Schedule of data collection [/fig_ref]. The safety data will include adverse events potentially associated with the study medication, the occurrence of major adverse cardiac events and bleeding, categorised using the Bleeding Academic Research Consortium (BARC) definitions of bleeding [fig_ref] Table 1: Bleeding Academic Research Consortium [/fig_ref] [bib_ref] Standardized bleeding definitions for cardiovascular clinical trials a consensus report from the..., Mehran [/bib_ref].
All patients will then be transferred to ASP 75 mg once per day, thereby returning to standard clinical care. This treatment will continue until study visit 3 which will be 28 days (acceptable range 21-35 days) after study visit 2, and thereafter, long-term. Patients who are unable to attend study visit 2 within 35 days from study visit 1 will be advised to transfer to ASP 75 mg once per day at 35 days after study visit 1 and will be invited to attend for a final study visit at the same time as their planned study visit 3.
## Study visit 3
Patients will be interviewed by the study team to collect safety data and to check compliance with the prescribed medication. Patients will then exit the study and will be instructed to continue ASP 75 mg once per day as per standard clinical care under the ongoing supervision of their primary care practitioner [fig_ref] Figure 2: Schedule of data collection [/fig_ref].
## Definition of end of trial
Patients will end their involvement in the study after they have completed study visit 3, at approximately 8 weeks after study visit 1. The end of the study is when analyses of blood samples are completed, data queries are resolved, the database is locked and the analysis completed.
## Loss to follow-up and withdrawals
Each patient has the right to withdraw at any time. In this event, we will analyse any data already collected, unless the patient expresses a wish for their samples to be destroyed and any associated data not to be included in the analysis. We expect low rates of patient withdrawal and losses to follow-up because the study interval is only approximately 8 weeks and requires only three study visits. It is also unlikely that patients will have to be withdrawn from their allocated study treatment as the study intervention (TIC + ASP or TIC) is well tolerated and is prescribed for only a 4-week period. However, in the event of a serious adverse event that the investigator believes may be attributable to the intervention, the study medication will be discontinued and the patient will be returned to standard care with ASP 75 mg once daily. The trial medication may also be discontinued if any contraindicated therapy is initiated whilst the patient is participating in the trial. Data collected before discontinuation of trial medication will be included in the analyses.
## Laboratory analyses
The blood samples obtained at the three study visits will be analysed using a panel of platelet function and biomarker tests in a central laboratory. All laboratory Trial schema. Schema showing the recruitment pathway with the number of patients to be recruited, anticipated eligibility, recruitment and follow-up rates. ASP aspirin, CLOP clopidogrel, TIC ticagrelor tests will be performed in accordance with reagent manufacturers' instructions and with locally generated standard operating procedures. The following analyses will be performed:
(i) Light transmission aggregation (LTA) in plateletrich plasma (PRP) in response to stimulation with collagen-related peptide (CRP), adenosine diphosphate (ADP), arachidonic acid (AA), thromboxane receptor agonist (U46619) and thrombin-related activator peptide 6 (TRAP 6). (ii)Flow cytometry to quantify surface CD62P expression and PAC-1 binding in unstimulated platelets and in platelets stimulated with CRP and TRAP 6. (iii)Adhesion and aggregation of platelets in whole blood on a collagen surface under flow conditions. (iv)Plasma concentrations of soluble CD40 ligand (sCD40L) and thromboxane B2 (TXB2) to assess baseline in vivo platelet activation and thromboxane A2 biosynthesis respectively. These biomarkers will be measured by enzyme-linked immunosorbent assay using frozen plasma samples, after the end of recruitment.
## Primary outcome
The primary outcome is the maximum amplitude of the LTA response to TRAP 6 expressed as a percentage of the absolute difference in light transmission between PRP and platelet-poor plasma. This was selected as the primary outcome since the functional response to TRAP 6 represents the best marker of the overall extent of platelet inhibition.
## Secondary outcomes
The secondary outcomes will be the following laboratory test results:
(i) Maximum amplitude of the LTA responses to CRP, ADP, AA and U46619. (ii)Median fluorescence intensity of anti-P selectin and PAC-1 binding to platelets in unstimulated platelets and after stimulation with CRP and with TRAP 6. (iii)Aggregation and adhesion parameters of platelets exposed to a collagen surface under flow conditions. (iv)Plasma concentrations of sCD40L and TXB2.
## Sample size
The sample size has been set at 110 patients, which is sufficient to detect a standardised difference in the primary outcome between the two study interventions of 0.35 standard deviations. The sample size has been calculated with the following assumptions:
1. Correlation between the primary outcome measured at visit 1 and at visit 2 = 0.5 2. No cross-over anticipated 3. 10% loss to follow-up (i.e. missing outcome data at visit 2)
DAPT with ASP + any P2Y 12 blocker is prescribed as part of standard current care at the BHI to approximately 1500 adults per year who undergo PCI and coronary stent insertion. We estimate that approximately 1200 (80%) will complete the prescribed duration of treatment of 6-12 months. Based on previous observational studies at the BHI, we conservatively estimate that the overall recruitment and completion rate will be approximately 10% of eligible patients. Therefore, we predict that 110 patients will occur in 14-16 months.
# Statistical analysis
The laboratory test results will be expressed as continuous variables. The results for the primary comparison of the maximum amplitude of the LTA response to TRAP 6 between the TIC + ASP and TIC groups will be analysed using linear mixed regression. The baseline (visit 1) and post-intervention values (visit 2) will be modelled jointly in preference to the pre-randomisation value being modelled as a covariate to avoid the necessity either to exclude cases with missing preintervention measures or to impute missing preintervention values. The secondary outcome measures will be compared similarly.
The study will also enable the extent of platelet inhibition to be described within subjects over time as their treatment changes (i.e. from visit 1 to visit 2 to visit 3).
Comparative analyses will be on the basis of intention to treat and mean differences will be quantified with 95% confidence intervals. The width of the individual confidence intervals for the secondary outcomes will be adjusted for the number of tests carried out (Bonferroni correction).
The primary analysis will take place when follow-up and laboratory analyses are complete for all recruited patients. No interim or subgroup analyses are planned other than those outlined above. Safety data will be reviewed regularly by the trial management group who meet at frequent intervals to discuss trial progress. During these meetings the rate of adverse events in the study population will be monitored against the anticipated rate of events that are expected for the target population.
## Research approval
Research ethics approval and clinical trial authorisation were granted in October 2014 by the South-Central Oxford A Research Ethics Committee (reference 14/SC/ 1309) and the Medicines and Healthcare Regulatory Agency (EUDRACT No. 2013-002734-20) respectively. The trial is managed by the Clinical Trials and Evaluation Unit Bristol (CTEU Bristol) and sponsored by the University Hospitals Bristol NHS Foundation Trust (www.uhbristol.nhs.uk/research-innovation/). Trial governance is through the sponsor's standard operating procedures. The trial is registered as ISRCTN84335288 and conforms to SPIRIT guidelines (see [fig_ref] Figure 2: Schedule of data collection [/fig_ref] and Additional file 1: SPIRIT checklist).
## Changes to the protocol since first approved
There have been three substantial amendments made to the protocol since it was first approved. The first amendment was approved in May 2015 and incorporated minor clarifications in addition to the new recommendations from the European Society of Cardiologists. These recommendations led to a change in duration of standard DAPT specified in the protocol, from 12 months to between 6 and 12 months after PCI and insertion of a drug-eluting stent; this change increased the pool of eligible patients.
The second amendment to the protocol was approved in August 2015, removing the plan to establish a Data Monitoring and Safety Committee (DMSC). This change was made because, on review of the study design, it became apparent that the study would not benefit from a Type 1 Bleeding that is not actionable and does not cause the patient to seek unscheduled performance of studies, hospitalisation, or treatment by a healthcare professional; may include episodes leading to self-discontinuation of medical therapy by the patient without consulting a healthcare professional.
Type 2 Any overt, actionable sign of haemorrhage (e.g. more bleeding than would be expected for a clinical circumstance, including bleeding found by imaging alone) that does not fit the criteria for type 3, 4, or 5 but does meet at least one of the following criteria: requiring nonsurgical, medical intervention by a healthcare professional, leading to hospitalisation or increased level of care, prompting evaluation.
Type 3a Overt bleeding plus haemoglobin drop of 3 to < 5 g/dL, corrected for transfusion (provided haemoglobin drop is related to bleed) OR any transfusion with overt bleeding. The third amendment was approved in September 2016. This amendment widened the inclusion criteria from receiving ASP + CLOP for 6-12 months to receiving ASP + any P2Y 12 blocker for 6-12 months. This widening of the inclusion criteria was made to enhance study enrolment because shortly after starting the study, the institutional standard care changed from using CLOP as the preferred P2Y 12 blocker, to other P2Y 12 blockers for some patient subgroups. After this amendment, the enrolment rate increased to that predicted in the original study design.
# Discussion
The TEMPLATE study design will enable unbiased comparison of the effects of TIC versus TIC + ASP on platelet activity in patients with coronary artery disease. It will also enable a longitudinal comparison of the effects TIC and TIC + ASP with the effects of ASP alone in the same patients. The laboratory tests selected for this study will enable measurement of functional platelet responses to a panel of activating agonists using LTA, flow cytometry and flow chamber tests, selected to measure the extent of inhibition of the multiple platelet activation pathways. We will also measure the extent of baseline platelet activity by testing unstimulated platelets by flow cytometry and with the soluble platelet activation biomarker tests. Together, these data will provide a comprehensive description of the overall pharmacodynamic effects of the different antiplatelet treatments. This information has not been reported previously in cohorts of patients with coronary artery disease receiving TIC or TIC + ASP. To our knowledge, there are no similar previous studies in this patient group that have evaluated the pharmacodynamic effect of monotherapy with other P2Y 12 blockers, including CLOP.
The TEMPLATE study design allows the information to be obtained in a low-risk setting. We are enrolling patients at the point in the standard clinical care pathway when there is a transition from DAPT to singleantiplatelet treatment with ASP, typically at 6 to 12 months after the PCI procedure. This is after the interval of greatest risk of major adverse cardiac events and abnormal bleeding following PCI [bib_ref] Shorter (</=6 months) versus longer (>/=12 months) duration dual antiplatelet therapy after..., Pandit [/bib_ref]. Furthermore, the duration of exposure to the study treatment (i.e. either TIC or TIC + ASP) is only 4 weeks, which will minimise the likelihood of adverse events attributable to the study drugs. The short overall study participation of 8 weeks during which there are only three study visits should facilitate study delivery and minimise patient drop-out.
The main benefit of the TEMPLATE study is that the findings will provide evidence to either support or refute the hypothesis that TIC monotherapy enables platelet inhibition to the same extent as dual therapy with TIC + ASP. This is significant clinically because abnormal bleeding is a major complication of all antiplatelet therapy, but particularly of ASP which adversely affects gastric mucosa, increasing the risk of gastrointestinal bleeding [bib_ref] Endogenous biosynthesis of prostacyclin and thromboxane and platelet function during chronic administration..., Fitzgerald [/bib_ref]. If the findings of the TEMPLATE study indicate no difference in platelet inhibition between the TIC treatment group and the TIC + ASP treatment group, then this would support avoidance of ASP in this patient group, potentially conferring a safety advantage. This is also part of the rationale of the large-scale GLOBAL LEADERS randomised controlled trail (NCT01813435) of 16,000 participants in which the study group receive TIC + ASP for 1 month after PCI and coronary stent insertion followed by TIC monotherapy for a further 23 months. The control group is DAPT with TIC + ASP or CLOP + ASP for 12 months [bib_ref] Long-term ticagrelor monotherapy versus standard dual antiplatelet therapy followed by aspirin monotherapy..., Vranckx [/bib_ref]. The outcomes of the GLOBAL LEADERS study are all-cause mortality or non-fatal new Q-wave myocardial infarction (composite primary outcome) and bleeding (secondary outcome), but do not include pharmacodynamic measures of the effect of the different antiplatelet regimes. We anticipate that reporting of TEMPLATE study results in a peer-reviewed publication will precede reporting of the GLOBAL LEADERS study results and that our results will assist interpretation and help inform future treatment guidelines.
## Study status
The trial opened to recruitment in December 2015 with the first patient recruited in January 2016. Recruitment is ongoing.
## Trial steering committee members
The Trials Steering Committee comprises a consultant anaesthetist, interventional consultant cardiologist and a consultant cardiac surgeon.
## Additional file
Additional file 1: SPIRIT 2013 checklist: recommended items to address in a clinical trial protocol and related documents. (DOC 123 kb)
Abbreviations AA: Arachidonic acid (a platelet metabolite that stimulates platelets in vivo and is also used as a laboratory reagent to study platelet activation responses); ACS: Acute coronary syndrome; ADP: Adenosine diphosphate (a P2Y 12 receptor agonist that stimulates platelets in vivo and is also used as a laboratory reagent to study platelet activation responses); ASP: Aspirin; BARC: Bleeding Academic Research Consortium; BHI: Bristol Heart Institute; CABG: Coronary artery bypass grafting; CLOP: Clopidogrel; CTEU Bristol: Clinical Trials and Evaluation Unit Bristol; CRP: Collagen-related peptide (a platelet activating chemical that is used as a laboratory reagent to study platelet activation responses); DAPT: Dual antiplatelet therapy; DMSC: Data monitoring and safety committee; LTA: Light transmission aggregation; NHS: National
[fig] Figure 2: Schedule of data collection (SPIRIT) [/fig]
[table] Table 1: Bleeding Academic Research Consortium (BARC) definitions [/table]
|
Nalmefene attenuates neural alcohol cue-reactivity in the ventral striatum and subjective alcohol craving in patients with alcohol use disorder
Rationale Alcohol use disorder is a common and devastating mental illness for which satisfactory treatments are still lacking. Nalmefene, as an opioid receptor modulator, could pharmacologically support the reduction of drinking by reducing the (anticipated) rewarding effects of alcohol and expanding the range of treatment options. It has been hypothesized that nalmefene acts via an indirect modulation of the mesolimbic reward system. So far, only a few imaging findings on the neuronal response to nalmefene are available.Objectives We tested the effect of a single dose of 18 mg nalmefene on neuronal cue-reactivity in the ventral and dorsal striatum and subjective craving. Methods Eighteen non-treatment-seeking participants with alcohol use disorder (67% male, M = 50.3 ± 13.9 years) with a current high-risk drinking level (M = 76.9 ± 52 g of pure alcohol per day) were investigated using a cue-reactivity task during functional magnetic resonance imaging (fMRI) in a double-blind, placebo-controlled, cross-over study/design. In addition, selfreported craving was assessed before and after exposure to alcohol cues. Results An a priori defined region of interest (ROI) analysis of fMRI data from 15 participants revealed that nalmefene reduced alcohol cue-reactivity in the ventral, but not the dorsal striatum. Additionally, the subjective craving was significantly reduced after the cue-reactivity task under nalmefene compared to placebo. Conclusion In the present study, reduced craving and cue-reactivity to alcohol stimuli in the ventral striatum by nalmefene indicates a potential anti-craving effect of this drug via attenuation of neural alcohol cue-reactivity.
# Introduction
Alcohol use disorder (AUD) is one of the most prevalent substance use disorders worldwide. However, only 22% of AUD patients in Europe receive an addiction-specific treatment [bib_ref] Alcohol consumption, alcohol dependence and attributable burden of disease in Europe: potential..., Rehm [/bib_ref]. One possible reason for this low treatment rate could be a lack of willingness for abstinence. Even if abstinence should be the primary goal of addiction treatment, reducing alcohol consumption as a harm reduction may be an alternative treatment option. Against this background, reducing this unwanted treatment gap might be achieved by offering patients to choose between abstinence and reduced drinking as their individual treatment goal [bib_ref] Reduced-risk drinking as a treatment goal: what clinicians need to know, Ambrogne [/bib_ref] [bib_ref] Alcohol dependence and harmful use of alcohol, Batra [/bib_ref] [bib_ref] Controlled drinking after 25 years: how important was the great debate?, Sobell [/bib_ref].
Treating alcohol use disorder pharmacologically, disulfiram (approved by U.S. Food and Drug Administration but not approved by European Medicines Agency [EMA] anymore), acamprosate (approved by FDA and EMA), as well as naltrexone (approved by FDA and EMA), and nalmefene (approved by EMA) present the currently available and approved options [bib_ref] Pharmacotherapy of alcoholism-an update on approved and off-label medications, Soyka [/bib_ref]. Naltrexone, which was indicated for relapse prevention, to remain abstinent and to reduce craving in alcohol use disorder (Anton 2008) is a (μ-, δ-, and κ-) opioid receptor antagonist. By this means, opioid antagonists play a key role in mediating the rewarding effects of alcohol [bib_ref] Influence of the endogenous opioid system on high alcohol consumption and genetic..., Gianoulakis [/bib_ref] , by suppressing the alcohol-induced release of dopamine in the mesolimbic reward system [bib_ref] The dopamine hypothesis of reward: past and current status, Spanagel [/bib_ref] [bib_ref] Anti-craving compounds for ethanol: new pharmacological tools to study addictive processes, Spanagel [/bib_ref] , which again might reduce the subjective alcohol craving [bib_ref] Naltrexone and coping skills therapy for alcohol dependence: a controlled study, O'malley [/bib_ref] [bib_ref] Naltrexone in the treatment of alcohol dependence, Volpicelli [/bib_ref]. The approval by the FDA was based on these data showing that naltrexone decrease relapse rates to alcohol use, reduced drinking days, and alcohol craving [bib_ref] Naltrexone and coping skills therapy for alcohol dependence: a controlled study, O'malley [/bib_ref] [bib_ref] Naltrexone in the treatment of alcohol dependence, Volpicelli [/bib_ref] probably by reducing the positive reinforcing effects of alcohol and/or the anticipation of such effects [bib_ref] Translating the neuroscience of alcoholism into clinical treatments: from blocking the buzz..., Heilig [/bib_ref]. This is also reflected in a decreased fMRI cue-reactivity by naltrexone in the ventral striatum of non-treatment-seeking alcoholics [bib_ref] Effect of naltrexone and ondansetron on alcohol cue-induced activation of the ventral..., Myrick [/bib_ref]. The ventral striatum has been associated to motivational reward processes, which is manifested in a greater activation by alcohol respectively reward-associated stimuli, while the dorsal striatum has been linked to stereotyped and automated behavior [bib_ref] Alcohol-associated stimuli activate the ventral striatum in abstinent alcoholics, Braus [/bib_ref] [bib_ref] Neural systems of reinforcement for drug addiction: from actions to habits to..., Everitt [/bib_ref] [bib_ref] Drug addiction: updating actions to habits to compulsions ten years on, Everitt [/bib_ref] [bib_ref] Functional neuroimaging studies of alcohol cue reactivity: a quantitative meta-analysis and systematic..., Schacht [/bib_ref]. Also, imaging studies indicate that there is a shift from ventral to dorsal cue processing [bib_ref] Initial, habitual and compulsive alcohol use is characterized by a shift of..., Vollstädt-Klein [/bib_ref] and overreliance on habitual learning with increased activation in dorsal striatum in the presence of alcohol dependence [bib_ref] Behavioral and neuroimaging evidence for overreliance on habit learning in alcohol-dependent patients, Sjoerds [/bib_ref] in the course of addiction development. Against this background, cue-induced brain activation measured with fMRI in the striatum and the ACC was associated with an increased amount of drinking at follow-up [bib_ref] Neural substrates of cue reactivity: association with treatment outcomes and relapse, Courtney [/bib_ref] [bib_ref] Cue-induced activation of the striatum and medial prefrontal cortex is associated with..., Grüsser [/bib_ref]. In addition to these neurobiological findings, a metaanalysis of [bib_ref] Pharmacotherapy for adults with alcohol use disorders in outpatient settings: a systematic..., De [/bib_ref] showed that the numbers needed to treat for benefit (NNTs) for naltrexone (50 mg/day) were 20 to prevent return to any drinking and 12 to prevent return to heavy drinking. Considering this, recently, nalmefene has been approved by the EMA specifically for the reduction of alcohol consumption in adult patients suffering from AUD and a high drinking risk level (> 60 g pure alcohol on a single drinking day for men and > 40 g for women), without physical withdrawal symptoms and not requiring immediate detoxification (Online document. Online document. European Medicines Agency [EMA] 2013). In 2005, about 11 million people in Europe aged between 18 and 64 years suffered from alcohol dependence, and in 2009, 15% of men and 8% of woman consumed alcohol at a high or very high risk level in Europe [bib_ref] Alcohol consumption, alcohol dependence and attributable burden of disease in Europe: potential..., Rehm [/bib_ref] according to which nalmefene would address a sizeable group.
Nalmefene acts as an opioid system modulator with antagonistic activity at the μand δ-receptors, as well as naltrexone, and partial agonistic activity at the κ-receptor [bib_ref] Nalmefene induced elevation in serum prolactin in normal, Bart [/bib_ref] , distinguishing the drug from other drugs that act within the opioid system, such as naltrexone which is a full kappa receptor antagonist [bib_ref] Initial, habitual and compulsive alcohol use is characterized by a shift of..., Vollstädt-Klein [/bib_ref].
Besides this different pharmacological principle, the treatment with nalmefene also differs from previous treatment strategies. While therapy with naltrexone takes place continuously, nalmefene should be taken as needed before high-risk drinking situations (Online document. Online document. European Medicines Agency [EMA] 2013). On this occasion, nalmefene should reduce the reinforcing effect of alcohol due to its influence on the mesolimbic reward system, alleviating the reduction of alcohol consumption.
So far, previous investigations showed that nalmefene as on-demand medication is superior to placebo in reducing the number of heavy drinking days [bib_ref] Extending the treatment options in alcohol dependence: a randomized controlled study of..., Mann [/bib_ref] [bib_ref] Nalmefene for the management of alcohol dependence: review on its pharmacology, mechanism..., Mann [/bib_ref] [bib_ref] Efficacy of as-needed nalmefene in alcohol-dependent patients with at least a high..., Van Den Brink [/bib_ref]. More recent investigations have also shown that nalmefene given as needed reduces the number of heavy drinking days after 12 weeks compared to placebo [bib_ref] Nalmefene in alcohol-dependent patients with a high drinking risk: randomized controlled trial, Miyata [/bib_ref] and treatment as usual [bib_ref] Nalmefene, given as needed, in the routine treatment of patients with alcohol..., Castera [/bib_ref]. One further study [bib_ref] Nalmefene reduces reward anticipation in alcohol dependence-an experimental fMRI study, Quelch [/bib_ref] demonstrated that nalmefene reduces in the presence of the alcohol infusion the neuronal response regarding the anticipation of reward in striatal areas if compared to placebo.
Nevertheless, hitherto, the neurobiological mechanism of action of nalmefene is not well understood and needs further investigation. Therefore, we investigated the effect of single-dose nalmefene on neural alcohol-cue-reactivity during fMRI and subjective alcohol craving. We hypothesized that a single dose of nalmefene is superior over placebo in decreasing neural reactivity in the ventral and dorsal striatum, following the presentation of alcohol associated visual stimuli. This would suggest that nalmefene can dampen the hedonistic effects of alcohol or counteract the anticipation of this effect. Due to the shift from ventral to dorsal striatal cue processing [bib_ref] Initial, habitual and compulsive alcohol use is characterized by a shift of..., Vollstädt-Klein [/bib_ref] and preliminary fMRI findings that show that opiate antagonists like naltrexone and nalmefene are able to reduce brain activity in mesolimbic pathway after cue exposure [bib_ref] AUDIT. The alcohol use disorders identification test. Guidelines for Use in Primary..., Babor [/bib_ref] [bib_ref] Effect of naltrexone and ondansetron on alcohol cue-induced activation of the ventral..., Myrick [/bib_ref] [bib_ref] Nalmefene reduces reward anticipation in alcohol dependence-an experimental fMRI study, Quelch [/bib_ref] , we concentrate in particular on the ventral and dorsal striatum.
# Methods and materials
The effect of a single-dose nalmefene (18 mg) on cuereactivity was examined prospectively, using a double-blind, placebo-controlled study design (cross-over design) and functional magnetic resonance imaging (fMRI) (registration at clinicaltrials.gov; NCT02372318). Heavy-drinking (alcohol consumption > 60 g for men and > 40 g for women; at least 5 days/week), non-treatment-seeking participants with alcohol use disorder (≥ five DSM-5 AUD criteria) were selected as the target population. Recruitment was conducted through local notices, several press releases, advertisements, and bulletin in social media. All participants provided written informed consent prior to study participation.
## Study design
The study consisted of a telephone screening, a baseline screening, and two investigational days including fMRI examination of cue-reactivity (see [fig_ref] Figure 1: Study procedures [/fig_ref]. Participants had to meet all following inclusion criteria in order to take part in the study: (1) participants had to be aged between 18 and 70 years; (2) they had to meet at least five diagnostic criteria for an alcohol use disorder according to the Diagnostic Statistical Manual of Mental Disorders (DSM-5); (3) the average amount of consumed pure alcohol should be at least ≥ 60 g for men and ≥ 40 g for women per day (at least 5 days/week); (4) they should have a sufficient visual acuity (binocular [corrected] ≥ 0.8). Participants were excluded if they met one or more of the following exclusion criteria: (1) previous inpatient detoxification treatment; (2) current withdrawal symptoms (CIWA-Ar > 4; [bib_ref] Assessment of alcohol withdrawal: the revised clinical institute withdrawal assessment for alcohol..., Sullivan [/bib_ref] , previous severe withdrawal or any withdrawal complications; (3) other Axis I psychiatric diagnoses than alcohol or tobacco use disorder in the last 12 months screened with Structured Clinical Interview (SKID-I) for DSM-4 [bib_ref] SKID-I. Strukturiertes Klinisches Interview für DSM-IV. Achse I: Psychische Störungen, Wittchen [/bib_ref] due to the unavailability of a SKID for DSM-5 at the time of examination; (4) neurological disorders respectively history of brain injury; (5) at the time of the examination psychotropic medication within the last 14 days; (6) an intoxication (breath alcohol concentration > 0.3‰); (7) positive drug screening (opioids, cannabinoids, benzodiazepines, barbiturates, cocaine, amphetamines); (8) positive pregnancy test or (9) contraindications to the prescription of nalmefene (e.g., known intolerance, current use of opioid analgesics or opioid-containing antidiarrheal, positive opioid in urine, opiate withdrawal syndrome, or severe liver dysfunction); and (10) exclusion criteria for MRI (e.g., metal implants and claustrophobia).
On screening day, participants were informed about study procedures and possible risks such as medication side effects or headaches after fMRI. Basic sociodemographic information was documented, and history of somatic illnesses and neurological and mental disorders as well as current medication was recorded. Health status was assessed by medical examination. Absence of current illicit drug abuse was verified via urine screening; for women, a pregnancy screen was conducted additionally. An alcohol breath test was performed to confirm abstinence (breath alcohol concentration > 0.3). After medical check-up, participants underwent the Structured Clinical Interview for DSM-4 to verify the absence of Axis-I disorder (SKID-I; [bib_ref] SKID-I. Strukturiertes Klinisches Interview für DSM-IV. Achse I: Psychische Störungen, Wittchen [/bib_ref]. This interview was chosen as there was no SKID interview available for DSM-5 at the time of the investigation. Alcohol consumption during the last 3 months was recorded using the Form 90 interview [bib_ref] Reliability and validity of the form 90 interview, Scheurich [/bib_ref] , as well as nicotine status and consumption during the last 3 months. Psychometric assessment comprised the Fagerstrøm Test of Nicotine Dependence (FTND; [bib_ref] The Fagerstrom test for nicotine dependence: a revision of the Fagerstrom Tolerance..., Heatherton [/bib_ref] , the Alcohol Dependence Scale (ADS; [bib_ref] Alcohol dependence syndrome: measurement and validation, Skinner [/bib_ref] , the Inventory of Drinking Situations (IDS; [bib_ref] Inventory of drinking situations (IDS): user's guide. Addiction Research Foundation, Annis [/bib_ref] , the Beck Depression Inventory (BDI; [bib_ref] An inventory for measuring depression, Beck [/bib_ref] , and the Alcohol Use Disorders Identification Test (AUDIT; [bib_ref] The Alcohol Use Disorders Identification Test (AUDIT): a review of recent research, Reinert [/bib_ref] , recorded via an electronic platform (Social Science Survey, www.soscisurvey.de).
At the investigational day, participants underwent a medical check-up for occurrence of somatic illnesses representing a contraindication for assessment or nalmefene intake. An alcohol breath test was performed to ensure abstinence. Current withdrawal symptoms were captured with the Clinical Institute Withdrawal Assessment for alcohol scale (CIWA-Ar; [bib_ref] Assessment of alcohol withdrawal: the revised clinical institute withdrawal assessment for alcohol..., Sullivan [/bib_ref]. After assuring the absence of any exclusion criteria valid at MRI investigation, study medication was handed out to the participant, and intake was supervised. Alcohol consumption since baseline measurement or investigational day one respectively was recorded using the Form 90 interview [bib_ref] Reliability and validity of the form 90 interview, Scheurich [/bib_ref]. A second investigational day was scheduled at a 1-week interval and differed only in the double-blind, randomized, oral administration of nalmefene (18 mg) or placebo 2 h before the fMRI measurement. These timeframes were determined on the basis of the pharmacokinetic parameters for nalmefene, namely a time to peak concentration of 0.8 h (median) and opioid receptor occupancy up to 74 h after oral administration of (20 mg) nalmefene [bib_ref] Prolonged central mu-opioid receptor occupancy n g, Ingman [/bib_ref]. Participants were then discharged for a recreational period until a 2-h period after medication intake was fulfilled to ensure substance invasion. Before fMRI measurement, participants' severity of craving was captured via paper-pencil assessment applying the Alcohol Urge Questionnaire (AUQ, 8 items on a seven-point rating scale; [bib_ref] Development and initial validation of a measure of drinking urges in abstinent..., Bohn [/bib_ref] and the Alcohol Craving Questionnaire (ACQ, 30 items on a seven-point rating scale; German version, [bib_ref] The assessment of craving: psychometric properties, factor structure and a revised version..., Raabe [/bib_ref]. Subsequently, participants underwent fMRI measurement comprising (1) our alcohol cue-reactivity task [bib_ref] Initial, habitual and compulsive alcohol use is characterized by a shift of..., Vollstädt-Klein [/bib_ref] and (2) an emotional faces task, which has been analyzed elsewhere [bib_ref] Initial, habitual and compulsive alcohol use is characterized by a shift of..., Vollstädt-Klein [/bib_ref]. Directly afterwards the fMRI scan, the questionnaires (AUQ, ACQ) were applied one more time. After a quick medical check-up by a physician, participants were dismissed. Test for successful blinding was conducted by asking the participants about their estimation for the time point of the medication administration (56% did not make any estimate, 39% correctly identified it, and 6% made an incorrect estimate).
## Alcohol cue-reactivity task
For the assessment of neural response to alcohol-related stimuli, a cue-reactivity task was used. In this task, 60 alcohol-related and 45 neutral stimuli were presented using a blocked design with five stimuli each block. Each image was presented for 4 s, so each block took 20 s. Alcohol-related pictures were taken from a validated picture series [bib_ref] Initial, habitual and compulsive alcohol use is characterized by a shift of..., Vollstädt-Klein [/bib_ref] , and neutral control cues were taken from the International Affective Picture System. Following each block, the participants were asked for the current intensity of their alcohol craving on a visual analogue scale (VAS) ranging from 0 (no craving) to 100 (extremely extensive craving). Task duration was 12 min.
## Functional mri acquisition
Functional and anatomical brain images were acquired using a 3T whole-body tomograph (MAGNETOM Trio, Siemens Medical Systems, Erlangen, Germany). Task-related blood oxygen level-dependent (BOLD) response was measured using T2*-weighted echo planar imaging (EPI) sequences (TR = 2.41 s, TE = 25 ms, flip angle = 80°, 42 slices, slice thickness 2 mm, 1 mm gap, voxel dimensions 3 × 3 × 3 mm 3 , FOV 192 × 192 mm 2 , 64 × 64 in-plane resolution). The T2*weighted EPI sequences were acquired in a transversal orientation 30°clockwise to AC-PC-line covering the whole brain. This short TE and the 30°flip to AC-PC orientation were chosen to minimize susceptibility artifacts. The number of images measured at the ALCUE Task for each participant was 305.
In addition, high-resolution anatomical scans using T1weighted 3-D magnetization-prepared rapid acquisition gradient-echo (MPRAGE) sequences consisting of 192 sagittal slices (slice thickness 1 mm, 1 × 1 × 1 mm voxel size, FOV 256 × 256 mm 2 , TR = 2300 ms, TE = 3.03 ms, TI = 900 ms, flip angle = 9°) were acquired for each participant.
fMRI pre-processing Pre-processing and statistical analyses of brain imaging data were performed using SPM8 (Wellcome Department of Cognitive Neurology, London, UK). The first five scans were excluded from the analyses to avoid artifacts due to magnetic saturation effects. The remaining scans were realigned spatially to correct for head motion over the course of the session and then normalized to an MNI (Montreal Neurological Institute, Quebec, Canada) EPI template. Subsequent smoothing was performed using an isotropic Gaussian kernel for group analysis (8 mm full width at half maximum .
# Statistical analysis
Statistical analyses of the pre-processed fMRI data on the first (individual) level were performed by modeling the different conditions (alcohol-associated versus neutral control stimuli; boxcar functions convoluted with the hemodynamic response function) as explanatory variables within the context of the general linear model (GLM) on a voxel-by-voxel basis with SPM8. Realignment parameters were included as regressors of no interest. The following contrast images were calculated for each participant and each drug condition: (1) favorite drink > neutral; (2) neutral > favorite drink (e.g., [1 0 0 -1] for beer drinkers and [1 1 0 -2] for beer and wine drinkers).
Individual contrast images described above of the participants were included in a second level analysis (full factorial model) to identify the main effects as well as differences between the verum and the placebo condition. The sequence of drug administration (placebo first/nalmefene first) was considered as a covariate of no interest in the analysis. We conducted a region of interest (ROI) analysis because of the strong a priori hypotheses previously defined in the study protocol (clinicaltrials.gov, NCT02372318) of cue-induced activation of the ventral (VS) and the dorsal striatum (DS) especially due to previous findings of reduce brain activity in mesolimbic regions after intake of opiate antagonists and cue exposure [bib_ref] Effect of naltrexone and ondansetron on alcohol cue-induced activation of the ventral..., Myrick [/bib_ref] [bib_ref] Nalmefene reduces reward anticipation in alcohol dependence-an experimental fMRI study, Quelch [/bib_ref]. The ROI mask for the ventral striatum was created by placing two 10 mm spheres bilaterally on the MNI coordinates [+ − 12, 8, − 8]. These coordinates were determined using a term-based metaanalyses search at the platform www.neurosynth.org. For the dorsal striatum, a self-created and already established mask of Vollstädt-Klein et al. (2010) was used (see [fig_ref] Figure 2: Masks for region of interest [/fig_ref]. For the ROI analyses, FWEcorrected p-values are reported at cluster level.
In addition to the ROI analysis, we conducted a whole brain analysis for exploratory analysis. To control for multiple statistical testing, the probability of a family wise error (FWE) was set to .05. For this purpose, we used the AlphaSim (3dClustSim) method. A voxel wise threshold of p < .02 was combined with a cluster extent threshold of 795 (placebo, contrast favorite drink > neutral), 882 (nalmefene, contrast favorite drink > neutral), and 841 voxels (placebo > nalmefene, contrast favorite drink > neutral), determined by the AlphaSim using 25000 Monte Carlo simulations for the whole brain analysis of the ALCUE task. Estimation of smoothness based on the residual images was conducted using SPM by taking the maximum of the 3 estimated parameters in x, y, and z directions.
The self-reported craving was analyzed in SPSS (Statistical Package of the Social Sciences, version 24; SPSS; IBM Corporation, Armonk, NY, USA) using paired t-tests.
# Results
Twenty-three participants were randomized to one of the two groups; one group was administered nalmefene at the first time point, and in the other one, nalmefene was given at the second time of measurement. Of the 23 randomized participants, ten participants (44%) reported adverse side effects (five participants of group 1 and five participants of group 2). A mean of 5.8 (SD = 2.7) symptoms occurred and lasted a mean time of 37.5 h (SD = 22.3 h). For a detailed overview of the adverse side effects that occurred, see Supplement [fig_ref] Table 1: Effects of single-dose nalmefene on cue-reactivity and craving in alcohol use disorder... [/fig_ref] , Reported side effects. Five individuals had to be excluded from the analyses because of either withdrawal of informed consent (in most cases due to adverse side effects of nalmefene, e.g., insomnia, vertigo, and nausea, n = 4 from group 1) or unsuitability for fMRI scanning (n = 1, metal implant). Therefore, the final sample consisted of 18 participants who finished the whole experimental procedure. Because of technical problems (n = 1), or too much head movement (n = 2), only 15 participants were included in the final fMRI-data analysis. However, the sample of the 18 participants was used to evaluate the behavioral data. Of these 18 participants, 12 were male (67%), and 10 of them were smokers (59%). The mean age was 50.3 years (SD = 13.9 years), and a mean of 6.4 (SD = 1.4) DSM-5 criteria for alcohol dependence was met. On average, participants consumed 6.4 (SD = 4.3) standardized alcoholic beverages (12 g each per drink) per day, for a mean amount of 76.9 (SD = 52) grams per day. In the 90 days before the baseline examination, participants had an average of 54 (SD = 31.6) heavy drinking days. On the ADS, participants scored an average of 8.7 (SD = 4.4). For the extent of harmful use and alcohol dependence (AUDIT), the mean score was 17.2 (SD = 5.8). For detailed demographics information of these 18 respectively and 15 participants, see [fig_ref] Table 1: Effects of single-dose nalmefene on cue-reactivity and craving in alcohol use disorder... [/fig_ref].
# Fmri results
Overall, data of 15 heavy drinkers was compared between the conditions nalmefene and placebo. There was a significant lower neural cue-reactivity (favorite drinks > neutral) in the ventral striatum (a priori ROI) in the nalmefene condition compared to the placebo condition [(x,y,z) = (− 4, 4, − 8), t = 6.20, p FWE corrected = .007; (x,y,z) = (4, 4, − 10), t = 4.50, p FWE corrected = .032]; see . Such a difference could not be observed for the dorsal striatum. For the ventral striatum, the actual data demonstrated an effect size of d = 0.74 resulting in a power of 86%. Aside from that, there was no main effect for the contrast alcohol (fav) > neutral in the ROI analyses for placebo or nalmefene condition.
However, the explorative whole brain analysis (i.e., non-ROI-based approach) revealed a broadly activated network (for the contrast favorite drink > neutral) including striatal areas, limbic regions (hippocampus and anterior cingulate), and inferior and middle frontal gyrus under placebo, but not under the nalmefene condition (see for a detailed overview Supplement material and Supplement .
In addition, the exploratory whole brain analysis for the direct comparison placebo vs. nalmefene showed an interaction effect, driven by a decreased activation under nalmefene (i.e., placebo > nalmefene and favorite drink > neutral) in . There was no significant increase in brain activation in any brain region by nalmefene.
## Subjective craving data (behavioral data)
No significant difference in pre-scanning AUQ or ACQ craving between the nalmefene and placebo condition was present. However, participants reported significantly lower cueinduced craving (measured by AUQ) directly after the ALCUE fMRI task in the nalmefene compared to the placebo condition ( Regarding the alcohol craving measured during the cuereactivity task using a VAS ranging from 0 (no craving) to 100 (extremely extensive craving), we found no significant differences between the drug conditions neither for the craving after favorite alcohol cues (t = 0.65, p = 0.53; mean ± SD placebo: 26.46 ± 29.51 and nalmefene: 22.45 ± 28.45) nor after neutral cues (t = − 1.01, p = 0.33; mean ± SD placebo: 08.83 ± 14.98 and nalmefene: 10.96 ± 12.98). Also the difference score (favorite alcohol cues − neutral cues) did not differ significantly under either the placebo (mean ± SD: 17.63 ± 21.17) or the nalmefene (mean ± SD: 11.49 ± 20.87) condition, t = 1.04, p = 0.32.
Linear correlations between questionnaires and VS and DS activations and linear correlations between difference scores (verum versus placebo condition) were examined. There were no significant correlations between (changes in) neural cuereactivity responses and (changes in) craving.
# Discussion
To the best of our knowledge, we are the first to show that a single-dose nalmefene reduces the neuronal response to alcohol associated stimuli in the ventral striatum. Nalmefene seems to influence neuronal brain responses, responsible for reward-associated behavior [bib_ref] A neural substrate of prediction and reward, Schultz [/bib_ref] [bib_ref] Dysfunction of reward processing correlates with alcohol craving in detoxified alcoholics, Wrase [/bib_ref] and may affect substance-related behavior consequently. Such an effect was not revealed for the dorsal striatum. In addition, we could show a reduced subjectively reported craving measured by AUQ after a cue-reactivity task under nalmefene compared to placebo.
This finding expands the results of a first fMRI study by [bib_ref] Nalmefene reduces reward anticipation in alcohol dependence-an experimental fMRI study, Quelch [/bib_ref] who demonstrated similar to the present investigation decreased brain activation in the mesolimbic system under the influence of nalmefene during reward anticipation. In contrast to the present study, [bib_ref] Nalmefene reduces reward anticipation in alcohol dependence-an experimental fMRI study, Quelch [/bib_ref] used a monetary incentive delay task during an intravenous alcohol challenge. We on the other hand administered nalmefene 2 h before the presentation of alcohol-related stimuli respectively alcohol consumption according to a preventive approach as recommended in the European public assessment report (EPAR) for Selincro® (Online document. Online document. European Medicines Agency [EMA] 2013). This naturalistic approach reflects the use of nalmefene as medication for the reduction of alcohol consumption during confrontation with alcohol-associated stimuli. Results of ROI analysis for ventral striatum; p < .001 (uncorr), 10 voxel, contrast: placebo > nalmefene, alcohol (favorite) > neutral Comparison sum score of AUQ after fMRI cue-reactivity task between nalmefene and placebo Furthermore, one possible explanatory approach for the reduced brain activation in the ventral striatum but not in the dorsal striatum after the administration of nalmefene could be an earlier stage of alcohol use disorder in the examined sample. The participants scored an average of 8.7 (SD = 4.4) on the ADS [bib_ref] Alcohol dependence syndrome: measurement and validation, Skinner [/bib_ref]. This mean score corresponds to the cut off proposed by [bib_ref] Diagnostic validity of the MAST and the alcohol dependence scale in the..., Ross [/bib_ref] indicating alcohol dependence. For the extent of harmful use and alcohol dependence (AUDIT; [bib_ref] The Alcohol Use Disorders Identification Test (AUDIT): a review of recent research, Reinert [/bib_ref] , the mean score was 17.2 (SD = 5.8), indicating a high level of alcohol-related problems (cut-off > 16, [bib_ref] AUDIT. The alcohol use disorders identification test. Guidelines for Use in Primary..., Babor [/bib_ref]. It has been hypothesized that a "ventro-dorsal shift" characterizes the switch from "choice to habit" on a neural basis mirroring the progression of the alcohol dependence [bib_ref] Neural systems of reinforcement for drug addiction: from actions to habits to..., Everitt [/bib_ref] [bib_ref] Drug addiction: updating actions to habits to compulsions ten years on, Everitt [/bib_ref] [bib_ref] Initial, habitual and compulsive alcohol use is characterized by a shift of..., Vollstädt-Klein [/bib_ref]. In the present study, all participants met DSM-5 criteria for AUD. However, none of the participants received any previous (semi-) inpatient detoxification treatment. In addition, in the present sample, the mean ADS score and an average of 81 g (SD = 55.3) of consumed pure alcohol per day may indicate that the investigated individuals are "between" light and heavy drinkers and therefore are not yet fully habitual-but rather reward-motivated. This might suggest that especially individuals in early stages of alcohol use disorder could benefit from an add-on therapy with nalmefene. Thus, the results found here suggest that nalmefene could be an agent supporting regaining control over substance use, especially at an early stage of substance use disorder. This approach to regaining control in early stages of substance use disorders (SUD) needs further investigation because it has not yet been sufficiently examined [bib_ref] Addiction research consortium: losing and regaining control over drug intake (ReCoDe)-from trajectories..., Heinz [/bib_ref].
In a further analysis of Vollstädt-Klein et al. (2019), we investigated the effect of nalmefene on neural activation using fMRI during the presentation of emotional faces pictures in this sample. Results of this suggest that nalmefene is able to influence neuronal processes responsible for social skills and empathy. According to this, nalmefene could have additional useful effects, in addition to its anti-craving effect, especially in the social context, which in turn could support the reduction in amount of drinking [bib_ref] Initial, habitual and compulsive alcohol use is characterized by a shift of..., Vollstädt-Klein [/bib_ref].
Intriguingly and complementing the above described findings, previous studies have shown that increased activation of the ventral striatum by alcohol-associated cues was associated with an increased risk of relapse in alcohol-dependent individuals. This also suggests that cue-induced activation of ventral striatum might increase the risk of relapse [bib_ref] Alcohol-associated stimuli activate the ventral striatum in abstinent alcoholics, Braus [/bib_ref] [bib_ref] Cue-induced activation of the striatum and medial prefrontal cortex is associated with..., Grüsser [/bib_ref] [bib_ref] Identifying the neural circuitry of alcohol craving and relapse vulnerability, Heinz [/bib_ref]. Against this backdrop, it has to be emphasized that a reduction of the cue-induced brain activation in the ventral striatum through nalmefene might reduce the risk of relapsing into severe drinking patterns.
For the structurally similar opiate antagonist naltrexone, a reduced neuronal activity on alcohol stimuli in the ventral striatum could be demonstrated by fMRI investigation likewise [bib_ref] Effect of naltrexone and ondansetron on alcohol cue-induced activation of the ventral..., Myrick [/bib_ref] which is also expressed on the behavioral level. Treatment with naltrexone reduces both subjectively craving and the amount of consumed alcohol compared to placebo treatment.
Moreover, in the present study, a significantly lower selfreported craving, as measured with the AUQ, after the fMRI scan was revealed in the nalmefene condition, compared to placebo. This finding could be based on the fact that the AUQ is useful to capture substance cravings before and after a cue-reactivity task [bib_ref] Factor structure of the alcohol urge questionnaire under neutral conditions and during..., Mackillop [/bib_ref] , and that this questionnaire is a quick self-report instrument (< 1 min) to assess craving with high internal consistency and high test-retest reliability [bib_ref] Assessing craving for alcohol, Drobes [/bib_ref]. In addition, the AUQ was shown to correlate significantly positively with alcohol dependence severity and with scores on the Obsessive Compulsive Drinking Scale indicates construct validity [bib_ref] Development and initial validation of a measure of drinking urges in abstinent..., Bohn [/bib_ref]. reported significantly reduced alcohol craving as measured by AUQ after naltrexone administration which was also associated with kappa opioid receptor (KOR) availability. This is also in line with our finding, due to a partial agonistic activity at the kappa opioid receptor of nalmefene. Taking together, the reported finding might indicate that nalmefene is able to reduce craving and thus the hedonic effect of alcohol. The particular strength of the within-subjects study design is that each participant serves as their own control and thus reduces the error variance.
On the other hand, not all craving measurements showed significant reduction by nalmefene. An explanation for low and not significant different craving measured by ACQ and the fMRI task between the two conditions may also be an effect of social desirability. Even if direct questioning about current substance demand is a common approach, this can be distorted by social desirability. Additionally, many participants reported that the investigation took place at a time (in most cases, during the day) that was outside of their usual time to drink (in the evening). This aspect in combination with the fact that the participants were in an examination situation could have additionally influenced the self-reported craving. In addition, the ACQ measures the aspect of loss of control relatively insufficiently, so that a multidimensional craving assessment is recommended by combining it with other measures such as the OCDS [bib_ref] The assessment of craving: psychometric properties, factor structure and a revised version..., Raabe [/bib_ref]. Future studies could use additional (indirect) measurements of physiological outcomes, such as skin conductance or heart rate, to support the measurement of craving [bib_ref] Assessing craving for alcohol, Drobes [/bib_ref]. Moreover, this could be useful as there is evidence that not all individuals can consciously perceive their substance craving [bib_ref] A cognitive processing model of alcohol craving and compulsive alcohol use, Tiffany [/bib_ref] , which could be problematic when using self-report measures.
Unexpectedly, there was no main effect for the type of stimulus in the ROI analysis, i.e., there was no increased striatal activation during favorite alcoholic stimuli under the nalmefene or the placebo condition. However, the explorative whole brain analysis with a more liberal threshold indicated that alcohol stimuli induced increased activation compared to neutral stimuli in the ventral and dorsal striatum. Here, also other relevant brain regions implicated in cue and reward processing as middle frontal gyri and anterior cingulate cortex [bib_ref] Extended-release naltrexone (XR-NTX) attenuates brain responses to alcohol cues in alcohol-dependent volunteers:..., Lukas [/bib_ref] [bib_ref] Functional neuroimaging studies of alcohol cue reactivity: a quantitative meta-analysis and systematic..., Schacht [/bib_ref] were shown to be increased in the whole brain analysis under placebo and reduced by nalmefene.
In addition to the lack of observable main effects in the ROI analysis for the type of stimulus, the small sample size should be mentioned as a further weakness of this study. On this account, small effects may have been lost. In addition, the type of stimuli may not be suitable for every participant to provoke cue-reactivity [bib_ref] Smoking stimuli from the terminal phase of cigarette consumption may not be..., Mucha [/bib_ref]. Although we analyzed the contrast "favorite drink > neutral," the picture stimulus set might not have been optimal for each participant.
Another limitation in this specific sample is an unequal distribution of participants between the two groups (see [fig_ref] Figure 1: Study procedures [/fig_ref] and group 1: n = 5, group 2: n = 10). This is partly explained by dropout due to adverse side effects (e.g., insomnia, vertigo, and nausea; for detailed description, see Supplement [fig_ref] Table 1: Effects of single-dose nalmefene on cue-reactivity and craving in alcohol use disorder... [/fig_ref] , Reported side effects). In the group of participants who received the nalmefene at the first time point (group 1), four participants dropped out because of adverse side effects, while in group 2 (placebo T1, nalmefene T2), no dropout after the first time point occurred (see [fig_ref] Figure 1: Study procedures [/fig_ref] CONSORT flow diagram). Therefore, drop-out might be affected by medication, which could limit the results as an influence of the sequence of drug administration.
Nevertheless, further research is needed to assess long-term effects of nalmefene on neuronal response since even nonmedical interventions like cue-exposure therapy led to reductions in cue-reactivity [bib_ref] Cue exposure therapy for the treatment of alcohol use disorders: a meta-analytic..., Mellentin [/bib_ref].
Taken together, nalmefene attenuated alcohol cuereactivity in the ventral striatum of non-treatmentseeking AUD patients compared to placebo and reduced subjective alcohol craving in our study, supporting the anti-craving effect of this drug. So our findings support the assumption that nalmefene might reduce the amount of drinking via attenuation of neural alcohol cuereactivity.
As previous studies suggested that especially individuals with a high cue-reactivity in the ventral striatum [bib_ref] Predicting naltrexone response in alcohol-dependent patients: the contribution of functional magnetic resonance..., Mann [/bib_ref] might particularly benefit from treatment with opioid-antagonists, future studies could establish predictors for successful treatment with nalmefene.
Author contribution SVK, KM, and FK were responsible for the study design. DK and CD contributed to the acquisition of fMRI and psychometric data. DK and SVK performed the data analysis. DK, SVK, PB, and JMB interpreted the data. DK and SVK drafted the manuscript. AK and DH contributed to the recruitment of patients and conducted physical examinations. All authors revised the manuscript critically for important intellectual content and approved the final version.
Funding Open Access funding enabled and organized by Projekt DEAL. The original study was supported by Lundbeck A/S, Denmark. The project was supported in part by grants from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) -TRR 265 Project-ID 402170461 -Project ID-421888313, Project-ID 437718741 and GRK2350-1 Project-ID 324164820.
# Declarations
Ethical statement All study procedures were in accordance with the Declaration of Helsinki. The study was approved by the ethics committee of the University of Heidelberg (Ethics Commission II: 2014-607N-MA).
## Conflict of interest
Outside the submitted work, Derik Hermann received honoraria for participating in advisory boards of the pharmaceutical companies Indivior, Camurus, and Servier. Falk Kiefer received honoraria as a consultant for Amomed, Desitin, Indivior, Lundbeck und Otsuka. We do not have further commercial or financial involvements that might present an appearance of a conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
[fig] Figure 1: Study procedures [/fig]
[fig] Figure 2: Masks for region of interest (ROI) analysis, ventral striatum (red) two 10 mm spheres bilaterally on the MNI coordinates [+ − 12, 8, − 8]; the dorsal striatum (blue) self-created and already established mask of Vollstädt-Klein et al. (2010) [/fig]
[table] Table 1: Effects of single-dose nalmefene on cue-reactivity and craving in alcohol use disorder Notes. M mean, SD standard deviation, ADS Alcohol Dependence Scale, AUDIT Alcohol Use Disorders Identification Test, BDI Beck Depressions Inventory, FTND Fagerström Test for Nicotine Dependence a 12 g per drink [/table]
|
West Riding Asylum Reports
1873.] West Riding Asylum Reports.
IV.--West Riding Asylum Reports.^ We devoted considerable space in our number for January, 1872, to an analytical review of the first volume of the ' West Riding Lunatic Asylum Reports/ inasmuch asit was a new adventure and a novelty as a literary production. Asylum reports of the usual stamp have, indeed, no novelty, but are "as plenty as blackberries," and like this common-place fruit get little consideration. They are the natural outcome of the painfully elaborate legislation applied to lunatic asylums, and are presumably addressed to the ratepayers of the county, who are supposed to be interested in the details of the sugar and the soap consumed in the establishment, and of other domestic items of expenditure. Our impression, however, is, that the ratepayers' interests would be better consulted by saving the cost of a mass of most unedifying and useless printing which is scarcely ever read even by members of the visiting committees. They contain, indeed, a medical report also, but, as a rule, few are the grains of knowledge and experience to be gathered therefrom; for the medical superintendents who write such reports have no sufficient inducement to make them what they might and should be?the records of work done, of observations made, and of lessons in management and treatment acquired. These gentlemen feel that a long medical report, worthy of them as professional men earnest in the pursuit and advancement of their speciality, would be out of place in a pamphlet addressed to non-medical readers, in which, too, the minutice of the domestic economy of the establishment is a prominent feature, and the excellent and prudent administration of the committee a principal theme.
Entertaining these convictions we were, therefore, pleased to bring prominently under the notice of our readers the praiseworthy enterprise of Dr. Crichton Browne to record the medical work done at the large West Riding Asylum in a form likely to be useful to the profession and of lasting value. On the same grounds Ave are now glad to call attention to this second volume, and to find that Dr. Browne has been sufficiently encouraged to repeat his venture.
The volume consists of thirteen essays, each written, with two exceptions, by gentlemen who are, or who have been, members of the medical staff of the asylum. Now, in noting the authorship of the several papers and the general character of those papers, we are struck by a too evident tendencv to depart from the announced principle and object of the volume, viz. to make it a record of the medical and scientific work accomplished in the institution. In place of this it approaches too nearly to a collection of essays to which the experience of the asylum contributes either more or less, or even nothing at all. We would warn Dr. Browne against this tendency. It is not that we fail to recognise the value of the essays submitted to us, even when destitute of teachings immediately derived from the science and practice to be encountered within the wards of the asylum in question ; but by publishing such, and also by serving up theoretical rather than practical disquisitions, the editor surrenders what should be the distinctive feature of his volume, and places that publication in rivalry with journals whose business it is to give circulation to such productions.
If the results of actual observation and research within the walls of the asylum are insufficient to make a book of the dimensions now before us yearly, the editor should set smaller limits to his ambition, and content himself with fewer pages.
It is time to notice the matter of the several essays before us, but our analysis must be brief. The first one is by Dr. Burman, who has experimented on the value and effects of the subcutaneous injection of conia. The conclusions he has arrived at are?1. That conia is too powerful and too irritant to be administered internally alone; but, when neutralised with acid and in bland solution, it may be so used and will produce cicutism without topical irritation. 2. Pure conia may be injected under the skin without other effects than those of local irritation at the site of injection. 3. Conia, neutralised with acetic or hydrochloric acid, and dissolved in spirit and water, acts very rapidly and powerfully. In doses of Hiss to 1H. ij in the healthy human subject, it produces well-marked cicutism. 4. Thus administered therapeutically it subdues the motor excitement of mania, wards off emaciation and exhaustion, and promotes recovery. Commencing with doses of 1)1. the conia may be gradually increased in proportion to the motor activity of the patient until decided physiological effects are produced. 5. So used the digestive functions and the circulation are not interfered with and no considerable local irritation follows. 6.
The most suitable cases for the drug are those of acute mania, without organic brain lesion and where medicine, if given by the mouth, would require to be administered by the stomachpump. 7. As conia acts as a sedative on the motor centres, and morphia on the sensori-motor and ideo-motor centres, these two substances may be used together, as concurring to allay maniacal excitement.
8. Conia might be very useful as a subcutaneous injection in cases of poisoning by strychnia, as well as in tetanus, hydrophobia and other spasmodic diseases. 9. Specimens of conia obtained from different sources vary much in strength and appearance; hence the propriety of first testing the specimen to be employed. 10. The best and purest is prepared from the seeds of the uncultivated plant. 11. niss of the best conia subcutaneously injected is equivalent in action to about fl. 3j of the best succus conii administered by the mouth.
Mr. Major in the next article records the changes in the minute structure of the cortical substance of the brain in a case of chronic brain wasting, and in a subsequent paper describes a simple instrument for determining the depth of the grey matter of the convolutions. In the article first named he takes occasion to call attention to the structural variety obtaining in almost every segment of the brain, so that what may be predicated as normal or abnormal in one part is not so in another, even within the limits of the same convolution. This circumstance adds immensely to the difficulty of determining and describing changes in brain matter, and renders an attempt to give a satisfactory or exhaustive account of a post-mortem examination a work of so great labour that we cannot often look for it. However, Mr. Major has had the industry and perseverance to examine microscopically a diseased brain, and to compare it with a healthy one " to the extent of the frontal, parietal, and occipital lobes in each, by means of numerous sections," noting the appearance of each layer in succession. In so doing he has followed the method of preparation recommended by Dr. Lockhart Clarke, and has observed the effects of staining with carmine. We cannot enumerate the changes detected in the morbid brain here described, as to do so would involve the reproduction of Mr. Burman's paper, but may state generally that distinct alterations were remarked in the number and disposition of the nerve-cells, in the size, configuration and processes of those cells, in their granular and nucleated appearance, and also in the walls of the minute blood-vessels.
The essay following is " On Menstrual Irregularities," by Dr. Henry Sutherland. After some general observations he advances and then illustrates seven principal conclusions derived from notes of upwards of 500 inmates of the asylum. These conclusions are that? "1. In idiocy and cretinism puberty is delayed or absent. 2. In epileptic insanity the fits generally increased in number, and are the patients frequently become excited at the catamenial period. 3. That in mania exacerbations of excitement usually occur at the menstrual period, and that a state of intense excitement is almost continuous in patients suffering from menorrliagia. 4. That in Reviews.
[April, melancholia a large proportion of patients suffer from amenorrhoea. 5. That in dementia the patients usually menstruate in a normal, healthy manner. G. That in general paralysis the change of life frequently occurs early. 7. That very rarely the catamenia reappear in aged insane women after a prolonged cessation."
As Dr. Sutherland justly remarks, " some of tlie above conclusions have already been arrived at by various independent authorsit is well, however, to have such confirmed.
Dr. Samuel Mitchell, now the Medical Superintendent of the South Yorkshire Asylum, follows with an account of some experiments conducted by himself with a mixture of ether and nitrous oxide, but he fails to summarise his results, except so far as to state that its inhalation, even when the profoundest anaesthesia has been produced, has not been followed by tlie lividity nor by the convulsive twitchings seen when nitrous oxide alone is administered, and that no vomiting ensues. He also mixed chloroform with nitrous oxide and finds that, as in the case of the mixture with ether, the sense of suffocation induced by ether and chloroform given alone does not occur, and that there is a prolonged feeling of numbness in the extremities not present when nitrous oxide is given by itself.
But the more interesting part of his paper is that occupied with an examination of the theories of the causation of anaesthesia and of the rationale of stimulation. A very excellent essay " On the Relations between Cranial Injuries and Mental Diseases," illustrated by some few cases, follows from the able pen of the editor, Dr. Crichton Browne, but its extent and character forbid analysis in this place. Mr.
Pedler's paper " On Puerperal Mania" presents no notable features; what value it has is derived from statistics of cases of the disorder obtained from the case-books of the asylum. Mr. Major in the next article describes an instrument, before referred to, for determining the depth of the grey matter of the cerebral convolutions, which is known to vary much in the same brain in different situations and in different brains in corresponding situations. The instrument is simple enough, being nothing more than a graduated glass tube, of equal calibre throughout, but bevelled on the outside at one extremity so as to give it the sharpness necessary for its ready penetration through the brain substance. The thickness of the grey lamina may at once be read off by means of the degrees marked on the tube. The inventor is very precise in his account of the way of using the instrument, and those who desire to test it will do well to refer to that account. The paper is accompanied by some tables of results and general conclusions.
" The Mental Symptoms of Ordinary Disease" is a general but which, except in the way of reference to a few cases that have happened, as we suppose, during the writer's former connection with the asylum, has no direct association with the experience or with the work of that institution. It is, nevertheless, a well written and instructive essay, but not one of original and novel conception.
"The Electric Treatment of the Insane" is a paper by our able contributor, Dr. Clifford Allbutt, of Leeds, based upon experiments made among the patients of the West Riding Asylum. It is, therefore, a genuine record of work carried out there, although by a gentleman not connected with the establishment. Its lesson is, that electricity is of little avail as a curative agent in insanity. A marked improvement followed in acute primary dementia; a less degree of good was found in mania, atonic melancholia, and, perhaps, in recent secondary dementia; no change occurred in chronic dementia and some cases of melancholia, and unfavorable results happened in hypochondriacal melancholia, and, perhaps, in brain-wasting.
Ophthalmoscopic observations made in cases of general paralysis, and after the administration of certain toxic agents, are placed on record by Mr. C. Aldridge, and will be read with interest and instruction. They present a continuation of the like researches made in the wards of the same asylum, by Dr. Clifford Allbutt, the general results of which are condensed in that writer's work on " The Use of the Ophthalmoscope in Diseases of the Nervous System." " Most of his observations were made upon patients in this asylum," writes Mr. Aldridge, "but as they occurred fouryears ago, I have not been able to find one patient alive who was then examined." This statement is in itself curious as a commentary 011 the rapid fatality of general paralysis. The circumstance it represents also was fatal to the wish Mr. Aldridge had to verify Dr. Allbutt's facts, and to report any additional ones with regard to the self-same cases. The affection of the optic nerve detected is a " descending neuritis," with first increased vascularity and engorgement, subsequently replaced by absorption of the exudation and extending whiteness of the disc, and with, in most cases, atrophy. Mr. Aldridge is unable to confirm Dr. Allbutt's opinion that there is ataxy of the ocular muscles, the phenomena suggestive of that condition being due to inability to arrest or sustain the attention by reason of the excitement or fatuity of the patients. In twenty-three of the forty-three paralytics examined there was inequality of the pupils, a circumstance known to the profes-Reviews.
[April, sion for many years as common in such patients, and now shown to depend on atrophy of the optic disc, the atrophy being most advanced in the eye where the pupil is largest.
The ophthalmoscopic condition of the optic discs as above determined has a certain pathological value, but no diagnostic importance, inasmuch as, so far as observation has yet extended, it is one that becomes manifested only when the second stage of general paralysis is arrived at, and when the general symptoms of the malady have become unmistakeable. One remarkable fact, however, emerges from the observations recorded, viz. that atrophic and white optic discs may resume gradually their capillary tint with recovery of function. This is, as Mr.
Aldridge expresses it, a hopeful fact.
A valuable supplementary portion of the paper under notice is to be found in the researches of Mr. Aldridge respecting the operation of various toxic agents on the fundus oculi. In the first volume of these f. West Riding Reports' the same industrious and able observer noted the changes produced after the administration of chloral, ergot, nitrite of amyl, and nitrous oxide. He has since experimented with belladonna, hyoscyamus, and picrotoxine, and with laburnum. The three first mentioned induce hyperaunia of the retina; the last named, together with ergot, causes ancemia.
The history of his experiments and observations is preceded by an interesting discussion of the hypothesis that the changes observed by the ophthalmoscope in the vessels within the eye are to be regarded as signs of analogous events taking place within the skull, and that the drugs producing such changes exercise the like effects in the vascular system of the brain. His next attempt is to solve the question whether the changes discoverable have their origin in the nervous or in the vascular system. The conclusion Mr. Aldridge appears to arrive at is, that the origin may be in either one or the other, and that it differs in relation with the toxic agent employed. We commend this disquisition to our readers, as it opens up many important questions in physiology and pathology, and, we may add, in therapeutics.
We are very pleased to find a champion in Mr. Courtenay of the virtues of opium in the treatment of the insane, which have on theoretical grounds been of late years so mercilessly assailed. We consider this young physician has fully re-established the value of opium in melancholia. Apart from theoretical considerations and speculations which are always ready to hand to ingenious minds, and, owing to the imperfections of physiological science, are not easily disposed of, the decriers of opium in the treatment of insanity had little else to rely upon in their assault than some experiments conducted by Dr. Clouston a few years since at the Cumberland Asylum. Mr. Courtenay well shows the faultiness and inconclusiveness of these experiments, and then proceeds to examine the physiological effects of opium and the pathological condition of the nervous system, as far as yet known in melancholia. This done he details the results of treatment by opium in forty-nine cases of the disorder, and disproves the assertions made of the damaging effects of that drug on the weight, temperature, pulse, and appetite of patients. One cause of opium getting a bad name he shows to be in the unnecessarily large doses in which it has mostly been administered, and he sums up his article in the following sentence:?" No drug at present known possesses actions which meet so well the different symptoms of melancholia/'
The next essay, on " Impairment of Language the result of Cerebral Disease/' is from the pen of the editor's father, the much respected Dr. W. A. F. Browne, who for many years was the energetic physician of the Crichton Asylum, Dumfries, and subsequently a commissioner in lunacy for Scotland. Knowing his sad affliction, the loss of sight, it is gratifying to find him still an earnest labourer for the advancement of science.
The essay in question exhibits much thought and power of analysis. It is not merely a summary of the facts and doctrines of aphasia, but an endeavour to place that condition on a wider basis, harmonising with it various analogous states " which nullify or impair the articulate expression of thought." He moreover contends that " a very large number of different deviations from the normal use of language must be taken into consideration besides its abolition before we are in a position to generalise confidently upon the result."
The concluding paper is by Mr. G. Thompson on " The Sphygmograph in Epilepsy." It consists of tracings made with the instrument in some,epileptics, accompanied by comments on the vascular condition pourtrayed, and constitutes a valuable chapter on the history of sphygmography.
We have devoted that extent of consideration to this volume which indicates our appreciation of the value of its contents, notwithstanding the expression of some strictures 011 its general features; and we are persuaded that those of our readers who will study the volume at leisure will bear us out in the good opinion we have formed of the scientific matter found in its pages. In conclusion we hope Dr. Crichton Browne will persevere with his most praiseworthy endeavours to render the large asylum over which he so ably presides a real centre of work, and to utilise it as a large storehouse of experience and observation; for it must be confessed that the generality of
|
Long-term antibody persistence and exceptional vaccination response on previously SARS-CoV-2 infected subjects
a b s t r a c tBackground: The first COVID-19 vaccines are being distributed to the general population. However, the shortage of doses is slowing down the goal of reaching herd immunity. The aim of the study was to verify whether previously SARS-CoV-2 infected subjects, a considerable portion of the population, should receive the same vaccination treatment of seronegative individuals. Methods: Health-professionals either recovered from COVID-19 or never infected by SARS-CoV-2 were serologically tested at different time-points right before, and several days after, vaccination. Results: Previously infected individuals showed humoral immune responses, 21 days after the first dose, that was approximately 10-folds higher than the seronegative group 21 days after the second dose. Seropositivity persists for at least 11 months. Conclusion: During a shortage of COVID-19 vaccine doses, previously SARS-CoV-2 infected individuals should be dispensed from the vaccination campaign. When dose availability returns to normality, injection of a single dose for seropositive individuals should be considered.
# Introduction
The coronavirus disease (COVID-19) is a pandemic threatening the health and economy of the world's population. Exceptional research efforts led to the rapid development of vaccines, which are now starting to be distributed to the general population. The BioNTech/Pfizer vaccine BNT162b2 Comirnaty was the first to be approved, in both Europe and US, showing a remarkable 95% efficacy [bib_ref] Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine, Polack [/bib_ref] [bib_ref] Phase I/II study of COVID-19 RNA vaccine BNT162b1 in adults, Mulligan [/bib_ref]. Unlike conventional vaccines, Comirnaty is a lipid nanoparticle-formulated, nucleoside-modified RNA (modRNA) encoding the SARS-CoV-2 full-length spike-protein (S-protein) [bib_ref] Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine, Polack [/bib_ref] , locked in the pre-fusion conformation [bib_ref] Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation, Wrapp [/bib_ref]. Comirnaty represents the first large-scale mRNA vaccination campaign thus, data regarding the serological response in the general population are scarce [bib_ref] mRNA vaccines-a new era in vaccinology, Pardi [/bib_ref] [bib_ref] Mechanism of action of mRNA-based vaccines, Iavarone [/bib_ref]. Furthermore, this vaccine was never thoroughly tested on individuals previously infected by SARS-CoV-2 and only preliminary information is available on its effect on this relatively large portion of the population [bib_ref] Antibody responses in seropositive persons after a single dose of SARS-CoV-2 mRNA..., Krammer [/bib_ref]. Most importantly, due to the shortage of vaccine doses, scientists are wondering for how long antibodies against SARS-CoV-2 persist in previously infected individuals and whether they should be administered with two, one or no vaccination doses.
The aim of this study was 1) to evaluate if, and for how long, previously infected subjects retain a degree of immunity (i.e., immunological memory), and 2) to compare immune response mounting upon vaccination in naturally seropositive and seronegative individuals.
# Methods
## Covidiagnostix study
## Contents lists available at sciencedirect
Vaccine j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / v a c c i n e A serological test right before the first vaccination dose (T 0 ) was implemented to discriminate between naturally SARS-CoV-2 seropositive and seronegative individuals.
## Inclusion criteria
All seropositive individuals at T 0 , with also a diagnostic history of previous COVID-19 (n = 157) allowing the dating of the disease (i.e., positive RT-PCR swab test and/or positive serological during the COVID-19 period), as well as 10 seronegative individuals, yet having a diagnostic history of previous COVID-19, were included in the study (Scheme 1). This group was called the ''COVIDgroup". An equal number (n = 167) of randomly selected (alphabetical order) seronegative individuals with a diagnostic history of no previous COVID-19 (i.e., negative RT-PCR swab test and/or negative serological test during the COVID-19 period), matching the COVIDgroup for gender distribution, were also included in the study (Scheme 1). This group was called the ''NO-COVID-group".
## Study procedures
Blood samples, collected as described elsewhere [bib_ref] Biochemical, immunochemical and serology analytes validation of the lithium heparin BD Barricor..., Ferrari [/bib_ref] [bib_ref] Alcohol and illicit drugs in drivers involved in road traffic crashes in..., Ferrari [/bib_ref] into clotactivator BD vacutainer tubes (cat# 369032) without a separator (Becton, Dickinson and Company, NJ, US), were withdrawn at T 0 , right before receiving the first vaccination dose, and time-1 (T 1 ), 21 days later, before the injection of the second dose. For the 167 seronegative subjects, a third blood sample was withdrawn at time-2 (T 2 ), 21 days after the second vaccination dose (Scheme 1). Scheme 1. Study design scheme. Light blue background is associated with serological test targeted anti-N-protein antibodies. Light green background is associated with serological test targeted anti-S-protein antibodies.
At T 0, all serum samples were tested for the presence of SARS-CoV-2 nucleocapsid-protein (N-protein) specific antibodies using the Roche electrochemiluminescence immunoassay (ECLIA) Elecsys Anti-SARS-CoV-2 test (Ref# 09203095190), on a COBAS-601platform (Roche Diagnostic, Basel, Switzerland), targeted on total immunoglobulins (IgTot) against the N-protein. The test was chosen because of its specificity near 100% (manufacturers' suggested cutoff: 1.0 U/mL) as described in several studies [bib_ref] Side-by-Side Comparison of Three Fully Automated SARS-CoV-2 Antibody Assays with a Focus..., Perkmann [/bib_ref] [bib_ref] Clinical evaluation of commercial automated SARS-CoV-2 immunoassays, Kittel [/bib_ref]. Thanks to an instrumental query, upon a positive result, same samples were further tested on the same platform with the Roche SARS-CoV-2-S test (Ref# 09289275190), targeted on IgTot against the receptor binding domain (RBD) of the viral S-protein.
At T 1 , serum samples of both groups were tested for the presence antibodies specific for the S-protein RBD using the Anti-SARS-CoV-2-S assay. The latter has a signal interval ranging from 0.4 to 250 U/ml. Above 250 U/mL the instrument automatically performs a 1:10 dilution that further extends the upper limit to 2500 U/mL. As reported in the manufacturer's datasheet, the positivity cutoff is set at 0.8 U/mL.
The NO-COVID-group was further tested at T 2 with the Anti-SARS-CoV-2-S test.
## Covid-19 diagnostic data
From the beginning of the COVID-19 pandemic, thanks to a follow up institutional program, RT-PCR swab tests were performed routinely but also whenever a healthcare professional showed symptoms consistent with COVID-19. Samples were taken from back of the throat and analyzed using the Tib-Molbiol's 2019-nCoV Real-Time Reverse-Transcription PCR Kit (cat# 61011896) on a Roche Cobas Z480 thermocycler (Roche Diagnostic, Basel, Switzerland). RNA purification was performed using the Roche Magna pure system (cat# A42352) [bib_ref] Routine blood tests as a potential diagnostic tool for COVID-19, Ferrari [/bib_ref]. Additionally, during May 2020, OSR health professionals were subjected to a serological evaluation using the LIAISON SARS-CoV-2 S1/S2 IgG chemiluminescence immunoassay (CLIA) (Ref# 311450) targeted on IgG specific for the viral S-protein.
# Statistical analysis
Statistical analyses were performed with the software Excel (Microsoft, Redmond, WA, USA). Median and, when possible, average ± standard deviation (SD) were quoted. Comparisons of the quantitative antibody titers were performed by a two-tailed, unequal variances t-test (Welch test). Differences were considered statistically significant if the p-value was lower than 0.05.
# Results
## Covid-group
Of the 167 health professionals (40.7 ± 11.1 years) belonging to the COVID-group, 96 were females (41.5 ± 11.0 years) and 71 were males (39.5 ± 11.2 years). At T 0 , 157 were seropositive for the Nprotein (supplementary and were further tested for the presence of antibodies against the S-protein RBD [fig_ref] Figure 1: Serological results of the subjects involved in the study [/fig_ref]. They showed a median value at the Anti-SARS-CoV-2-S test of 110 U/mL, with one subject above the upper 2500 U/mL instrumental limit and one subject below the 0.4 U/mL lower instrumental limit [fig_ref] Figure 1: Serological results of the subjects involved in the study [/fig_ref]. By excluding these two extreme values the arithmetic mean, and its corresponding standard deviation, were 224.6 ± 345.0 U/mL. In addition to the subject showing a titer below the 0.4 U/mL instrumental limit, two subjects were below the 0.8 U/mL instrumental cutoff limit (0.6 and 0.7 U/mL respectively), resulting ''negative" for the presence of antibodies specific for the S-protein RBD. The 10 subjects (7 females 38.4 ± 12.6 years and 3 males 36.3 ± 7.2 years) negative at the T 0 N-protein serological screening and assigned to the COVID-group because of their proved diagnostic history of COVID-19, were missing the T 0 Sprotein RBD titer, which was measured, according to an instrumental query, only for the N-protein positive subjects. Seven of them had a positive RT-PCR test in March/April 2020 (9-10 months before vaccination) and experienced COVID-19 with very light symptoms (2-3 days of fever,~37.5C°). The remaining three were completely asymptomatic and discovered to have recovered from the disease through the institutional serological test in May 2020. Thus, they experienced COVID-19 more than 8 months before vaccination.
At T 1 , 165 subjects (98.8%) showed an antibody titer above the 2500 U/mL upper instrumental limit [fig_ref] Figure 1: Serological results of the subjects involved in the study [/fig_ref]. The remaining 2 subjects showed antibody titers increased with respect to T 0 (<0.4, 40 and 547, 755 U/mL for T 0 and T 1 , respectively). It must be noted that one of them was the only subject showing an Anti-Sprotein RBD result at T 0 below the instrumental limit (<0.4 U/ mL). The 10 subjects that, at T 0 , showed no presence of anti Nprotein antibodies but were included in the COVID-group because of their COVID-19 diagnostic history, as well as the two subjects ''negative", at T 0 , for the presence of S-protein RBD specific antibodies, all showed titers at T 1 above the instrumental limit. To better inquire into the 98.8% of values above the 2500 U/mL instrumental limit, we diluted (1:50) 19 randomly chosen samples (accounting for >10% of the total samples above the limit) with a single pool of human pre-pandemic serum to bring the instrumental response within the instrumental range [fig_ref] Figure 1: Serological results of the subjects involved in the study [/fig_ref]. After adjusting for the dilution factor, the 19 samples showed a median value of 22,650 U/mL (the arithmetic mean, and its corresponding standard deviation were 30589 ± 26893 U/mL), consistent to what previously observed in a study, including 51 participants, using the same Roche instrumentation [bib_ref] Antibody response to first BNT162b2 dose in previously SARS-CoV-2-infected individuals, Manisty [/bib_ref]. Thus, the median anti-S titer have increased 205-fold with respect to pre-vaccine levels. Two of the diluted samples were from subjects negative at the T 0 N-protein serological screening and showed titers of 37,820 and 12,280 U/mL.
## Covid-group: time-interval between disease and first vaccination dose
The documented diagnostic history of COVID-19, available at the OSR database, showed that the longer time-intervals between a positive swab test and the first vaccination dose were 11 months (1 subject), 10 months (11 subjects) and 9 months (6 subjects). Ninety-one subjects showed a positive serological test in May 2020, thus they had experienced COVID-19 at least 8 months before T 0 . The remaining 58 health professionals had timeintervals ranging from 1 to 7 months. Among the 109 subjects (65.2%) showing the longer time-intervals (8 to 11 months), all of them (100%) showed antibody titers at T 1 > 2500 U/mL.
## No-covid-group
The NO-COVID-group was formed by randomly selecting 167 subjects, matching the COVID-group for gender distribution (96 females aged 42.8 ± 9.3 years, and 71 males aged 42.6 ± 11.1 years), showing both a negative result at the T 0 N-protein serological assay (supplementary and a negative diagnostic history of COVID-19. For the latter reasons, anti-S-protein RBD antibody were also assumed to absent in these subjects. At T 1 no subjects showed an anti-S-protein RBD titer above the 2500 U/mL and only one showed a titer still below the 0.8 U/mL cutoff limit [fig_ref] Figure 1: Serological results of the subjects involved in the study [/fig_ref]. The median value was 44.1 U/mL (arithmetic mean, and corresponding standard deviation, equal to 112.8 ± 269.9 U/mL). At T 2 , 21 days after the second dose, all of the 167 subjects, belong- ing to the NO-COVID-group, showed antibody titers above the 0.8 U/mL cutoff limit and 58 of them (34.7%) were above the 2500 U/mL upper instrumental limit [fig_ref] Figure 1: Serological results of the subjects involved in the study [/fig_ref]. Their median value was 1806 U/mL. As for the COVID-group, we diluted a number (n = 13) of randomly chosen samples (>2500 U/mL), with human pre-pandemic serum. In contrast to the COVID-group, as for the COVID-group, a 1:50 dilution was applied to bring the instrumental response within the instrumental range [fig_ref] Figure 1: Serological results of the subjects involved in the study [/fig_ref]. The 13 samples showed a median value of 3290 U/mL (arithmetic mean, and corresponding standard deviation, equal to 4328 ± 2920 U/mL), , consistent to what previously observed by Manisty et al. [bib_ref] Antibody response to first BNT162b2 dose in previously SARS-CoV-2-infected individuals, Manisty [/bib_ref].
## Covid-group vs no-covid-group
At T 1 , 98.8% of the individuals included in the COVID-group were above the 2500 U/mL upper instrumental limit whereas none of the NO-COVID was. At T 2 only 34.7% of the latter group was above that limit. By comparing the instrumental signal outcomes, the T 1 diluted samples of the COVID-group showed an averaged value (31240 ± 28089 U/mL) statistically significantly (p-value < 0.0001) higher (>7-folds) than that of the NO-COVID-group at T 2 (4328 ± 2920 U/mL). It must be noted that, in contrast with the COVID-group, only 34.7% of the NO-COVID-group were above the 2500 U/mL upper instrumental limit at T 2 , thus the 4328 ± 2920 U/mL (obtained from diluted samples) largely overestimated the average of the entire group.
# Discussion
The comparison between the humoral immune responses upon Comirnaty vaccination in naturally seropositive and seronegative individuals showed different behaviors. Those with a past SARS-CoV-2 infection showed antibody titers, 21 days after the first vaccination dose, exceeding the upper limit of detection in almost 99% of the cases. This was consistent with previous studies [bib_ref] Antibody responses in seropositive persons after a single dose of SARS-CoV-2 mRNA..., Krammer [/bib_ref] [bib_ref] Antibody response to first BNT162b2 dose in previously SARS-CoV-2-infected individuals, Manisty [/bib_ref] , one of them using the same Roche instrumentation [bib_ref] Antibody response to first BNT162b2 dose in previously SARS-CoV-2-infected individuals, Manisty [/bib_ref] , showing that in subjects with a previous SARS-CoV-2 infection, the anti-S titers (three weeks after vaccination) increase approximately 200-fold with respect to pre-vaccine levels. In contrast, none of the seroneg-ative individuals was above the upper instrumental limit 21 days after the first vaccination dose, whereas, 21 days after the second dose, 34.7% of them was above the 2500 U/mL upper limit. Similar results were obtained by Manisty et al. [bib_ref] Antibody response to first BNT162b2 dose in previously SARS-CoV-2-infected individuals, Manisty [/bib_ref]. Further investigation of the samples showed that seropositive subjects, 21 days after the first vaccination dose, have antibody titers approximately one order of magnitude higher than their seronegative counterpart 21 days after the second dose.
Interestingly, the 10 subjects with no presence of anti N-protein antibodies, yet included in the COVID-group, and two of the three subjects showing very low anti-S RBD antibody titers at T 0 (below the 0.8 U/mL instrumental cutoff limit), also exhibited titers at T 1 above the 2500 U/mL instrumental limit. The 10 subjects with a documented COVID-19 diagnostic history, yet negative at the anti N-protein test, might be considered as false negative tests. However, the Elecsys Anti-SARS-CoV-2 test has a very high specificity and the lack of N-protein antibodies could also be related to a physiological decay. Unfortunately, the anti-S titers, which could have been useful to discriminate between the two above hypotheses, were not available for these subjects. We might speculate that, the fact that all of them experienced the disease at least 8 months before vaccination and had either asymptomatic or very light disease course, could be consistent with a physiological antibody decay as shown in previous studies [bib_ref] Clinical and immunological assessment of asymptomatic SARS-CoV-2 infections, Long [/bib_ref]. Yet, dilution of two of these samples showed that their titers were comparable with those displaying a clear anti-N seropositivity at T 0 . Thus, it seems that even very low antibody levels are sufficient to induce the observed strong humoral immune response, upon the first vaccination dose, in individuals previously infected by SARS-CoV-2.
Furthermore, available diagnostic data showed that all of the subjects with the longer disease-to-vaccination time-intervals (8-11 months), representing 65.2% of the COVID-group, showed antibody titers at T 1 above the 2500 U/mL upper instrumental limit. Thus, immunological memory (the ability of the immune system to specifically recognize an antigen that has been previously encountered and initiate a corresponding immune response), seems to last for at least 11 months, much longer than the six months previously hypothesized [bib_ref] Antibody persistence in the first six months following SARS-CoV-2 infection among hospital..., L'huillier [/bib_ref]. Noteworthy, the first Italian autochthonous case was on February 21st, 2020, meaning that 11 months was the largest disease-to-vaccination time-interval possibly available for this study. In summary, our data showed that in previously SARS-CoV-2 infected individuals immunological memory lasts for at least 11 months and that the first dose of the Comirnaty vaccine acts in these individuals as a boost approximately 10-fold higher than it does the second vaccine dose in seronegative subjects. Based on this data, it is plausible to think that certain measures can be adopted to improve the COVID-19 vaccination campaign. For instance, in case of a critical scarcity of doses, vaccination of previously infected individuals can be safely postponed, whereas, in case of a full vaccine doses availability, we suggest that previously infected individuals should be injected with a single vaccine dose only.
# Funding
This project was supported by Ministry of Health of Italy, ''Bando Ricerca COVID-19"; project number: COVID-2020-12371619; project title: COVIDIAGNOSTIX -Health Technology Assessment in Covid serological diagnostics.
## 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: Serological results of the subjects involved in the study. COVID-group: Panel A: anti-S-protein RBD assay results at T 0 (n = 157 black dots) and T 1 (n = 167, red dots). Panel B: 19 samples, exceeding 2500 U/mL (upper instrumental limit) at T 1 , were diluted with pre-pandemic serum to bring the signal within the instrumental range. Dilution factor back-calculated values are shown. NO-COVID-group: Panel C: anti-S-protein RBD assay results at T 1 (n = 167, red dots) and T 2 (n = 167, blue dots). Panel D: 13 samples, exceeding 2500 U/mL (upper instrumental limit) at T 2 , were diluted with pre-pandemic serum to bring the signal within the instrumental range. Dilution factor backcalculated values are shown. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) [/fig]
|
Estimating the causal effect of treatment with direct-acting antivirals on kidney function among individuals with hepatitis C virus infection
BackgroundDirect-acting antivirals (DAA) are highly effective at treating Hepatitis C virus (HCV) infection, with a cure rate >95%. However, the effect of DAAs on kidney function remains debated.MethodsWe analyzed electronic health record data for DAA-naive patients with chronic HCV infection engaged in HCV care at Boston Medical Center between 2014 and 2018. We compared the following hypothetical interventions using causal inference methods: 1) initiation of DAA and 2) no DAA initiation. For patients with normal kidney function at baseline (eGFR>90 ml/ min/1.73m 2 ), we estimated and compared the risk for reaching Stage 3 chronic kidney disease (CKD) (eGFR�60 ml/min/1.73m 2 ) under each intervention. For patients with baseline CKD Stages 2-4 (15<eGFR�90 ml/min/1.73m 2 ), we estimated and compared the mean change in eGFR at 2 years after baseline under each intervention. We used the parametric g-formula to adjust our estimates for baseline and time-varying confounders.ResultsFirst, among 1390 patients with normal kidney function at baseline the estimated 2-year risk difference (95% CI) of reaching Stage 3 CKD for DAA initiation versus no DAA was -1% (-3, 2). Second, among 733 patients with CKD Stage 2-4 at baseline the estimated 2-year mean PLOS ONE difference in change in eGFR for DAA initiation versus no DAA therapy was -3 ml/min/ 1.73m 2 (-8, 2).ConclusionsWe found no effect of DAA initiation on kidney function, independent of baseline renal status. This suggests that DAAs may not be nephrotoxic; furthermore, in the short-term, HCV clearance may not improve CKD.PLOS ONEEffect of DAA therapy on renal function PLOS ONE | https://doi.
[formula] a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 [/formula]
# Introduction
Hepatitis C virus (HCV) infection is a major cause of morbidity and mortality throughout the world and has important extrahepatic sequelae [bib_ref] Hepatitis C and kidney disease: A narrative review, Barsoum [/bib_ref] [bib_ref] Association of hepatitis C seropositivity with increased risk for developing end-stage renal..., Tsui [/bib_ref] [bib_ref] Association of hepatitis C viral infection with incidence and progression of chronic..., Molnar [/bib_ref]. In addition to diabetes and cardiovascular disease, patients with HCV infection are at increased risk for chronic kidney disease (CKD) and end stage renal disease (ESRD) compared with the general population, possibly due to systematic inflammation and vascular damage [bib_ref] Higher prevalence of chronic kidney disease and shorter renal survival in patients..., Satapathy [/bib_ref] [bib_ref] Association of hepatitis C virus infection with risk of ESRD: A population-based..., Su [/bib_ref] [bib_ref] Hepatitis C virus infection increases risk of developing end-stage renal disease using..., Lee [/bib_ref].
Treatment with direct-acting antivirals (DAA) is highly effective at clearing HCV: approximately 95% of patients treated with DAAs achieve a sustained virological response (SVR) equivalent to cure [bib_ref] Access to costly new hepatitis C drugs: Medicine, money, and advocacy, Trooskin [/bib_ref] [bib_ref] Efficacy and safety of ledipasvir/ sofosbuvir for genotype 1b chronic hepatitis C..., Okubo [/bib_ref] [bib_ref] Effectiveness, treatment completion and safety of sofosbuvir/ledipasvir and paritaprevir/ritonavir/ombitasvir + dasabuvir in..., Butt [/bib_ref]. Clinical trials have shown that an SVR following treatment with DAAs is associated with liver function improvements for patients with and without CKD [bib_ref] Clinical outcomes in patients with chronic hepatitis C after direct-acting antiviral treatment:..., Carrat [/bib_ref] [bib_ref] Outcomes after successful direct-acting antiviral therapy for patients with chronic hepatitis C..., Cheung [/bib_ref]. Furthermore, treatment with DAAs has been associated with a reduction in both incidence and severity of extrahepatic manifestations of HCV infection such as diabetes and cardiovascular disease [bib_ref] Extrahepatic manifestations of chronic HCV infection, Cacoub [/bib_ref]. However, the effect of treatment with DAAs on kidney function remains ambiguous and needs to be further investigated.
Clinical trials that led to DAA approval generally had relatively short follow-up duration and did not estimate the mid-and long-term effect of DAA treatment on kidney function. Recent observational studies using real-world data comparing changes in estimated glomerular filtration rate (eGFR) before and after DAA treatment provide inconsistent results. Some studies observed a stabilization of or an increase in eGFR, while one demonstrated a decline in eGFR, specifically among DAA-treated individuals with normal baseline kidney function [bib_ref] Efficacy and safety of ledipasvir/ sofosbuvir for genotype 1b chronic hepatitis C..., Okubo [/bib_ref] [bib_ref] Direct-acting antiviral therapy slows kidney function decline in patients with hepatitis C..., Sise [/bib_ref] [bib_ref] Direct acting antiviral HCV treatment does not influence renal function, Driedger [/bib_ref] [bib_ref] Renal safety after one year of sofosbuvir-based therapy for chronic hepatitis C:..., Medeiros [/bib_ref] [bib_ref] Long-term Evolution of Estimated Glomerular Filtration Rate in Patients With Antiviral Treatment..., Liu [/bib_ref] [bib_ref] Evolution of eGFR in chronic HCV patients receiving sofosbuvir-based or sofosbuvir-free direct-acting..., Liu [/bib_ref]. However, current HCV guidelines do not recommend routine, frequent renal monitoring after DAA treatment in individuals with normal kidney function. Therefore, studies that compare the pre-and post-DAA rate of eGFR decline could be biased by the frequency of and cause for creatinine measurements. In addition, while clinical trials of DAA therapies did not indicate renal toxicity, there have been concerns about kidney adverse effects, especially in connection with sofosbuvir, which is mainly cleared renally .
Target trial emulation is a novel approach to the design of comparative effectiveness studies [bib_ref] Using big data to emulate a target trial when a randomized trial..., Hernán [/bib_ref]. With this method, the study design and statistical analysis of an observational study closely emulate the protocol components of a target trial, i.e., the ideal trial a researcher would conduct to answer the research question. We have, therefore, used the target trial approach to estimate the effect of DAA treatment on kidney function by comparing two hypothetical interventions: if all patients had initiated DAA treatment versus if all patients had not initiated DAA treatment.
# Methods
## Study data
We extracted de-identified electronic health record (EHR) data from the Boston Medical Center Clinical Data Warehouse on all patients with a record of chronic HCV infection between 2014 and 2018. Boston Medical Center is the largest safety net hospital in New England; it serves as a leading primary care provider for Boston's vulnerable populations, including minority and low-income patients; and it has a large HCV testing and screening program [bib_ref] HCV screening, linkage to care, and treatment patterns at different sites across..., Calner [/bib_ref] [bib_ref] Impact of the 2020 COVID-19 pandemic on ambulatory hepatitis C testing, Sperring [/bib_ref]. We collected information on inpatient, outpatient, emergency room visits; laboratory measures; prescriptions; active problem lists (i.e., clinician-compiled list of a patient's health diagnoses); procedures; insurance status; vital status; and demographic characteristics. Chronic HCV infection was defined as having an HCV-RNA result above the laboratory level of detection or a positive HCV genotype result. We used the CKD-EPI algorithm to calculate eGFR [bib_ref] A new equation to estimate glomerular filtration rate, Levey [/bib_ref] [bib_ref] Methods of estimating GFR-different equations including CKD-EPI, Florkowski [/bib_ref]. The Boston University Medical Campus Institutional Review Board approved this study. The need for participant consent was waived by the committee because the data was de-identified.
## Target trial emulation
In our real-world data, individuals with CKD or impaired kidney function tended to have more frequent renal monitoring than those with normal kidney function (S1 and S2 Figs). The differential monitoring and the potential bias due to it were important considerations. Additionally, the interpretation of changes in eGFR differs for individuals with and without CKD. Therefore, we designed two separate target trials. In Trial 1 we included patients with normal baseline kidney function, defined as eGFR>90 ml/min/1.73m 2 . In Trial 2 we included patients with CKD Stage 2-4 at baseline, defined as eGFR between 15 and 90 ml/min/1.73m 2 . The target trials were identical in all components except for the baseline eGFR inclusion criteria and the outcome definition. We then designed two cohort studies to emulate the protocols of each target trial. S1 and S2 Tables detail the target trials and trial emulations.
## Eligibility criteria
Eligible individuals met the following criteria between January 2014 and December 2018: age �18 years, chronic HCV infection as described above, no history of kidney transplant or ESRD (defined as CKD Stage 5 or on dialysis), and the presence of a serum creatinine (to compute the baseline eGFR), transaminase, and the platelet measurements within three months of each other (to calculate a Fibrosis-4 [FIB-4] score). In addition, to capture the patients actively engaged in HCV care, we restricted eligibility to individuals with an outpatient visit for HCV infection within three months of their identified HCV infection in the EHR. Individuals who met these eligibility criteria were split into cohort 1 and cohort 2 according to their baseline eGFR and they were analyzed separately. A flow chart of these samples is presented in [fig_ref] Fig 1: Flow chart of study participants [/fig_ref]
## Interventions
For both cohorts we compared the following two DAA initiation strategies: 1) initiation of DAA within 3 months of baseline and 2) no DAA initiation. S3 Table details the DAA therapies utilized.
## Outcome
The outcome definitions differed by cohort. The outcome for cohort 1 (individuals with normal kidney function) was the development of Stage 3 CKD, defined as the first recorded eGFR�60 ml/min/1.73m 2 .
The outcome for cohort 2 (individuals with baseline CKD Stage 2-4) was change from baseline eGFR to the last recorded eGFR between 24-and 30-months post-baseline.
## Follow-up
For both cohorts, follow-up started at baseline, defined as the earliest date all eligibility criteria were met, and ended at the earliest of one of the following events: death, kidney transplant, ESRD (defined as CKD Stage 5 or on dialysis), 2 years post-baseline, administrative end of follow-up (December 2018), or, in Trial 1, date of outcome. A summary of the study design is presented in [fig_ref] Fig 2: Description of cohort entry timeline [/fig_ref] [bib_ref] Graphical depiction of longitudinal study designs in health care databases, Schneeweiss [/bib_ref].
## Statistical analyses
We summarized the baseline characteristics of each cohort separately using descriptive statistics.
For cohort 1, we used a time-to-event framework to estimate and compare the 2-year cumulative incidence of reaching Stage 3 CKD under the two DAA initiation strategies. For cohort 2, we estimated and compared the mean 2-year change from baseline eGFR under each intervention. Because standard statistical methods cannot appropriately adjust for time-varying confounding, we used the parametric g-formula to generate adjusted estimates [bib_ref] A new approach to causal inference in mortality studies with a sustained..., Robins [/bib_ref] [bib_ref] Estimation of the causal effects of time-varying exposures, Robins [/bib_ref] [bib_ref] A structural approach to selection bias, Hernán [/bib_ref]. Covariates were chosen a priori based on the co-authors' clinical judgement and review of the published literature. Like all causal inference methods, the parametric g-formula estimates the causal effect of a hypothetical intervention (here, the DAA therapy initiation) under the identifiability assumptions (positivity, exchangeability, and consistency) and assumes a correct model specification [bib_ref] A new approach to causal inference in mortality studies with a sustained..., Robins [/bib_ref]. The parametric g-formula algorithm has been described in detail elsewhere [bib_ref] Comparative effectiveness of dynamic treatment regimes: An application of the parametric g-formula, Young [/bib_ref] [bib_ref] Comparative effectiveness of immediate antiretroviral therapy versus CD4-based initiation in HIV-positive individuals..., Lodi [/bib_ref] [bib_ref] Intervening on risk factors for coronary heart disease: An application of the..., Taubman [/bib_ref]. Briefly, it works in two steps: 1) parametric regression models estimate the joint distribution of the outcome, treatment, and time-varying covariates conditional on previous treatment and covariate history and 2) the estimates in step 1 are used to simulate the distribution of the outcome under each intervention using the Monte Carlo method.
In the first step, continuous and binary time-varying covariates were modeled with separate linear and logistic regression models, respectively. All models adjusted for baseline covariates and the most recent lagged value of time-varying covariates. Baseline covariates included age, sex, race/ethnicity (Hispanic or Latino, White, Black, other), body mass index, education level, HCV genotype, insurance type, degree of liver fibrosis (by fibroscan), HIV coinfection, and diabetes. Time-varying covariates were alanine transaminase (ALT), aspartate transaminase (AST), platelet count, change from baseline eGFR, drug or alcohol use disorder [bib_ref] Identifying injection drug use and estimating population size of people who inject..., Janjua [/bib_ref] , mental illness, outpatient visits, and hypertension. We also treated the timing of the laboratory measurements (AST, ALT, platelet count, eGFR) as potential confounders: we fit logistic regression models for these visit processes with a binary outcome of 1 if a measurement was taken at that time and 0 if not. For cohort 1, continuous eGFR was included as an additional baseline covariate, and for cohort 2 CKD stage was included as an additional baseline covariate. Skewed continuous covariates (ALT and AST) were log-transformed for normality. Details on the definition and ascertainment of covariates are presented in S4 [fig_ref] Table: Cohort 1 included 1442 eligible patients with normal baseline kidney function [/fig_ref] In the second step, we used the parameter estimates from the first step to simulate the postbaseline time-varying covariates and the outcome separately under each of the two interventions. In our real-world data, eGFR is measured sporadically and long periods of time can elapse between measurements. Therefore, we imposed an eGFR measurement after 12 months in individuals without a new measurement.
We used nonparametric bootstrapping with 500 samples to obtain 95% confidence intervals (CI) around our estimates based on percentiles.
To explore the validity of our parametric assumptions, we compared the observed outcome means and time-varying covariates with those predicted by our models .
All analyses were conducted in SAS version 9.4 using the publicly available GFORMULA macro.
## Subgroup analyses
As secondary analyses, we reran the estimation procedure in the following patient subgroups with elevated risk of CKD [bib_ref] Predictors of estimated GFR decline in patients with type 2 diabetes and..., Zoppini [/bib_ref] : those with 1) type II Diabetes, 2) hypertension, and 3) age>45 years. The cut-off for age was chosen according to the bimodality of the distribution of age in the study sample (S5 Fig).
## Sensitivity analyses
We conducted several sensitivity analyses to assess the robustness of our results. First, we reran the analysis for both cohort 1 and cohort 2 excluding HCV and HIV coinfected patients [bib_ref] Hepatic decompensation in antiretroviraltreated patients co-infected with HIV and hepatitis C virus..., Lo Re [/bib_ref]. Second, we reran the analysis for both cohorts restricting them to outpatient eGFR measurements only. This was done to attempt to exclude acute renal injuries. Third, we reran the analysis for cohort 1 with the follow-up extended to the database close (2018), and we reran the analysis for cohort 2 with the follow-up extended to 3 years. Fourth, we conducted the analyses with time discretized into months, instead of weeks as was done in the main analysis. Fifth, we ran a series of analyses reordering the time-varying covariates, since each parametric model conditions on past covariates. Sixth, we reran the analyses by 1) imposing an eGFR measurement after 6 months, instead of 12 months, and 2) not imposing a new eGFR measurement. In addition, for cohort 1, we ran a sensitivity analysis defining Stage 3 CKD as the first of two eGFR measurements <60 ml/min/1.73m 2 to be consistent with the official CKD diagnosis guidelines [bib_ref] Safety and efficacy of sofosbuvir-containing regimens in hepatitis C-infected patients with impaired..., Saxena [/bib_ref]. We also reran the primary analysis additionally adjusting for concomitant use of nephrotoxic medications (nonsteroidal anti-inflammatory drugs, diuretics, and angiotensin-converting enzyme inhibitors or angiotensin receptor blockers). Finally, we reran the primary analyses to estimate the effect of treatment with Sofosbuvir-containing regimens on renal function. We did this by censoring patients treated with non-Sofosbuvir-containing regimens at their treatment start date.
# Results
Descriptive characteristics at baseline for each cohort are presented in [fig_ref] Table 1: Baseline characteristics stratified by cohort [/fig_ref]. Descriptive characteristics stratified by treatment with DAA are presented in S5 In cohort 1, 101 (7%) individuals developed Stage 3 CKD events and 13 died during 3,048 person-years of follow-up. As presented in [fig_ref] Table 2: Estimated causal effect of DAA initiation interventions on kidney function stratified by... [/fig_ref] , the adjusted estimated 2-year risk (95% CI) of reaching Stage 3 CKD was 5% (3%, 6%) and 5% (4%, 8%) under DAA initiation and no DAA initiation, respectively (risk difference -1% (-3%, 2%)). [bib_ref] Association of hepatitis C seropositivity with increased risk for developing end-stage renal..., Tsui [/bib_ref] and 435 (59%) were male. At baseline, 68% had hypertension and 26% had diabetes. In cohort 2, 514 (70%) patients initiated DAA during follow-up and of these, 316 (61%) initiated DAA within 3 months of baseline. Among the 514 who initiated DAA, 494 (96%) had a posttreatment HCV-RNA measured, and 481 of 494 (97%) achieved SVR defined as an undetectable HCV-RNA.
In cohort 2, the observed mean (standard deviation, SD) change from baseline eGFR was 0 (16) ml/min/1.73m 2 . The adjusted estimated mean (95% CI) change from baseline in eGFR was 0 (-3, 2) ml/min/1.73m 2 and 3 (-1, 8) ml/min/1.73m 2 under DAA initiation and no DAA initiation, respectively. Compared to no DAA initiation, the mean difference was -3 (-8, 2) ml/ min/1.73m 2 under DAA initiation. There were 14 deaths during the follow-up in cohort 2.
Results of subgroup analyses are presented in S6-S8 Tables. Individuals with diabetes, with hypertension, or aged >45 years had a higher incidence of CKD Stage 3 (cohort 1) and a faster eGFR decline (cohort 2) at 2 years post-baseline compared to their counterparts without diabetes, without hypertension, or aged �45 years. Effect estimates were similar to those in the main analysis, although with lower precision due to smaller sample sizes. The subgroup of patients aged �45 years in cohort 2 had too few patients (N = 78) to allow for estimation.
Results from sensitivity analyses were consistent with main results for both cohorts (S6 and S7 Figs). The distribution of the time-varying means predicted by our models were similar to the observed means (S3 and S4 Figs).
# Discussion
We estimated the causal effect of initiating versus not initiating DAA treatment on kidney function in two cohorts of patients with HCV infection: one with normal kidney function (cohort 1) and one with baseline CKD Stage 2-4 (cohort 2). For those with normal kidney function at baseline, we observed no difference in the risk of reaching Stage 3 CKD had all patients initiated DAA therapy versus had all patients not initiated DAA therapy. Similarly, for those with CKD Stage 2-4 at baseline, we observed no difference in 2-year post-baseline change in eGFR had all versus no patients initiated DAA. These results were robust to several sensitivity analyses that varied analytic decisions and model specification. Additionally, in our secondary analysis we evaluated patient subgroups which might particularly benefit from treatment and found no effect of DAAs on renal function. Altogether these findings suggest that treatment with DAA for HCV infection neither impairs nor improves renal function over a 2-year period. Previous observational studies estimated the association between DAA treatment and eGFR trends. Specifically, Driedger et al. reported no notable change in renal function before and after DAA initiation [bib_ref] Direct acting antiviral HCV treatment does not influence renal function, Driedger [/bib_ref]. In another study, found no association between sofosbuvir-based DAA regimens and eGFR among a group of patients with baseline eGFR<45 ml/min/1.73m 2 , but observed improvements in renal function among those with eGFR>45 ml/min/1.73m 2 [bib_ref] Renal safety after one year of sofosbuvir-based therapy for chronic hepatitis C:..., Medeiros [/bib_ref]. An observational study by Sise et al. compared eGFR slope one year before and one year after DAA initiation and found no difference among those with baseline eGFR>60 ml/min/1.73m 2 but reported improvements among those with baseline eGFR<60 ml/min/1.73m 2 [bib_ref] Direct-acting antiviral therapy slows kidney function decline in patients with hepatitis C..., Sise [/bib_ref]. In addition, in a retrospective cohort study of patients treated with ledipasvir/sofosbuvir for HCV genotype 1b, Okubo et al. reported no changes in eGFR from pretreatment to 12 weeks post-treatment among those with CKD Stage 3 [bib_ref] Efficacy and safety of ledipasvir/ sofosbuvir for genotype 1b chronic hepatitis C..., Okubo [/bib_ref]. Finally, a prospective observational study with biannual eGFR measurements found no significant difference in eGFR evolution among DAA-treated individuals who did and did not achieve SVR [bib_ref] Long-term Evolution of Estimated Glomerular Filtration Rate in Patients With Antiviral Treatment..., Liu [/bib_ref]. All these studies compared kidney function before and after DAA treatment, therefore, they were only analyzing patients who received DAA treatment. While this approach allows for control of baseline confounding factors, whereby patients act as their own controls, these estimates might be biased if patients who received DAA treatment had worse (or better) renal function than those who did not [bib_ref] Access to costly new hepatitis C drugs: Medicine, money, and advocacy, Trooskin [/bib_ref] [bib_ref] Simple, effective, but out of reach? public health implications of HCV drugs, Ward [/bib_ref] [bib_ref] Adoption of direct-acting antiviral medications for hepatitis C: A retrospective observational study, Zullig [/bib_ref]. In contrast, we designed our study to include all patients eligible to receive DAA treatment and we adjusted for potential confounders. Despite a different study design, our results are not only consistent with the lack of a clear association between treatment with DAAs and renal function in the literature, but also are generalizable to all patients eligible to receive DAAs.
Our study has important strengths. We used longitudinal data from an urban safety net hospital with a notable HCV testing and screening program, which allowed us to include a diverse patient population. We analyzed this rich source of real-world data with causal inference methods that explicitly adjusted for time-varying confounding and selection bias. Timevarying confounding is intrinsic to this study setting: elevated liver function tests and the presence of certain covariates indicate a more pressing need for treatment, and treatment improves liver function [bib_ref] Clinical outcomes in patients with chronic hepatitis C after direct-acting antiviral treatment:..., Carrat [/bib_ref]. Selection bias is also present, whereby patients with CKD at baseline have more frequent post-baseline eGFR measurements than individuals with normal renal function [bib_ref] A structural approach to selection bias, Hernán [/bib_ref]. We address both methodological issues by adjusting for variables associated with renal function, including explicit adjustment for renal monitoring frequency. By stratifying patients into cohorts according to baseline renal function, we aimed to reduce the impact of selection bias on our results. Our study explicitly estimated what would have happened had all patients initiated versus not initiated DAA, which differs from, but complements, previous studies that compared renal function before and after DAA therapy. As it would be unethical to conduct a randomized controlled trial (RCT) that assigns some patients to HCV treatment and others not, our study is as close methodologically as possible to an RCT and, notably, enables us to comment upon causation, not merely association.
Our study also has several limitations. First, our methods rely on the assumption of no residual confounding, yet we may not have captured all confounding characteristics. However, we separately analyzed two cohorts with different baseline renal function and clinical and demographic characteristics, and, thus, likely they also had different confounding structures. We found the null effect in both cohorts, therefore, it reinforced our conclusions. Second, retrospective use of EHRs is subject to error: we may have missed deaths, HCV treatments, or laboratory measurements done outside the hospital system. Third, we used prescriptions to ascertain DAA treatment initiation, but we had no record of medication adherence. Treatment completion rates for DAAs as low as 68% have been reported for HCV infected patients [bib_ref] Effectiveness, treatment completion and safety of sofosbuvir/ledipasvir and paritaprevir/ritonavir/ombitasvir + dasabuvir in..., Butt [/bib_ref]. However, a recent study in an overlapping sub-population at Boston Medical Center found, by comparing chart-reviewed DAA adherence with EHR prescription records, that only 7% of individuals prescribed DAAs never started treatment [bib_ref] Enhancing linkage to hepatitis C virus treatment following pregnancy in women identified..., Epstein [/bib_ref]. Fourth, because of the relatively high eGFR in our sample, we used a baseline eGFR cutoff of 90 ml/min/1.73m 2 . The cutoff for clinically significant CKD is 60 ml/min/1.73m 2 , and so future studies with larger sample size and/or more renally impaired patients should replicate this research with the more stringent cutoff. Finally, the majority of individuals who initiated DAA treatment received sofosbuvir-containing regimens, in line with DAA treatment in the United States during the study period. A future study with a larger database, perhaps pooling data across medical centers, may look at the effect of specific DAA combinations on renal function [bib_ref] Evolution of eGFR in chronic HCV patients receiving sofosbuvir-based or sofosbuvir-free direct-acting..., Liu [/bib_ref].
Despite these limitations, our results have important implications for renal monitoring among patients treated with DAA. Frequency of renal monitoring should remain the same after treatment with DAA as it was before for those with and without renal impairment. According to Kidney Disease: Improving Global Outcomes guidelines, this means at least annual eGFR monitoring for those with normal renal function and 1 to 4+ times per year for those with CKD, depending on severity. Our study suggests DAAs may not be nephrotoxic and that HCV clearance may not restore renal function, but future research should continue to work to elucidate these relationships. In conclusion, we used robust methodology combined with a rich source of data and found no causal effect of DAA treatment on kidney function.
## Supporting information
[fig] Fig 1: Flow chart of study participants. https://doi.org/10.1371/journal.pone.0268478.g001 [/fig]
[fig] Fig 2: Description of cohort entry timeline. a End stage renal disease (ESRD) includes dialysis and Stage 5 chronic kidney disease (CKD). b Baseline covariates: Fibroscan, genotype, body mass index, insurance type, alanine transaminase, aspartate transaminase, platelets, HIV, drug or alcohol use disorder, mental illness, diabetes, hypertension. c Time-varying covariates: alanine transaminase, aspartate transaminase, platelets, change in eGFR, diagnosis of drug or alcohol use disorder, mental illness, hypertension, clinic visits, treatment with DAAs. d Follow-up window ends at the earliest of one of the following events: death, kidney transplant, ESRD, 104 weeks (2 years) post-baseline, administrative end of follow-up (December 2018), or, in Trial 1, date of outcome. https://doi.org/10.1371/journal.pone.0268478.g002 [/fig]
[fig] S1: Fig. Proportion of patients with any visit over follow-up stratified by cohort. � p<0.05, �� p<0.01, ��� p<0.001 for difference in proportion of patients with any visit between cohort 1 and cohort 2. (TIF) S2 Fig. Proportion of patients with an eGFR measurement over follow-up stratified by cohort. � p<0.05, �� p<0.01, ��� p<0.001 for difference in proportion of patients with an eGFR measurement between cohort 1 and cohort 2. (TIF) S3 Fig. Observed versus simulated data for cohort 1. Observed (solid line) versus simulated data (dotted line) under the natural course (i.e., treatment is not imposed). The visit process models are for the timing of the laboratory measurements. (TIF) S4 Fig. Observed versus simulated data for cohort 2. Observed (solid line) versus simulated data (dotted line) under the natural course (i.e., treatment is not imposed). The visit process models are for the timing of the laboratory measurements. (TIF) S5 Fig. Histogram of age distribution in study sample. (TIF) S6 Fig. Sensitivity analyses for cohort 1. Administrative end of follow-up/database close date was December 2018. Reordering covariates refers to the temporal parametric assumptions made in the model. (TIFF) S7 Fig. Sensitivity analyses for cohort 2. Administrative end of follow-up/database close date was December 2018. Reordering covariates refers to the temporal parametric assumptions made in the model. (TIFF) [/fig]
[table] Table: Cohort 1 included 1442 eligible patients with normal baseline kidney function. In this group, the median [IQR] age was 49 [35, 57] years, the median [IQR] eGFR was 105 ml/min/1.73m 2 [98, 114] and 1022 (71%) were male. At baseline, 53% of patients had hypertension and 14% had diabetes. In [/table]
|
Chronically Activated T-cells Retain Their Inflammatory Properties in Common Variable Immunodeficiency
Purpose Immune dysregulation complications cause significant morbidity and mortality in common variable immunodeficiency (CVID), but the underlying pathophysiology is poorly understood. While CVID is primarily considered a B-cell defect, resulting in the characteristic hypogammaglobulinemia, T-cells may also contribute to immune dysregulation complications. Here, we aim to further characterize T-cell activation and regulation in CVID with immune dysregulation (CVIDid). Methods Flow cytometry was performed to investigate T-cell differentiation, activation and intracellular cytokine production, negative regulators of immune activation, regulatory T-cells (Treg), and homing markers in 12 healthy controls, 12 CVID patients with infections only (CVIDio), and 20 CVIDid patients. Results Both CD4 + and CD8 + T-cells in CVIDid showed an increased activation profile (HLA-DR + , Ki67 + , IFNγ +) when compared to CVIDio, with concomitant upregulation of negative regulators of immune activation PD1, LAG3, CTLA4, and TIGIT. PD1 + and LAG3 + subpopulations contained equal or increased frequencies of cells with the capacity to produce IFNγ, Ki67, and/or GzmB. The expression of PD1 correlated with serum levels of CXCL9, 10, and 11. Treg frequencies were normal to high in CVIDid, but CVIDid Tregs had reduced CTLA-4 expression, especially on CD27 + effector Tregs. Increased migratory capacity to inflamed and mucosal tissue was also observed in CVIDid T-cells. Conclusion CVIDid was characterized by chronic activation of peripheral T-cells with preserved inflammatory potential rather than functional exhaustion, and increased tissue migratory capacity. While Treg numbers were normal in CVIDid Tregs, low levels of CTLA-4 indicate possible Treg dysfunction. Combined studies of T-cell dysfunction and circulating inflammatory proteins may direct future treatment strategies.
# Introduction
Common variable immunodeficiency (CVID) is characterized by recurrent infections caused by low IgG, and IgA or IgM, for which patients are treated with immunoglobulin G replacement therapy (IgRT) [bib_ref] How I treat common variable immune deficiency, Cunningham-Rundles [/bib_ref]. IgRT has significantly decreased the risk of infectious complications in CVID, but nonetheless, over a third of patients develop immune dysregulation complications [bib_ref] Clinical picture and treatment of 2212 patients with common variable immunodeficiency, Gathmann [/bib_ref] [bib_ref] Common variable immunodeficiency disorders: division into distinct clinical phenotypes, Chapel [/bib_ref] , resulting in significant morbidity and mortality [bib_ref] Morbidity and mortality in common variable immune deficiency over 4 decades, Resnick [/bib_ref] [bib_ref] The burden of common variable immunodeficiency disorders: a retrospective analysis of the..., Odnoletkova [/bib_ref]. A wide range of immune dysregulation phenomena can be observed in CVID, including granulomatous-lymphocytic interstitial lung disease (GLILD), enteritis, autoimmune cytopenias, lymphoproliferation, and hematological malignancies [bib_ref] Common variable immunodeficiency disorders: division into distinct clinical phenotypes, Chapel [/bib_ref] [bib_ref] The spectrum of disease manifestations in patients with common variable immunodeficiency disorders..., Maarschalk-Ellerbroek [/bib_ref]. The underlying pathophysiology of immune dysregulation in CVID is currently poorly understood, which complicates diagnostics and treatment [bib_ref] British Lung Foundation/United Kingdom Primary Immunodeficiency Network consensus statement on the definition,..., Hurst [/bib_ref] [bib_ref] Current understanding and recent developments in common variable immunodeficiency associated autoimmunity, Gereige [/bib_ref].
While the defining hypogammaglobinemia in CVID is considered to be primarily the result of B-cell dysfunction, several lines of evidence suggest an additional role for T-cells in CVID with immune dysregulation (CVIDid). Biopsies of lung granulomas in CVIDid show predominance of CD4 + T helper (Th) cells [bib_ref] Granulomatous and lymphocytic interstitial lung disease: a spectrum of pulmonary histopathologic lesions..., Rao [/bib_ref] [bib_ref] Use of combination chemotherapy for treatment of granulomatous and lymphocytic interstitial lung..., Chase [/bib_ref] , while regulatory T-cells (Tregs) are often absent [bib_ref] Granulomatous and lymphocytic interstitial lung disease: a spectrum of pulmonary histopathologic lesions..., Rao [/bib_ref]. In peripheral blood of patients with CVIDid, a decreased CD4/CD8 ratio was observed with decrease of naïve T-cells [bib_ref] Unravelling the complexity of T cell abnormalities in common variable immunodeficiency, Giovannetti [/bib_ref] , Tregs, Th17 cells, and follicular helper T (Tfh) cells [bib_ref] Predominantly antibody-deficient patients with non-infectious complications have reduced naive B, Treg, Th17,..., Edwards [/bib_ref] [bib_ref] The TH1 phenotype of follicular helper T cells indicates an IFN-γ-associated immune..., Unger [/bib_ref]. Moreover, there are indications that CVID T-cells may be functionally exhausted, including reduced capacity to respond to bacterial antigens and increased expression of PD1 [bib_ref] Exhaustion of bacteria-specific CD4 T cells and microbial translocation in common variable..., Perreau [/bib_ref] [bib_ref] Common variable immunodeficiency-associated endotoxemia promotes early commitment to the T follicular lineage, Coz [/bib_ref]. Our group and others have previously demonstrated that serum cytokines in CVIDid [bib_ref] Targeted proteomics reveals inflammatory pathways that classify immune dysregulation in common variable..., Berbers [/bib_ref] [bib_ref] Plasma protein profiling reflects TH1-driven immune dysregulation in common variable immunodeficiency, Hultberg [/bib_ref] are shifted towards a Th1 phenotype, and we observed an upregulation of proteins associated with immune regulation-IL10, LAG3, and 4-1BB. Monogenic primary immunodeficiencies caused by mutations in immune regulation genes such as CTLA4and ICOS [bib_ref] Homozygous loss of ICOS is associated with adult-onset common variable immunodeficiency, Grimbacher [/bib_ref] often result in a CVIDid phenotype. However, how the interplay between immune regulation and immune activation results in CVIDid remains poorly understood.
To further study the balance between immune activation and immune regulation in CVIDid, we used flow cytometry to evaluate naïve T-cell subsets, T-cell activation and cytokine production, exhaustion, negative regulators of immune activation, regulatory T-cells, and T-cell homing markers.
# Methods
# Ethics statement
Ethical approval for this study for all participants was received from the Medical Ethical Committee of the
## Study population and sample collection
Patients were diagnosed with CVID according to the European Society for Immunodeficiencies criteriaand were included at the outpatient clinics of the UMC Utrecht and the Erasmus MC University Medical Center Rotterdam, The Netherlands. Patients were eligible if they were aged seven or older, and they and/or their legal guardians signed informed consent. Household members of patients were recruited as healthy controls (HC). Medication use up to 3 months prior to sampling was recorded.
## Sample processing
Peripheral blood mononuclear cells (PBMC) were isolated from blood by Ficoll density centrifugation (GE Healthcare-Biosciences, AB), and frozen at -180 °C until use. Cells were subsequently thawed, counted and plated at 1,000,000 live cells per panel per sample. For the panels including intracellular cytokine measurement, cells were first restimulated with 20 ng/mL phorbol 12-myristate 13-acetate (PMA, MilliporeSigma) and 1 μg/mL ionomycin (MilliporeSigma) for 4 h at 37 °C with addition of monensin (Golgistop, BD Biosciences, 1:1500) during the last 3.5 h. Cell death was stained in all panels using Fixable Viability Dye eFluor 506 (eBioscience). Next, cells were incubated with the surface antibodies [fig_ref] Table 1: Cohort characteristics [/fig_ref] for 20 min at 4 °C and washed. Cells were then permeabilized with fixation/permeabilization reagent (eBioscience) for 30 min at 4 °C, washed, and incubated with the intracellular antibodies (Supplementary [fig_ref] Table 1: Cohort characteristics [/fig_ref]. Cells were stored at 4 °C until the next day, when they were measured on the LSR Fortessa (BD Biosciences).
# Analysis and statistics
For flow cytometric data, median fluorescence intensities (MFI) and percentages of positive cells were analyzed in FlowJo. All statistical analyses and graphic representations were done in R 3.2.0. Continuous variables were compared using the Mann-Whitney rank test, or paired Wilcoxon-rank test for paired samples. Correlation was calculated using Spearman's correlation. P values below 0.05 were considered statistically significant.
# Results
## The overall t-cell profile of cvidid patients differs from cvidio and hc
To investigate immune activation and regulation in CVID, PBMC were isolated from 12 healthy controls (HC), 12 CVID patients with infections only (CVIDio), and 20 patients with CVIDid [fig_ref] Table 1: Cohort characteristics [/fig_ref]. Patients were selected from a cross-sectional Dutch primary immunodeficiency cohort when they received IgRT and did not use any immunomodulatory medication during and the last 3 months prior to sampling. Three of the CVIDid patients had a known CVIDassociated monogenetic disease (CTLA4 haploinsufficiency, STAT1 gain of function, and PIK3R1). Flow cytometry was performed as described in the Supplementary Information (see also Supplementary [fig_ref] Table 1: Cohort characteristics [/fig_ref] First, pooled flow cytometry data was analyzed in an unsupervised manner using principal component analysis (PCA, [fig_ref] Figure 1: General description of T-cell subsets in CVID [/fig_ref]. CVIDio and HC clustered closely together and were distinct from most CVIDid samples. This suggests that most variation in the flow cytometric T-cell data related to immune dysregulation and not to the hypogammaglobulinemia shared by CVID patients. CVIDid patients did not cluster by treatment history or location of autoimmunity [fig_ref] Figure 5: Migratory [/fig_ref] , suggesting that peripheral blood T-cell skewing may be generalized among CVIDid patients.
## Cvidid t-cells are th-1-skewed and chronically activated
Next, the distribution of CD4 + and CD8 + T-cell subsets in CVIDid was assessed. Frequencies in CD4 + , CD8 + , and naïve/memory subsets were similar to that reported in previous studies [bib_ref] Unravelling the complexity of T cell abnormalities in common variable immunodeficiency, Giovannetti [/bib_ref] [bib_ref] Frequency of treg cells is reduced in CVID patients with autoimmunity and..., Arumugakani [/bib_ref] : a decreased CD4/CD8 ratio [fig_ref] Figure 1: General description of T-cell subsets in CVID [/fig_ref] , and a trend of decreased naïve (CD45RA + CCR7 +) CD4 + T-cells in CVIDid compared to CVIDio (p = 0.07) [fig_ref] Figure 1: General description of T-cell subsets in CVID [/fig_ref]. In addition, a subset of CVIDid patients showed a high percentage of CD4 + T effector memory cells, and CD4 + T effector memory cells re-expressing CD45RA (TEMRA), which are associated with chronic activation such as observed in viral infections [bib_ref] Unique phenotypes and clonal expansions of human CD4 effector memory T cells..., Tian [/bib_ref] [fig_ref] Figure 1: General description of T-cell subsets in CVID [/fig_ref]. Within the naïve CD4 + T-cells, CD31 + recent thymic emigrants
[formula] 0 (0.00%) 0 (0.00%) 8 (40.00%) Hematological 0 (0.00%) 0 (0.00%) 2 (10.00%) Gastrointestinal 0 (0.00%) 0 (0.00%) 8 (40.00%) Rheumatological 0 (0.00%) 0 (0.00%) 6 (30.00%) Dermatological 0 (0.00%) 0 (0.00%) 4 (20.00%) Hematological malignancy 0 (0.00%) 0 (0.00%) 2 (10.00%) Lymphoproliferation (incl splenomegaly) 0 (0.00%) 0 (0.00%) 10 (50.00%) Other 0 (0.00%) 0 (0.00%) 4 (20.00%) DMARD-naïve/subclinical disease NA NA 12 (60.00%) Disease in remission NA NA 8 (40.00%) Genetics [/formula]
Not done 12 (100.00%) 12 (100.00%) 12 (60.00%) Nothing found 0 (0.00%) 0 (0.00%) 2 (10.00%) VUS found 0 (0.00%) 0 (0.00%) 2 (10.00%) Pathogenic mutations found 0 (0.00%) 0 (0.00%) 3 (15.00%) + 1 TNFRSF13B (TACI) mutation were more abundant in CVIDid than the CD31-central naïve T-cells [fig_ref] Figure 6: Patterns of CD4 + T-cell activation [/fig_ref]. No differences in CD8 + T-cell distribution were observed [fig_ref] Figure 1: General description of T-cell subsets in CVID [/fig_ref]. Within the effector/memory (CD45RO +) subpopulation, proportions of CD4 + T-cells expressing HLA-DR, Ki67, and IFNγ were significantly increased in CVIDid, while IL17a, IL13, and TNF-α expressing T-cells were not different between CVIDid and CVIDio . A similar activation pattern was observed in the CD8 + effector memory population for HLA-DR, Ki67, and GzmB . We previously described [bib_ref] Targeted proteomics reveals inflammatory pathways that classify immune dysregulation in common variable..., Berbers [/bib_ref] serum cytokine and chemokine levels in an overlapping cohort, allowing comparison between soluble serum markers and T-cell characteristics. Pooling the data from HC, CVIDio, and CVIDid, we observed that the proportion of IFNγ + Th cells correlated with serum levels of interferon-inducible chemokines CXCL9, 10, and 11 . While we observed an increase of serum IL17a in CVIDid, there was no corresponding increase of IL17a-producing Th cells, and the frequency of these cells did not correlate with the serum IL17a levels .
## T-cells expressing negative regulators of immune activation retain their inflammatory potential in cvidid
Next, we investigated whether this immune activation resulted in immune exhaustion in CVIDid, which is known to happen in the context of chronic inflammation [bib_ref] PD-1+CD8+ T cells are clonally expanding effectors in human chronic inflammation, Petrelli [/bib_ref]. In CVIDid CD4 + -cells, we observed an increased proportion of cells expressing negative regulators of immune activation PD1, LAG3, CTLA4, ICOS, and TIGIT . In addition, CVIDid CD4 + T-cells expressed higher levels of CD95 (FAS-L), showing that they are terminally differentiated and may be more prone to apoptosis . In In order to assess whether the PD1 and LAG3 expressing cells were functionally exhausted, production of IFNγ, GzmB, Ki67, and CD95 was compared between the PD1/LAG3 + and PD1/LAG3 − populations and . In CD4 + T-cells, the PD1 + and LAG3 + subpopulations showed equal or increased expression of IFNγ, Ki67, and/or GzmB as the PD1 − and LAG3 − subpopulations, indicating retained functional capacity. Expression of PD1 correlated with serum levels of pro-inflammatory CXCL9, 10, and 11 as well as immune regulatory IL10 , further illustrating the association between PD1 and chronic inf lammation. In CD8 + T-cells, IFNγ and GzmB expression was also intact in PD1 + and LAG3 + cells, but Ki67 + cells were less frequent in the LAG3 + subpopulation, indicating reduced proliferative capacity [fig_ref] Figure 1: General description of T-cell subsets in CVID [/fig_ref].
## Cvidid regulatory t-cells fail to upregulate ctla4
In addition to negative regulators of immune activation, regulatory T (Treg) cells also contribute to limiting inflammation-related pathology. In contrast to previous studies [bib_ref] Low numbers of regulatory T cells in common variable immunodeficiency: association with..., Fevang [/bib_ref] , we did not observe a decreased proportion of CD25 + FOXP3 + Tregs in CVIDid, and a subgroup of CVIDid patients had even increased Treg frequencies [fig_ref] Figure 4: Regulatory T-cells in CVID [/fig_ref]. ICOS expression in CVIDid Tregs also indicated normal-to-increased activation [bib_ref] Inducible costimulator (ICOS) is a marker for highly suppressive antigen-specific T cells..., Vocanson [/bib_ref] of Tregs [fig_ref] Figure 4: Regulatory T-cells in CVID [/fig_ref] , but CTLA4 expression was reduced in CVIDid Tregs, both in cell frequency as MFI in the CTLA4-positive population [fig_ref] Figure 4: Regulatory T-cells in CVID [/fig_ref]. The failure to upregulate CTLA4 in CVIDid Tregs was isolated to the activated and usually highly suppressive [bib_ref] CD70 expression determines the therapeutic efficacy of expanded human regulatory T cells, Hornero [/bib_ref] CD27 + Treg population [fig_ref] Figure 4: Regulatory T-cells in CVID [/fig_ref] , and may result in reduced Treg function, as expression of CTLA4 is important for the suppressive function of Tregs. In addition, a subpopulation of CVIDid patients showed low levels of TIGIT-expressing Tregs, which may impair TIGIT-driven selective suppression of Th1 and Th17 effector cells [bib_ref] Lag-3, Tim-3, and TIGIT: Co-inhibitory receptors with specialized functions in immune regulation, Anderson [/bib_ref].
## Increased migratory capacity in cvidid t-cells
The immune dysregulation of most CVIDid patients occurs not only in the systemic immune system, but also locally in affected tissues, which are often mucosal sites (lung and/ or gut). Expression of the mucosal homing marker CCR9 was increased on naïve CD4 + and CD8 + populations of CVIDid T-cells [fig_ref] Figure 5: Migratory [/fig_ref] , but not on non-naïve (CD45RA −) CD4 + T-cells [fig_ref] Figure 5: Migratory [/fig_ref] , suggesting that this upregulation was at least not entirely antigen-driven. CVIDid CD8 + nonnaïve (CD45RA −) T-cells [fig_ref] Figure 5: Migratory [/fig_ref] did also show increased expression of CCR9 as well as integrin α4β1, which mediates homing to inflamed tissues, including the lung. CVIDid FOXP3 + CD4 + T-cells [fig_ref] Figure 5: Migratory [/fig_ref] also showed increased migratory capacity, as they were enriched for cells expressing integrin α4β1 and gut-homing marker integrin α4β7.
## Patterns of t-cell dysregulation in cvidid patients with glild
Finally, in order to illustrate patterns of T-cell immune dysregulation in individual patients with varying severity of specific immune dysregulation phenotypes, we selected markers most strongly associated with (chronic) immune activation in CD4 + T-cells (HLA-DR, Ki67, IFNγ, and IFNγ in PD1 + and LAG3 + , PD1 + , LAG3 + , TEMRA, and CD95) and immune regulation (Tregs, CTLA4 in Tregs, MFI of CTLA4 in Tregs, TIGIT in Tregs, ICOS in Tregs, and CTLA4 in CD4 + T-cells) in CVIDid. We investigated these in five patients with GLILD [fig_ref] Figure 6: Patterns of CD4 + T-cell activation [/fig_ref] , ranging from stably mild radiographic GLILD without lung function deterioration (patients 1 and 2), to clinical GLILD that required immunosuppressive treatment shortly after sampling (patients 4 and 5). In patients 1-4, immune activation increased with severity of disease. In patient 3, who at the time of sampling had subclinical GLILD that exacerbated 3 years later, as well as patients 4 and 5, the median fluorescence intensity of CTLA4 on Tregs decreased, which was not observed for the two subclinical GLILD patients that remained stable (1 and 2). Patient 4 with STAT1 GoF, who required treatment for GLILD shortly after sampling, showed high levels of immune activation and PD1 expression, and low Treg functional markers. Finally, patient 5 with CTLA4 haploinsufficiency demonstrated low CTLA4 expression on Tregs as expected, but also low T-cell activation (HLA-DR, Ki67, and IFNγ) in peripheral blood, despite progressive GLILD and the need for treatment shortly after sampling.
# Discussion
In this study, we showed that T-cells in patients with CVIDid were more often activated, proliferating, and Th1-skewed than those of patients with CVIDio. In addition, more CVIDid T-cells expressed immune co-inhibitory receptors PD1, LAG3, CTLA4, ICOS, and TIGIT, and these cells retained their inflammatory properties. Chronic activation was observed in both the CD4 + and CD8 + compartment. In the Treg compartment, we observed low CTLA4 expression in CVIDid, while ICOS expression remained intact. Finally, CVIDid T-cells showed increased migratory capacities to mucosal tissues.
High T-cell activation and Th-1 skewing in CVIDid were consistent with previous studies [bib_ref] Unravelling the complexity of T cell abnormalities in common variable immunodeficiency, Giovannetti [/bib_ref] [bib_ref] Plasma protein profiling reflects TH1-driven immune dysregulation in common variable immunodeficiency, Hultberg [/bib_ref]. In addition, we did not observe increased frequencies of IL17-producing T-cells, despite our previous finding of increased IL-17a in serum of the same patients sampled at the same time. This supports the hypothesis for an alternative source of IL17a production in CVIDid, such as type-3 ILCs [bib_ref] Expansion of inflammatory innate lymphoid cells in patients with common variable immune..., Cols [/bib_ref].
Previous studies have also reported increased PD1 expression in CVID, and interpreted this as a sign of functional exhaustion and impaired T-cell function [bib_ref] Exhaustion of bacteria-specific CD4 T cells and microbial translocation in common variable..., Perreau [/bib_ref] [bib_ref] Plasma protein profiling reflects TH1-driven immune dysregulation in common variable immunodeficiency, Hultberg [/bib_ref] [bib_ref] Common variable immunodeficiency patients with a phenotypic profile of immunosenescence present with..., Stuchlý [/bib_ref]. However, we observed that the PD1-and LAG3expressing cells in CVIDid retained the capacity to produce pro-inflammatory cytokines and to proliferate, and thus were not functionally exhausted. Upregulation of negative regulators of co-stimulation has been suggested to be a mechanism to limit inflammation-related damage to tissues in settings of chronic inflammation, while maintaining the ability to respond to pathogens [bib_ref] T cell differentiation in chronic infection and cancer: functional adaptation or exhaustion?, Speiser [/bib_ref]. In CVIDid, however, it is possible that this compensatory response is insufficient in severe states of immune dysregulation and that these chronically activated cells still contribute to immune dysregulation-related pathology.
In addition to this chronically activated T-cell state, we observed a decreased ability of Tregs to upregulate CTLA4, while CTLA4 expression was increased in the whole CD4 + population. Expression of CTLA4 by Tregs is an important mechanism by which Tregs mediate their suppressive function [bib_ref] Treg and CTLA-4: two intertwining pathways to immune tolerance, Walker [/bib_ref]. Clinical CTLA4 haploinsufficiency often results in a CVIDid phenotype with hypogammaglobulinemia and autoimmune disease, and functional Treg dysfunction has been described. In this study, the T-cell profile of the patient with CTLA4 haploinsufficiency was often not very different from the other non-genetic CVIDid patients. Therefore, the expression of CTLA4 in Tregs of non-genetic CVIDid patients may be relevant to the overall underlying pathophysiology of CVIDid and warrants further research. A recent study shows that abatacept, a CTLA4 fusion protein, was safe and effective in the treatment of CVIDid with interstitial lung disease [bib_ref] Abatacept use is associated with steroid dose reduction and improvement in fatigue..., Spee-Mayer C Von [/bib_ref]. As the population of CVIDid patients with low CTLA4 + CD27 + Tregs represented a mix of patients with pulmonary inflammation but also other organ-specific autoimmunity (data not shown), abatacept may be efficacious in other CVIDid patients as well. In addition, longitudinal monitoring of CTLA4 Treg expression in CVIDid may indicate whether it can be used as a biomarker for disease exacerbation and/or therapeutic response.
Despite these overall differences between CVIDid and CVIDio, the heterogeneity within the CVIDid group was substantial. Subgroup analyses of patients with organspecific autoimmunity did not yield insightful patterns, except that GLILD patients were often more extreme in A all observed differences (data not shown). In addition, disease severity did not always reflect immune activation. For example, the patient with CTLA4 haploinsufficiency did not show an inflammatory state in peripheral blood, while the patient that had the strongest IFNγ signature combined with low immune regulation markers (Supplementary [fig_ref] Figure 1: General description of T-cell subsets in CVID [/fig_ref] was clinically stable and did not require immunosuppressive therapy. It is possible that the peripheral blood immune phenotype only gives limited information that is clinically relevant, as T-cells may have migrated to the inflamed tissues in the sickest patients. In addition, this study may be limited by sample size to detect differences between subgroups of CVIDid. One other aspect that this study does not address is differences in absolute T-cell numbers. A recent study showed that while absolute T-cells are often lower in CVID, they do not differ between CVIDio and CVIDid [bib_ref] Predominantly antibody-deficient patients with non-infectious complications have reduced naive B, Treg, Th17,..., Edwards [/bib_ref].
To conclude, these results indicate that CVIDid Th cells are highly activated, and that, unlike in classically exhausted cells such as originally described in chronic viral infection and cancer [bib_ref] T cell differentiation in chronic infection and cancer: functional adaptation or exhaustion?, Speiser [/bib_ref] , PD1 and LAG3 expression in CVIDid CD4 + T-cells reflects chronic activation with preserved inflammatory potential rather than functional exhaustion, similar to previous findings in human auto-immune inflammation [bib_ref] PD-1+CD8+ T cells are clonally expanding effectors in human chronic inflammation, Petrelli [/bib_ref]. Combined studies of T-cell dysfunction and circulating inflammatory proteins in peripheral blood may help predict response to T-celltargeted therapies in individual CVIDid patients. Moreover, loss of CTLA4 upregulation in activated CVIDid Tregs may be mechanistically important in maintaining the inflammatory loop in CVIDid, and warrants further research.
## Supplementary information
The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s10875-021-01084-6.
Author Contribution Study was conceived by HL, FvW, and RB. Sample collection was performed by RB, JM, PH, VD, and PE. Wet lab work was performed by RB and MvdW. Data analysis and interpretation: HL, FvW, and RB. All authors substantially contributed to the acquisition, analysis, or interpretation of the data, and all approved the final manuscript.
Funding This study was financially supported by the Wilhelmina Children's Hospital Fund (Utrecht, the Netherlands).
Data Availability Not applicable.
# Declarations
Ethics Approval Ethical approval for this study for all participants was received from the Medical Ethical Committee of the Erasmus MC University Medical Center Rotterdam, the Netherlands (METC 2013-026). Written informed consent was obtained from all patients and controls according to the Declaration of Helsinki.
[fig] Figure 1: General description of T-cell subsets in CVID. A Principal component analysis of FACS data (all panels combined). B CD4 + and CD8 + T-cells. C Naïve (CD45RA + CCR7 +), central memory (CM: CD45RA − CCR7 +), effector memory (EM: CD45RA − CCR7 −) and terminally differenti-ated effector memory cells (TEMRA: CD45RA + CCR7 −) in CD4 + T-cells. CVIDid = CVID with immune dysregulation (n = 20), CVIDio = CVID with infections only (n = 12), HC = healthy controls (n = 12). Statistics: Mann-Whitney U-test. cells, LAG3, CTLA4, ICOS, and TIGIT, but not PD1 and CD95, were similarly increased in CVIDid (Supplementary Fig. 8). [/fig]
[fig] Figure 2, Figure 3: Expression Negative regulators of immune activation in CVID. A Proportions of PD1, LAG3, CTLA4, ICOS, and TIGIT in CD4 + CD45RO + T-cells. B Percentage and median fluorescence intensity of CD95 (FAS-L) in naïve (CD45RO −) and effectormemory (CD45RO +) CD4 + T-cells. C Comparison of IFNγ + and Ki67 + cells in PD1-and LAG3-positive and negative populations. Only samples with > 50 events in the PD1/LAG3-positive and PD1/LAG3-negative populations were included. D Spearman correlation between PD1 and IL10, CXCL9, CXCL10, or CXCL11. CVIDid = CVID with immune dysregulation (n = 20), CVIDio = CVID with infections only (n = 12), HC = healthy controls (n = 12). Statistics (A&C): Mann-Whitney U-test. Statistics B: paired Wilcoxon-Rank test. *p < 0.05, **p < 0.01, ***p < 0.001 [/fig]
[fig] Figure 4: Regulatory T-cells in CVID. A Gating strategy and proportion of CD25 + FOXP3 + T-cells within CD4 + population. B Expression of CTLA4, ICOS, TIGIT, and T-BET in the Treg population. C Decreased fraction of CTLA4 + Tregs was confined to the CD27 + population. CVIDid = CVID with immune dysregulation (n = 20), CVIDio = CVID with infections only (n = 12), HC = healthy controls (n = 12). Statistics: Mann-Whitney U-test. *p < 0.05, **p < 0.01, ***p < 0.001 [/fig]
[fig] Figure 5: Migratory [/fig]
[fig] Figure 6: Patterns of CD4 + T-cell activation: % of HLA-DR, Ki67, IFNg, IFNg in the PD1 + population and the LAG3 + population, %PD1, %LAG3, %CD45RA + CCR7 − (TEMRA) cells, and % of CD95. And patterns of T-cell regulation: %CD25 + FOXP3 + Treg, %CTLA4 in Treg, MFI of CTLA4 in Treg, %TIGIT in Treg, %ICOS in Treg, %CTLA4 in CD4 + T-cells in five patients with GLILD. Red line indicates individual patient legends, gray-shaded area indicates the median for the CVIDio group. Axis ranges are the minimum and the maximum for that marker for the entire cohort. MFI, median fluorescence intensity; GLILD, granulomatous-lymphocytic interstitial lung disease; VUS, variant of unknown significance; ITP, idiopathic thrombocytopenia purpura; RTX, rituximab; AZA, azathioprine; MMF, mycophenolate mofetil; SCT, stem-cell transplantation; GOF, gain of function [/fig]
[table] Table 1: Cohort characteristics. HC, healthy control; CVIDio, CVID with infections only; CVIDid, CVID with immune dysregulation; IQR, interquartile range; DMARD, disease modifying anti-rheumatic drug; VUS, variant of unknown significance [/table]
|
Expression and methylation patterns partition luminal-A breast tumors into distinct prognostic subgroups
Background: Breast cancer is a heterogeneous disease comprising several biologically different types, exhibiting diverse responses to treatment. In the past years, gene expression profiling has led to definition of several "intrinsic subtypes" of breast cancer (basal-like, HER2-enriched, luminal-A, luminal-B and normal-like), and microarray based predictors such as PAM50 have been developed. Despite their advantage over traditional histopathological classification, precise identification of breast cancer subtypes, especially within the largest and highly variable luminal-A class, remains a challenge. In this study, we revisited the molecular classification of breast tumors using both expression and methylation data obtained from The Cancer Genome Atlas (TCGA). Methods: Unsupervised clustering was applied on 1148 and 679 breast cancer samples using RNA-Seq and DNA methylation data, respectively. Clusters were evaluated using clinical information and by comparison to PAM50 subtypes. Differentially expressed genes and differentially methylated CpGs were tested for enrichment using various annotation sets. Survival analysis was conducted on the identified clusters using the log-rank test and Cox proportional hazards model. Results: The clusters in both expression and methylation datasets had only moderate agreement with PAM50 calls, while our partitioning of the luminal samples had better five-year prognostic value than the luminal-A/luminal-B assignment as called by PAM50. Our analysis partitioned the expression profiles of the luminal-A samples into two biologically distinct subgroups exhibiting differential expression of immune-related genes, with one subgroup carrying significantly higher risk for five-year recurrence. Analysis of the luminal-A samples using methylation data identified a cluster of patients with poorer survival, characterized by distinct hyper-methylation of developmental genes. Cox multivariate survival analysis confirmed the prognostic significance of the two partitions after adjustment for commonly used factors such as age and pathological stage. Conclusions: Modern genomic datasets reveal large heterogeneity among luminal breast tumors. Our analysis of these data provides two prognostic gene sets that dissect and explain tumor variability within the luminal-A subgroup, thus, contributing to the advancement of subtype-specific diagnosis and treatment.
# Background
Breast cancer is a heterogeneous disease exhibiting high tumor variability in terms of the underlying biological mechanisms, response to treatment, and overall survival rate [bib_ref] Review series breast cancer -one term, many entities?, Bertos [/bib_ref]. Accurate identification of the unique biological features characterizing each subtype is pivotal for improving our understanding of the disease, identifying subtype-specific biomarkers, targeted drug development, and better prediction of response to treatment.
Originally, therapeutic decisions in breast cancer were guided by clinicopathologic parameters like tumor size, presence of lymph-node/remote metastases, and histological grade. In addition, the status of three immunohistochemistry biomarkers -estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2/ERBB2) allowed the development of targeted therapies and proved predictive of treatment response [bib_ref] Thresholds for therapies: highlights of the St Gallen International Expert Consensus on..., Goldhirsch [/bib_ref].
With the emergence of global molecular profiling techniques, large genomic datasets became available for subtype discovery using unsupervised algorithms. By this methodology, breast samples are partitioned into subgroups using clustering algorithms, such as hierarchical clustering [bib_ref] Cluster analysis and display of genome-wide expression patterns, Eisen [/bib_ref] or K-Means, and then subgroup significance is evaluated using the clinical data associated with the samples.
Initially, microarray data were used to define four molecular breast cancer subtypes (basal-like, HER2-enriched, luminal and normal-like) based on characteristic gene expression signatures in correlation with clinical data [bib_ref] Molecular portraits of human breast tumours, Perou [/bib_ref]. These molecular subtypes correlated reasonably well with the immunohistochemical biomarker-based classification. Thus, basal-like samples are mostly triple-negative (ER-/ PR-/Her2-), luminal samples are mostly ER+, and HER2enriched tumors are characterized by amplification and high expression of the HER2/ERBB2 gene [bib_ref] Gene expression profiling in breast cancer: classification, prognostication, and prediction, Reis-Filho [/bib_ref] [bib_ref] Gene-expression signatures in breast cancer, Sotiriou [/bib_ref].
Subsequent analysis conducted on a larger dataset separated the luminal subtype into two distinct subgroups named luminal-A and luminal-B. Luminal-B tumors have higher expression of proliferation genes including Ki-67, and confer worse prognosis [bib_ref] Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications, Sørlie [/bib_ref] [bib_ref] Repeated observation of breast tumor subtypes in independent gene expression data sets, Sorlie [/bib_ref] [bib_ref] Molecular portraits of breast cancer: tumour subtypes as distinct disease entities, Sørlie [/bib_ref]. Moreover, luminal-B tumors respond better to chemotherapy, while patients with luminal-A cancer gain most benefit from antiestrogen treatment [bib_ref] Strategies for subtypes-dealing with the diversity of breast cancer: highlights of the..., Goldhirsch [/bib_ref].
As the partitioning of breast tumors into five molecular subtypes has gained acceptance and popularity, several expression-based predictors have been developed. A central predictor is PAM50, which maps a tumor sample to one of the five subtypes based on the gene expression pattern of 50 genes [bib_ref] Supervised risk predictor of breast cancer based on intrinsic subtypes, Parker [/bib_ref]. Though expected to be more robust than traditional classification systems that rely only on a few biomarkers, the separation between luminal-A and luminal-B by the various predictors is not consistent, suggesting that these molecular subtypes may not represent distinct coherent sample groups [bib_ref] Breast cancer molecular profiling with single sample predictors: a retrospective analysis, Weigelt [/bib_ref].
Other attempts to classify breast tumors were based on other profiling technologies such as miRNA arrays [bib_ref] MicroRNA signatures highlight new breast cancer subtypes, Bhattacharyya [/bib_ref] [bib_ref] MicroRNA expression profiling of human breast cancer identifies new markers of tumor..., Blenkiron [/bib_ref] , copy number variations [bib_ref] Molecular characterization of breast cancer with high-resolution oligonucleotide comparative genomic hybridization array, Andre [/bib_ref] or a combination of several different technologies [bib_ref] Integrative clustering of multiple genomic data types using a joint latent variable..., Shen [/bib_ref] [bib_ref] The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups, Curtis [/bib_ref]. The various studies have different levels of agreement with the expression-based molecular subtypes, but taken together they strongly indicate the existence of additional, subtler subtypes than the PAM50 subtypes [bib_ref] A new genome-driven integrated classification of breast cancer and its implications, Dawson [/bib_ref].
Epigenetic modifications such as DNA methylation arrays, which measure the methylation status of thousands of CpG sites across the genome [bib_ref] High density DNA methylation array with single CpG site resolution, Bibikova [/bib_ref] , were also used for breast cancer classification. DNA methylation changes were shown to play a pivotal role in cancer initiation and progression [bib_ref] The fundamental role of epigenetic events in cancer, Jones [/bib_ref] [bib_ref] Epigenetics in Cancer, Esteller [/bib_ref]. Particularly, promoter hypermethylation was associated with silencing of tumor suppressor genes [bib_ref] DNA methylation and gene silencing in cancer, Baylin [/bib_ref]. Several studies associated breast cancer molecular subtypes with specific methylation patterns [bib_ref] Molecular subtypes of breast cancer are associated with characteristic DNA methylation patterns, Holm [/bib_ref] , while others showed that methylation data may reveal additional complexity not captured at the expression level, possibly identifying finer patient groups of clinical importance [bib_ref] Methylation profiling with a panel of cancer related genes: association with estrogen..., Rønneberg [/bib_ref].
The large breast cancer dataset developed and provided by The Cancer Genome Atlas project [bib_ref] Comprehensive molecular portraits of human breast tumours, Koboldt [/bib_ref] includes more than a thousand breast tumor samples characterized by various modern high-throughput genomic technologies. This dataset constitutes a significant leap forward compared to the older microarray-based data. mRNA abundance levels are measured in TCGA dataset using the RNA-Seq technology. This technology has increased sensitivity and a higher dynamic range compared to microarrays [bib_ref] The fundamental role of epigenetic events in cancer, Jones [/bib_ref] [bib_ref] Epigenetics in Cancer, Esteller [/bib_ref]. DNA-methylation arrays applied on the same samples can help decipher biological tumor variability by epigenetic modifications not manifested at the gene expression level.
The aim of this study was to improve the classification of breast tumors based on the extensive TCGA expression and methylation data that have recently become available. We utilized these datasets to revisit the current classification of breast tumors into biologically distinct subgroups. Our improved and refined classification may contribute to the precision of diagnosis and thus, to more personalized treatment.
# Methods
## Study objectives
Our initial question was whether unsupervised clustering of all TCGA breast samples using the RNA-Seq data would reconstruct the partition defined by PAM50. As the luminal samples had the highest variability in our global clustering, we also asked how the luminal samples would cluster into two groups based on the RNA-Seq data, how the resulting sample groups would compare to the PAM50 partition into luminal-A and luminal-B, and whether that partition would have a clinical advantage over the PAM50 partition of the luminal samples. Looking into the internal structure of the highly variable luminal-A samples, we asked whether this PAM50 group can be further partitioned into finer subgroups with biological distinctness and clinical significance. We then used enrichment analysis to explore the biological mechanisms underlying the new luminal-A subgroups.
We asked similar questions about breast tumor variability at the epigenetic level. We evaluated the methylation-based partition of all breast tumors, all the luminal samples and the highly heterogeneous luminal-A, and compared the resulting partitions to PAM50. To examine the biological characteristics of differentially methylated CpGs (DMCs) separating the new methylation-based luminal-A subgroups, we conducted enrichment analysis. Finally, we performed multivariate COX survival analysis to determine whether the new subgroups have independent prognostic value.
## Data acquisition and preprocessing
TCGA data on invasive carcinoma of the breast were downloaded from the UCSC Cancer Browser web site [bib_ref] The UCSC Cancer Genomics Browser: update 2015, Goldman [/bib_ref] together with accompanying clinical information. The downloaded RNA-Seq gene expression dataset (Illumina HiSeq platform, gene level RSEM-normalized [bib_ref] RSEM: accurate transcript quantification from RNA-Seq data with or without a reference..., Li [/bib_ref] , log2 transformed) included 1215 samples of which 11 samples from male patients, 8 metastatic samples, and 30 samples of unknown tissue source were filtered out. PAM50 calls (obtained directly from UNC, including PAM50 proliferation scores) were available for 1148 of the filtered samples, and were distributed as follows: 183 basal-like, 78 HER2-enriched, 534 luminal-A, 203 luminal-B and 150 normal-like.
We also downloaded DNA methylation profiles (Illumina Infinium Human Methylation 450K platform, beta values) [bib_ref] High density DNA methylation array with single CpG site resolution, Bibikova [/bib_ref] containing 872 samples of which 8 male samples, 5 metastatic samples and 19 samples of unknown tissue source were filtered out. We used only 679 tumor samples for which PAM50 calls were available, including 124 basallike, 42 HER2-enriched, 378 luminal-A and 135 luminal-B samples. Our analysis used only the 107,639 probes of the Infinium-I design type for which a gene symbol was available. This allowed us to bypass the bias of the two probe designs included on the array, to focus on differentially methylated sites that are associated with known genes, and also to reduce the number of analyzed features.
## Unsupervised analysis of the tumor samples
Unsupervised analysis of the various sample subsets was executed by clustering the samples based on the 2000 features (genes or CpGs) showing the highest variability over the samples included in each analysis. We used the K-Means clustering algorithm in Matlab (release 2015a) with correlation distance and 100 replicates from which a solution minimizing the sum of point-to-centroid distances was chosen. Due to the high variability among sample subgroups in the breast cancer datasets, reselecting the top variable genes for the analysis of each sample set (and renormalizing accordingly) is crucial to ensure use of the features most relevant to that set. Each feature was independently centered and normalized over the analyzed samples prior to clustering.
Cohort descriptions for the samples used in each analysis are provided in Additional file 1 [fig_ref] Table 1: The main characteristics distinguishing between the luminal-A subgroups, LumA-R1 and LumA-R2 [/fig_ref]
## Sample cluster enrichment and survival analysis
To evaluate the clinical relevance of the sample clusters obtained in each unsupervised analysis, we used the extensive clinical information available from TCGA for each sample. Enrichment significance of sample clusters for categorical variables (such as the PAM50 subtype or histological type) was calculated using the false discovery rate (FDR)-corrected hypergeometric test. For numeric variables (such as age, percent tumor nuclei, and others) the difference between sample groups was evaluated using the Wilcoxon rank-sum test (Mann-Whitney U test).
Survival and recurrence-free survival curves were plotted using the Kaplan-Meier estimator [bib_ref] Nonparametric estimation from incomplete observations, Kaplan [/bib_ref] and p values for the difference in survival for each group versus all other groups were calculated using the log-rank (Mantel-Haenszel) test [bib_ref] Statistical aspects of the analysis of data from retrospective studies of disease, Horwitz [/bib_ref] [bib_ref] The logrank test, Bland [/bib_ref]. Cox univariate and multivariate survival analyses were conducted using Matlab implementation; p values were corrected using FDR. The analysis and visualization scripts are publicly available as an interactive graphical tool named PROMO.
## Analysis of differentially expressed genes and gene enrichment
A list of genes that have the highest differential expression between the two RNA-Seq-based sample groups LumA-R1 and LumA-R2 was generated by applying the Wilcoxon rank-sum test on all dataset genes exhibiting non-zero variance (n = 19,913) after flooring all dataset values to 1 and ceiling to 14. We selected the 1000 genes exhibiting the most significant p values that also have a median difference of at least 0.5 (log2-transformed RSEM expression values). All genes on the list had significantly higher expression in the LumA-R2 sample group (the lowest p value was 8.1e-28).
Gene enrichment tests were performed on these 1000 genes against a background of all genes included in the rank-sum test. The Expander software suite [bib_ref] EXPANDER-an integrative program suite for microarray data analysis, Shamir [/bib_ref] [bib_ref] Expander: from expression microarrays to networks and functions, Ulitsky [/bib_ref] was used to detect significant enrichments for Gene Ontology (GO) [bib_ref] Gene Ontology: tool for the unification of biology, Ashburner [/bib_ref] , Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways [bib_ref] Kyoto encyclopedia of genes and genomes, Kanehisa [/bib_ref] , Wiki-Pathways [bib_ref] WikiPathways: building research communities on biological pathways, Kelder [/bib_ref] and chromosomal location enrichments. GO tests were also performed using the GOrilla tool [bib_ref] GOrilla: a tool for discovery and visualization of enriched GO terms in..., Eden [/bib_ref]. The list of 1000 top differentially expressed genes and detailed results of the enrichment analysis are provided in Additional file 3.
## Analysis of differentially methylated cpgs, correlation to expression and cpg enrichment
To identify CpGs that are differentially methylated between LumA-M1 and LumA-M3 samples we applied the rank-sum test on all CpGs that survived our preprocessing and also had non-zero variability in the relevant samples (n = 93,880). We then selected the 1000 CpGs that had the highest significance and a minimal median difference of 0.2 (in Beta values). All selected CpGs had significantly higher mean methylation in the LumA-M1 compared to the LumA-M3 group.
To focus on DMCs with genes that had concomitant changes in expression, we calculated Spearman correlation between each CpG and the expression profile of its associated gene based on the Illumina probe-set annotation.
The correlation values enabled the identification of 586 DMCs (rank-sum p value <0.01, median difference >0.2) negatively correlated to expression (R < -0.2) and a second smaller group of 212 DMCs positively correlated (R > 0.2) with expression.
We used the array CpG annotations provided by Illumina to calculate enrichment of each one of the three CpG lists (top 1000 DMCs, 586 negatively correlated DMCs and 212 positively correlated DMCs) for features like differentially methylated regions (DMRs), enhancer regions, UCSC RefGene groups and regulatory feature groups. Gene enrichment analysis was performed on the unique genes composing each CpG list, using the Expander and Gorilla tools as described above. Enrichment for InterPro [bib_ref] The InterPro protein families database: the classification resource after 15 years, Mitchell [/bib_ref] terms was calculated using the Database for Annotation, Visualization and Integrated Discovery (DAVID) [bib_ref] Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources, Huang [/bib_ref]. Enrichment for tumor suppressor genes was calculated by hypergeometric test based on the TSGene [bib_ref] TSGene: a web resource for tumor suppressor genes, Zhao [/bib_ref] catalog. The lists of differentially methylated CpGs in addition to detailed results of the enrichment analysis are provided in Additional files 4, 5 and 6.
# Results
## Separation of luminal-a and luminal-b samples is not reconstructed by rna-seq unsupervised analysis
We started by evaluating the global sample structure within the RNA-Seq gene expression data obtained from TCGA. We applied unsupervised analysis on both tumor (n = 1035) and normal (n = 113) breast samples using the K-Means clustering algorithm over the top 2000 variable genes. As our initial goal was to compare the resulting partition into the four intrinsic molecular types, we used K = 5 (corresponding to the four types represented by PAM50 label classes in addition to normal). The results are shown in .
The resulting clusters exhibited moderate correspondence with PAM50 labels: most basal-like, normal and HER2-enriched samples fell into three different clusters (numbers 4, 5, and 3, respectively, listed in decreasing levels of homogeneity), whereas the luminal samples exhibited much greater variability. Importantly, most luminal-A sample were split between two different clusters -a homogenous luminal-A cluster (cluster 2), and a cluster composed of a mix of luminal-A and luminal-B samples (cluster 1).
Furthermore, the samples assigned to cluster 2 exhibited a very distinct expression pattern, overexpressing 1184 genes compared to cluster 1 (out of the 1421 differentially expressed genes, see "Methods"). Cluster 1 samples overexpressed only 229 genes compared to cluster 2 (see Additional file 1: According to these results, the variability within the luminal samples is not sufficiently captured by the PAM50 luminal-A and luminal-B subtypes. Specifically, they suggest that luminal-A samples can be further partitioned into finer subgroups, possibly having clinical meaning.
## Unsupervised partition of luminal samples predicts survival and recurrence better than pam50
To further investigate the variability among luminal samples, we clustered the 737 luminal samples (534 luminal-A and 203 luminal-B samples based on PAM50 labels) into two groups. The results are shown in . Similar to the global analysis, the luminal-A samples were divided between a luminal-A mostly homogenous cluster (cluster 2) and a cluster composed of both luminal-A and luminal-B samples (cluster 1).
Survival analysis performed on the two luminal partitions (the PAM50 luminal-A/luminal-B partition, and the two K-Means clusters shown in showed that the RNA-Seq-based clustering partition outperforms the luminal-A/luminal-B distinction in terms of both survival and recurrence (5-year survival plots are shown in ; also see Additional file 1: -2A for overall survival plots). Hence, the signal identified by our unsupervised analysis of the RNA-Seq data translates into a clinically relevant partition of the luminal samples that has better predictive power than the PAM50 luminal-A/luminal-B partition in terms of both survival and recurrence.
## Luminal-a samples have two distinct classes exhibiting clinical significance
As the luminal-A samples displayed the highest level of variability by consistently falling into two major subgroups in previous steps, we focused on this PAM50 class in an attempt to explore its underlying substructures. To this end, we re-clustered only the 534 luminal-A samples into two groups . As the resulting clusters were found to be significantly enriched for various clinical variables, we designated them as LumA-R1 (n = 258) and LumA-R2 (n = 276).
The most apparent property of the resulting partition was the general overexpression pattern in LumA-R2 samples compared to LumA-R1 samples. Indeed, out of the 2000 genes selected for clustering, 1276 were differentially expressed and 1068 of them were overexpressed in LumA-R2 samples (based on the FDR-corrected rank-sum test). A very similar partition (chi-square, p = 1.1e-40) with a parallel overexpression pattern was identified on a microarray gene expression dataset also available from TCGA for a subset of the luminal-A samples used here (n = 265). This supports the conclusion that the partition and distinct overexpression pattern we observed are not an artifact originating from RNA-Seq measurement technology or from any normalization protocols applied on the dataset (see Additional file 1, section 4).
Recurrence analysis performed on these two luminal-A subgroups identified that LumA-R2 samples were associated with a significantly reduced 5-year recurrence rate (p = 0.0076, . Enrichment analyses on additional clinical information available for the samples revealed that LumA-R1 and LumA-R2 subgroups are enriched with ductal (p = 2.1e-05) and lobular (p = 9.7e-12) histological types, respectively. LumA-R1 samples were associated with a higher proliferation score (p = 8.9e-25), older age (p = 2.6e-05), and a slight but significant decrease in normal cell percent (p = 2.8e-08) accompanied by an increase in tumor nuclei percent (p = 2.6e-12) compared with LumA-R2 samples (see [fig_ref] Table 1: The main characteristics distinguishing between the luminal-A subgroups, LumA-R1 and LumA-R2 [/fig_ref].
Comparing the luminal-A partition shown in to the groups formed when clustering all the luminal samples , we note that almost all LumA-R2 samples are contained within cluster 2 (composed of mainly Luminal-A subgroups exhibit distinct immune system expression profiles
In order to identify genes that distinguish best between LumA-R1 and LumA-R2 samples, we created a list of the 1000 most differentially expressed genes (see "Methods").
In agreement with the general expression pattern described earlier, all genes in the list were overexpressed in LumA-R2 compared to LumA-R1 samples. The most significant categories in the enrichment analysis performed in this list were related to the immune system regulation.
The more specific category of T cell receptor signaling genes appeared consistently in analyses based on various annotation databases (Gene Ontology: "T Cell activation" p = 1e-05, KEGG Pathway: "T Cell receptor signaling pathway" p = 3e-07, Wiki-Pathway: "T Cell receptor (TCR)
Signaling Pathway" p = 1.09e-07). Other enrichments of interest included the KEGG Pathways "Cytokine-cytokine receptor interaction" (p = 2.13e-13), "Chemokine signaling pathway" (p = 1.14e-09) and Wiki-Pathway "B Cell Receptor Signaling Pathway" (p = 1.72e-06). See [fig_ref] Table 2: The most enriched functional categories among the 1000 genes most differentially expressed... [/fig_ref] for a list of the most significant categories, and Additional file 1, section 5 for the full list. Careful examination of the gene list revealed that LumA-R2 samples overexpress genes that are typically expressed by various immune system cells (e.g., the leukocyte marker CD45/PTPRC, T cell marker CD3, and B cell marker CD19) [bib_ref] CD molecules 2006-human cell differentiation molecules, Zola [/bib_ref] [bib_ref] CD45: new jobs for an old acquaintance, Penninger [/bib_ref] [bib_ref] Deconstructing the form and function of the TCR/CD3 complex, Kuhns [/bib_ref] [bib_ref] Molecular mechanisms regulating CD19, CD20 and CD22 gene expression, Kehrl [/bib_ref]. A significant number of overexpressed genes are related to the T cell receptor (CD3D, CD3E, CD3G, and CD247) and the upstream part of its signaling pathway (ZAP70, LCK, FYN, LAT, a b Unsupervised analysis of luminal-A (LumA) breast samples. a Clustering of 534 RNA-Seq profiles partitions the data into two groups exhibiting distinct expression profiles. The clusters also show significant enrichment for clinical variables including recurrence, proliferation score, age, and histology. The bars below the heatmap show, from top to bottom, the partition of the samples, the designation of the samples according to the clustering of all luminal samples (see , histological type, and proliferation scores. b Five-year survival and recurrence analysis in the two luminal-A subgroups. LumA-R2 samples exhibit significantly reduced five-year recurrence rate compared with LumA-R1 PAK, and ITK) [bib_ref] Molecular mechanisms of T cell co-stimulation and co-inhibition, Chen [/bib_ref] [fig_ref] Figure 4: LumA-R2 samples overexpress genes in the T cell receptor signaling pathway [/fig_ref]. Interestingly, the overexpressed genes were related to T cell or natural killer (NK)-mediated cytotoxic activities (GZMA, GZMB, GZMH, GZMM, and PRF1) [bib_ref] Perforin and granzymes: function, dysfunction and human pathology, Voskoboinik [/bib_ref] [bib_ref] From T-cell activation signals to signaling control of anti-cancer immunity, Cronin [/bib_ref].
We also observed that the overexpression of immune receptor genes in LumA-R2 samples was accompanied by overexpression of several chemokine genes (CCL5, CCL17, CCL19, and CCL21) and their corresponding receptors (CCR5, CCR4, and CCR7). Topping the list of overexpressed genes in Lum-A-R2 samples (ranked by p value) is the Interleukin-33 (IL-33) gene, which drives T helper 2 (Th2) responses [bib_ref] Interleukin-33/ST2 axis promotes breast cancer growth and metastases by facilitating intratumoral accumulation..., Jovanovic [/bib_ref]. In summary, LumA-R2 samples exhibit better prognosis based on several clinical parameters while overexpressing a significant number of genes related to the immune system.
## Analysis of dna methylation identifies a luminal subgroup characterized by hyper-methylation and a significantly poorer outcome
The luminal-A tumors proved to be the most heterogeneous in our gene expression analysis. To further identify and characterize clinically meaningful subgroups within the luminal-A group, we explored breast tumor variability on the epigenetic level as well.
Using the Methylation 450K array dataset available from TCGA, we started our analysis as in the expression data, by clustering all 679 tumor samples into four groups, corresponding to the number of PAM50 classes. The resulting clusters [fig_ref] Figure 5: Unsupervised analysis of breast cancer tumors using DNA methylation data [/fig_ref] had modest agreement with the expression-based PAM50 classes; all basal-like samples were assigned to a single cluster exhibiting a distinct hypo-methylation pattern (cluster 4), whereas HER2enriched samples were scattered over three different clusters, indicating that this subtype has reduced manifestation at the methylation level. Notably, most luminal samples were assigned to three different clusters (1-3) with methylation-level gradation on the top 2000 variable CpGs. Cluster 1 exhibited a strong hyper-methylation pattern, contained the highest ratio of luminal-B samples, and was associated with significantly poorer survival compared to the three other clusters (p = 0.0001). Cluster 3, on the other hand, exhibited opposite characteristics: lower methylation levels, the lowest ratio of luminal-B samples and a better outcome (p = 0.0129). Similar results were obtained when we clustered only the 513 luminal A and B samples [fig_ref] Figure 5: Unsupervised analysis of breast cancer tumors using DNA methylation data [/fig_ref]. Here we used the top 2000 variable genes within these samples, to remove the effect of the other two subtypes on the clustering. Importantly, out of the 127 samples comprising the hyper-methylated cluster 1, which was associated with reduced survival (p = 2.6e-05), 76 samples were labeled as luminal-A, a subtype usually associated with good survival. In other words, approximately 20 % of the 378 luminal-A samples (as called by the expression-based PAM50) included in the analysis, could actually be assigned to a higher risk group based on methylation data (see Additional file 1, section 7 for more details).
The three-way partition by methylation levels and its association to differential survival risk also appeared when we repeated the analysis in the group of 378 luminal-A samples, using the top 2000 variable CpGs on these samples [fig_ref] Figure 5: Unsupervised analysis of breast cancer tumors using DNA methylation data [/fig_ref] Gene enrichment analysis associated these 483 genes hyper-methylated on LumA-M1 samples with GO terms related to development, signaling, cell differentiation and transcription regulation (p < 1e-15). The genes were also enriched for the homeobox InterPro term (p = 3.6e-35), in line with previous reports describing the methylation of homeobox genes during breast tumorigenesis [bib_ref] Methylation of homeobox genes is a frequent and early epigenetic event in..., Tommasi [/bib_ref] [bib_ref] Deregulated homeobox gene expression in cancer: cause or consequence?, Abate-Shen [/bib_ref] [bib_ref] The Hox genes and their roles in oncogenesis, Shah [/bib_ref]. Further, the 483 genes were enriched for tumor suppressor genes according to the TSGene catalog [bib_ref] TSGene: a web resource for tumor suppressor genes, Zhao [/bib_ref] (p = 1.5e-03), including 48 such genes (see column 1 in . Analysis for CpG features of the top 1000 DMCs revealed significant enrichment for enhancer elements, tissue-specific promoters and cancer-specific DMRs (see column 1 in [fig_ref] Table 4: Feature enrichment in the three subsets of differentially methylated CpGs in LumA-M1... [/fig_ref].
The databases Gene Ontology, InterPro and Tumor Suppressor Genes 2.0 were used to test the hypermethylated genes for enrichment. Group 1 is composed of the 1000 top differentially methylated CpGs with a mean difference of at least 0.2. All the CpGs on this list had significant hyper-methylation in the LumA-M1 samples compared to LumA-M3 samples. Group 2 is composed of the 586 CpGs with a differential methylation p value <0.01, a methylation mean difference >0.2 and Spearman-based correlation with expression <0.2. Group 3 is composed of 212 CpGs with a differential methylation p value <0.01, a methylation mean difference >0.2 and Spearman-based correlation with expression >0.2.
As DNA-methylation is known to regulate gene expression and as hyper-methylation of promoters is associated with gene silencing in cancer [bib_ref] Gene silencing in cancer in association with promoter hypermethylation, Herman [/bib_ref] , we focused on LumA-M1 hyper-methylated CpGs that affect the expression of their corresponding genes. To this end, we used the RNA-Seq-based expression data available from TCGA for the same 378 analyzed samples to generate a second list of CpGs that are both hyper-methylated in LumA-M1 samples (differential methylation p < 0.01, median difference of 0.2) and that have methylation levels inversely correlated to the expression level of their corresponding gene (Spearman correlation R < -0.2). As can be seen in [fig_ref] Table 4: Feature enrichment in the three subsets of differentially methylated CpGs in LumA-M1... [/fig_ref] , the 586 CpGs that passed this filter (corresponding to 340 genes) had significant overrepresentation of upstream parts of their corresponding genes (UCSC RefGene Group: TSS and 1 st exon p < =4.4e-05) and under-representation of gene body (p = 1.43e-16) and 3'UTR (p = 5.83e-04). In terms of the regulatory feature group, these 586 CpGs had over-representation of "Promoter Associated Cell type specific" elements (p = 1.40e-04) accompanied by highly significant under-representation of "Promoter Associated" elements (p = 2.94e-31), suggesting that the observed hyper-methylation pattern involves tissue-specific promoters. Among the 340 under-expressed genes containing the 586 hyper-methylated CpGs, there Gene enrichment in the three subsets of CpGs exhibiting differential methylation between the LumA-M1 and LumA-M3 subgroups [bib_ref] Review series breast cancer -one term, many entities?, Bertos [/bib_ref] (2)
Hyper-methylated CpGs Negative: R < -0. [fig_ref] Table 4: Feature enrichment in the three subsets of differentially methylated CpGs in LumA-M1... [/fig_ref] , respectively. Interestingly, the 212 LumA-M1 hyper-methylated CpGs that were positively correlated with expression (Spearman R > 0.2) had higher enrichment of development-related GO terms compared with negatively correlated CpGs ("pattern specification process" p = 1.07e-13, "embryonic morphogenesis" p = 1.05e-10, "cell fate commitment" p = 5.49e-10). In contrast to the negatively correlated CpGs, they had high over-representation of "gene body" and under-representation of "TSS" regions (UCSC RefGene Group, p = 9.48e-20 and p = 7.28e-14, respectively). For gene and CpG level enrichment for the positive correlations see column 3 in [fig_ref] Table 4: Feature enrichment in the three subsets of differentially methylated CpGs in LumA-M1... [/fig_ref] , respectively.
The differential methylation pattern distinguishing LumA-M1 from LumA-M3 samples could therefore be characterized by hundreds of CpGs that are hypermethylated in the LumA-M1 samples. Distinct subsets of these CpGs correlate negatively and positively with the expression of developmental genes.
# Cox survival analysis
In previous sections, we presented two different partitions of luminal-A tumors based on genomic profiles, with prognostic value: The LumA-R2 group (characterized by high expression of immune-related genes) was associated with a reduced chance of 5-year recurrence, whereas the LumA-M1 group (characterized by hypermethylation of CpGs located in developmental genes) was associated with poorer survival. To determine the prognostic contribution of the two partitions while adjusting for other relevant explanatory variables, we CpG enrichment tests show that hyper-methylated CpGs negatively correlated with gene expression are enriched for upstream gene parts, whereas positively correlated CpGs are enriched for the gene body. All three hyper-methylated CpG groups are enriched for informatically determined enhancer elements and experimentally determined differentially methylated regions and DNAse hypersensitive sites. The p values represent hyper-geometric-based over-representation or under-representation and are FDR corrected (significant p values are marked in bold). UTR untranslated region, DMR differentially methylated region performed multivariate Cox survival analysis on both LumA-R and LumA-M partitions (see [fig_ref] Table 5: Multivariate Cox analysis of luminal-A subgroups for five-year survival and five-year recurrence [/fig_ref]. Patients belonging to the LumA-M1 group had a 6.68-fold higher estimated 5-year death hazard compared with the other groups in the Cox multivariate model after adjustment for age, pathological stage, ER status, PR status and Her2 status. Patients belonging to the LumA-R2 group had a decreased recurrence hazard of 0.06 (that is, 94 % decrease) compared with LumA-R1 patients, after similar adjustment. The results reaffirm the independent prognostic value of the LumA-R2 and the LumA-M1 classes (see Additional file 1, section 10 for univariate analysis).
# Discussion
Gene expression profiling has become a useful tool for breast cancer classification and for direction of treatment [bib_ref] Clinical implications of the intrinsic molecular subtypes of breast cancer, Prat [/bib_ref]. Although the HER2-enriched and the basal-like subgroups are well-defined and indicative for anti-Her2 and chemotherapy treatment, respectively, the ER-positive luminal subgroup still presents a clinical challenge. In general, all luminal tumors are candidates for anti-hormonal therapy. However, some tumors within this class, often with a more proliferative potential and conferring poorer outcome, are considered for additional therapy. Accordingly, the common classification based on the molecular intrinsic subtypes divides the luminal tumors into the luminal-A tumors, which have a better outcome, and the more proliferative luminal-B tumor subgroups, which have a worse outcome. However, this classification is suboptimal for clinical decisions because the luminal tumors present a phenotypic and prognostic range rather than an exact partition to either group.
In this study, we applied unsupervised analysis on breast tumor samples using both expression and methylation profiles to reveal new genetic and epigenetic patterns that correlate with a clinical outcome, and compared them to the PAM50 subtypes. Overall, our analyses showed that the separation between luminal-A and luminal-B (as represented by PAM50 labels) is not clear-cut, but rather represents a phenotypic continuum (as previously observed [bib_ref] Breast cancer molecular profiling with single sample predictors: a retrospective analysis, Weigelt [/bib_ref] [bib_ref] Molecular classification of breast carcinoma, Alizart [/bib_ref] [bib_ref] Challenges translating breast cancer gene signatures into the clinic, Weigelt [/bib_ref]. In fact, each of the gene expression and methylation datasets used in our analysis separately enabled partitioning of the luminal samples into groups with better prognostic value than that of PAM50.
Furthermore, when we focused on the PAM50designated luminal-A samples only, the RNA-Seq expression profiles split the luminal-A samples into two subgroups . The lobular-enriched LumA-R2 sample group, characterized by a distinct gene overexpression pattern, was associated with significantly reduced recurrence risk compared with the more proliferative LumA-R1 subgroup. Interestingly, genes constituting that over-expression pattern were significantly enriched for functions related to the immune system, including the more specific enrichment of chemokines and genes of upstream T cell receptor signaling pathways. We postulate that the significantly elevated mRNA levels of immune related genes in LumA-R2 samples are indicative of increased infiltration levels of immune system cells into these tumors.
Typically, chemokines serve as ligands that by binding to their corresponding receptors, attract immune system cells to the site where they are secreted [bib_ref] Cancer and the chemokine network, Balkwill [/bib_ref] [bib_ref] Chemokine biology in cancer, Balkwill [/bib_ref]. LumA-R2 samples over-expressed several chemokines and their corresponding receptors. The simultaneous over-expression of both the chemokine CCL5 (previously found to be highly expressed by breast cancer cells [bib_ref] Elevated expression of the CC chemokine regulated on activation, normal t cell..., Luboshits [/bib_ref] and one of its receptors -CCR5 (expressed among others by CD8+ cytotoxic T cells), suggests that tumor cell-derived CCL5 attracts CD8+ cytotoxic T lymphocytes (CTLs) to LumA-R2 tumors. Similarly, the over-expressed chemokines CCL19 and CCL21 may be expressed by the tumor cells, whereas their CCR7 receptor may be expressed by licensed DC or (less typically) by naive and central memory T cells.
In line with this possibility, the over-expressed genes in LumA-R2 samples included genes typical of CTLs (and also natural killer (NK) cells), which may lead to anti-tumor cytotoxic activities exerted by the granzyme (GZMA and GZMB) and perforin pathways (PRF1). Accordingly, over-expression of T cell activation genes was also detected in patients with LumA-R2 tumors. Notably, the over-expressed genes are concentrated at the upstream part of the T cell receptor-signaling pathway Significant p values are marked in boldface. ER estrogen receptor, PR progesterone receptor, Her2 human epidermal growth factor receptor 2 [fig_ref] Figure 4: LumA-R2 samples overexpress genes in the T cell receptor signaling pathway [/fig_ref]. At this stage, it is not clear why downstream effectors are not enriched in LumA-R2 samples; however, it is of interest to see that the alpha chain of IL-15R was over-expressed in these samples, suggesting that T cell activation processes may indeed come into effect in this subgroup of patients. How could the over-expression of the immune genes by LumA-R2 samples be related, if at all, to reduced tumor recurrence? It is possible that only LumA-R2 tumors can release chemoattractants that induce the migration of antigen-specific, possibly beneficial, leukocyte subpopulations to the tumor site. Despite recent reports associating tumor infiltrating lymphocytes with a better prognosis [bib_ref] The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an..., Salgado [/bib_ref] [bib_ref] Tumorassociated lymphocytes as an independent predictor of response to neoadjuvant chemotherapy in..., Denkert [/bib_ref] [bib_ref] The immunogenicity of breast cancer-molecular subtypes matter, Denkert [/bib_ref] , it is yet to be determined how enhanced immunogenic activity in the LumA-R2 tumors may improve their outcome. Possibly in the future, this LumA-R2 characteristic pattern may direct emerging immune-checkpoint-related therapies [bib_ref] Prognostic and predictive immune gene signatures in breast cancer, Bedognetti [/bib_ref].
The role of epigenetic regulation in malignant processes is increasingly recognized. Indeed, our analysis of DNA methylation data partitioned the breast tumor samples into four clusters showing only moderate agreement with the expression-based PAM50 subtypes. In line with previous studies [bib_ref] Methylation profiling with a panel of cancer related genes: association with estrogen..., Rønneberg [/bib_ref] [bib_ref] A DNA methylation-based definition of biologically distinct breast cancer subtypes, Stefansson [/bib_ref] , one cluster had a hypomethylation pattern and corresponded with the PAM50 basal-like subgroup that was associated with poorer outcome. However, the luminal samples did not cluster neatly into the PAM50 luminal-A and luminal-B subgroups. Instead, three luminal clusters with increasing methylation levels were obtained (clusters 1-3 in [fig_ref] Figure 5: Unsupervised analysis of breast cancer tumors using DNA methylation data [/fig_ref] , of which the most hyper-methylated cluster was associated with significantly poorer 5-year prognosis. In fact, even when we clustered only the luminal-A samples [fig_ref] Figure 5: Unsupervised analysis of breast cancer tumors using DNA methylation data [/fig_ref] , the hyper-methylated cluster 1 (LumA-M1) was still associated with significantly poorer survival compared to the other two clusters (LumA-M2 and LumA-M3).
Notably, the top 1000 differentially methylated CpG loci, all hyper-methylated on LumA-M1 samples, had enrichment for genes involved in morphogenesis, differentiation, and developmental processes. Moreover, the CpG hypermethylation correlated with under-expression of developmental genes, including various tumor suppressor genes. Indeed, hyper-methylation of developmental genes in luminal breast tumors was previously reported [bib_ref] DNA methylation patterns in luminal breast cancers differ from nonluminal subtypes and..., Kamalakaran [/bib_ref] [bib_ref] Molecular rules governing de novo methylation in cancer, Nejman [/bib_ref] , secondary to repressive histone marks, which direct de novo methylation. Moreover, hyper-methylation was implicated in normal processes of cell aging and in tumorigenesis [bib_ref] Challenges translating breast cancer gene signatures into the clinic, Weigelt [/bib_ref]. Taken together, the methylation-based analysis suggests a poorer outcome for luminal tumors with a characteristic hyper-methylation pattern, whether in the luminal-A or in the luminal-B subgroups. The hyper-methylation-associated silencing of developmental and tumor suppressor genes may indeed explain these findings. More importantly, within the luminal-A subgroup that is generally associated with a better outcome, the hyper-methylation pattern of the LumA-M1 subgroup marks 84 samples (comprising 22 % of the 378 luminal-A samples) as a high-risk patient group that might benefit from more aggressive treatment.
Last, we showed that the sample partitions induced by the gene expression and DNA methylation patterns are related (p = 4.4e-08; see lower bar in [fig_ref] Figure 5: Unsupervised analysis of breast cancer tumors using DNA methylation data [/fig_ref] , mainly because the LumA-M3 samples that are associated with a better outcome are enriched for LumA-R2. However, our attempts to partition the luminal-A samples based on both patterns together did not yield a partition that is better than the separate partitions, in terms of survival prediction or clustering stability. This observation was confirmed by Cox multivariate analysis showing the independent prognostic contribution of each pattern to outcome prediction [fig_ref] Table 5: Multivariate Cox analysis of luminal-A subgroups for five-year survival and five-year recurrence [/fig_ref] , suggesting that gene expression and methylation hold complementary information, reflecting different aspects of the biological complexity of breast tumors.
Very recently, several novel partitions of luminal breast tumors were proposed [bib_ref] High density DNA methylation array with single CpG site resolution, Bibikova [/bib_ref] [bib_ref] The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an..., Salgado [/bib_ref] [bib_ref] Integration of genomic, transcriptomic and proteomic data identifies two biologically distinct subtypes..., Michaut [/bib_ref]. The partitions identified in this study are reinforced by partial though significant similarity to some newly defined groups. LumA-R1 and LumA-R2 clusters are enriched for the proliferative (p = 8.1e-04) and reactive-like (2.4-e04) classes of invasive lobular carcinoma (ILC), respectively, as defined in [bib_ref] Comprehensive molecular portraits of invasive lobular breast cancer, Ciriello [/bib_ref] (see Additional file 1, section 12). Furthermore, the LumA-M1 cluster is enriched (p = 1.6e-07) for the Epi-LumB group of tumors that are associated with poorer outcome, described by Stefansson et al. [bib_ref] A DNA methylation-based definition of biologically distinct breast cancer subtypes, Stefansson [/bib_ref] (named Epi-LumB, as it was largely composed of Luminal-B samples, see Additional file 1, section 13). Additional research is needed in order to consolidate the different partitions identified using different procedures into robust and meaningful categories for prognostic and diagnostic use in clinics.
# Conclusions
This study emphasizes the large heterogeneity of luminal breast tumors in general, and of luminal-A samples in particular, the inner variability of which was found to be inadequately captured by PAM50 molecular subtypes. Analysis of the RNA-Seq data revealed a partition of the luminal-A samples into groups associated with different risks of 5-year recurrence. We suggest that the overexpression of immune genes in the LumA-R2 group can be ascribed to a higher tendency of its samples to attract tumor-infiltrating lymphocytes, but this requires further research into the mechanism by which the higher infiltrates affect recurrence risk. In the DNA methylation data, a hyper-methylation pattern enriched for developmental genes defined a luminal-A subgroup that was
[fig] Figure 1, Figure 2: Global unsupervised clustering of 1148 breast samples using RNA-Seq data. Applying the K-Means algorithm using K = 5 on the RNA-Seq dataset yielded a partition exhibiting moderate agreement with PAM50 labels and the three immunohistochemical markers. Notably, luminal-A samples were split between a rather homogenous cluster 2 and cluster 1, which is composed of a mix of luminal-A and luminal-B. a K-Means clusters. b PAM50 calls. c Estrogen receptor (ER) status. d Progesterone receptor (PR) status. e Human epidermal growth factor receptor 2 (HER2) status luminal-A samples), whereas most LumA-R1 are contained within cluster 1 (composed of a mix of luminal-A and luminal-B samples) (see the second label bar in Fig. 3a). This suggests that LumA-R1 samples are more similar in their expression profile to luminal-B samples compared with LumA-R2 samples. Unsupervised analysis of luminal breast samples using RNA-Seq data. a Applying the K-Means algorithm on the 737 luminal samples using K = 2 splits the samples into two subgroups exhibiting better five-year prognostic value than the PAM50 luminal-A/luminal-B partition. b Five-year survival and recurrence for the two luminal breast cancer partitions. The partition into two RNA-Seq-based clusters outperforms PAM50 partition of the luminal samples in both survival and recurrence. P values were calculated using the log-rank test [/fig]
[fig] Figure 4: LumA-R2 samples overexpress genes in the T cell receptor signaling pathway. The list of top 1000 genes differentially expressed in LumA-R1 and LumA-R2 samples was found to be significantly enriched for the pathway genes (p = 1.3e-07). Genes marked in red are overexpressed in LumA-R2 samples. Pathway and graphics were taken from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database subgroup within the luminal-A subtype, which is distinguished by a robust hyper-methylation pattern.Analysis of differentially methylated CpGs between the LumA-M1 and LumA-M3 subgroups and their correlation to gene expressionTo uncover the biological features characterizing the distinct methylation patterns observed in the luminal-A subgroups, we examined the 1000 top DMCs (see "Methods") between the hyper-methylated LumA-M1 (n = 84) and the hypo-methylated LumA-M3 (n = 171). These two sample subgroups represent the two extremes of the methylation gradient observed in the luminal-A samples. Of note, all 1000 top DMCs (representing 483 genes) were hyper-methylated in the LumA-M1 samples compared to LumA-M3 samples. [/fig]
[fig] Figure 5: Unsupervised analysis of breast cancer tumors using DNA methylation data. Samples were clustered by K-Means based on correlation using the top 2000 variable CpGs over each sample subset. a All 679 tumors. b The 579 samples identified as luminal-A and luminal-B by PAM50 classification. c The 378 luminal A samples only. First bar below each expression matrix shows the assignment of the samples to methylation-based clusters. Second bar (a and b) shows PAM50 calls for the samples. Second bar (c) presents the RNA-Seq based LumA-R1/2 subgroups defined in Figure 3. Right panels show five-year Kaplan-Meier survival plots for the resulting groups [/fig]
[table] Table 2: The most enriched functional categories among the 1000 genes most differentially expressed between LumA-R1 and LumA-R2 samples [/table]
[table] Table 1: The main characteristics distinguishing between the luminal-A subgroups, LumA-R1 and LumA-R2 [/table]
[table] Table 4: Feature enrichment in the three subsets of differentially methylated CpGs in LumA-M1 and LumA-M3 subgroups [/table]
[table] Table 5: Multivariate Cox analysis of luminal-A subgroups for five-year survival and five-year recurrence [/table]
|
Male mammary gland development after apatinib therapy in advanced gastric cancer
Rationale: Most gastric cancer patients are diagnosed at mid-to late-stage and lose the chance of radical surgery, medical treatment is especially important to prolong the survival of patients. Apatinib mesylate, which is a small molecule vascular endothelial growth factor receptor 2 tyrosine kinase inhibitor, could be used as antiangiogenesis therapy for gastric cancer.Patient concerns: A 67-year-old man sought medical care for upper abdominal discomfort.Diagnosis: The patient was diagnosed as mixed medullary differentiated gastric adenocarcinoma, and immunohistochemistry suggested HER-2 (2+).Interventions: The patient received chemotherapy consisting of oxaliplatin combined with S-1 as first-line treatment, and targeted therapy with apatinib mesylate as second-line treatment.Outcomes: After 4 months of first-line chemotherapy, the patient received apatinib treatment immediately at a dose of 500 mg/d orally and died of cardiac arrest with 8.5 months of overall survival. During this period of targeted therapy with apatinib mesylate, this male patient suffered mammary gland development besides other common adverse reactions.Lessons: This case report is the first to report the case of male mammary gland development after oral apatinib.Abbreviations: VEFG = vascular endothelial growth factor, VEGFR-2 = vascular endothelial growth factor receptor 2.
# Introduction
Since most gastric cancer patients lose the chance of radical surgery when diagnosed at middle or late stage, medical treatment plays crucial roles in prolonged survival of patients.Apatinib mesylate blocks signal transduction by binding of vascular endothelial growth factor (VEGF) to its receptor, and could be used as antiangiogenesis therapy for multiple cancers.Although common adverse reactions of apatinib include hypertension, proteinuria, hand-foot syndrome, gastrointestinal bleeding, etc, so far there has been no reported case of male mammary gland development after oral apatinib. Here we report a case of a male gastric cancer patient suffered mammary gland development besides above-mentioned common adverse reactions of apatinib after 1month of treatment. To our knowledge, this is the first report on male mammary gland development caused by apatinib.
## Case presentation
A 67-year-old man sought medical care for upper abdominal discomfort in December 2016. Gastroscopy revealed gastric-togastric antrum lesions. Pathological examination suggested mixed medullary differentiated adenocarcinoma, and immunohistochemistry suggested HER-2(2+). On February 20, 2017, radical gastrectomy was performed. The tumor had invaded the main body of the pancreas, so palliative gastrojejunostomy was selected. Cancer cells were suspected in the postoperative peritoneal lavage fluid.
## First-line treatment
The patient was given systemic chemotherapy (March 30, 2017) consisting of oxaliplatin combined with S-1, and he declined detection of HER-2 FISH for personal reasons; III-degree myelosuppression occurred during treatment, manifesting as a decrease in white blood cells and platelet count. Oral S-1 was ceased 12 days later. On April 28, May 23, and June 19, 2017, the chemotherapeutic dose was decreased to 75%, II-degree myelosuppression occurred, and bone marrow suppression recovered slowly. On September 9, abdominal computed tomography was performed: the right posterior lobe of the liver showed a low-density lesion 4.5Â2.1 cm in size, and the enhanced scan showed a ring-shaped enhancement, suggesting a metastatic tumor. The assessment was for progressive disease (liver metastasis) with a progression-free survival of 4 months.
## Second-line treatment
Considering it is still difficult to recover bone marrow suppression after reducing the dose of chemotherapy this patient previously received, targeted therapy was selected for the secondline treatment. The targeted therapy was apatinib mesylate tablets at an initial dose of 500 mg/d (August 10 through October 20, 2017); 1 month after the beginning of treatment (September 13, 2017), a review of abdominal computed tomography findings showed gastric cancer, the range of postoperative gastrojejunostomy was relative stable, and the range of low density lesion of the right-posterior lobe of the liver was reduced to 3.9 Â 1.5 cm; the efficacy was classified as partial remission according to the RESIST 1.0 standard. The patient's conscious bilateral breast pain was obvious on September 10, and the bilateral mammary gland was palpable, hard lumps, and tenderness. Male mammary gland development was diagnosed according to the breast ultrasound and breast magnetic resonance imaging results, and the left and right ranges were 2.3 Â 0.9 cm, 1.5 Â 0.7 cm, respectively. The patient reported skin itching; urine routine prompts urine protein (3+); blood routine tips for white blood cells 2.0 Â 10 9 /L, platelets 74 Â 10 9 /L; and thickening of the stratum corneum at the finger joints with mild pain, though he had no trouble with normal physical activity. The patient's blood pressure was 150/100 mm Hg; apatinib was stopped. The patient was given auxiliary hematopoietic and antihypertensive symptomatic treatment and then reexamined. Urine routine analysis showed urine protein (-), blood routine results returned to normal, and blood pressure was controlled under 130/80 mm Hg. The breast ultrasound result showed male mammary gland development. Oral apatinib was continued, and the dose was kept at 500 mg/d because the patient continued to tolerate it. On October 20, routine urine analysis showed urinary protein 2+, and the patient produced black stool, suggesting upper gastrointestinal bleeding. Oral apatinib tablets were ceased. On November 2, the patient suddenly had a cardiac arrest and could not be resuscitated. As of the patient's death, second-line treatment progression-free survival had lasted 2.3 months, and his overall survival was 8.5 months.
# Discussion
Apatinib mesylate is a small molecule vascular endothelial growth factor receptor 2 (VEGFR-2) tyrosine kinase inhibitor that blocks signal transduction by binding of vascular endothelial growth factor (VEGF) to its receptor, potently inhibits tumor angiogenesis, and plays an anti-tumor therapeutic role.Due to his history of high blood pressure, the patient showed hand-foot syndrome, proteinuria, itchy skin, and continued high blood pressure after 1 month of apatinib treatment. During this period, proteinuria 3+ occurred, and recovered after 10 days of stopping apatinib treatment; the patient's skin was itchy, and no drug intervention was given because the side effects were mild. Unlike other patients, this patient developed the abovementioned adverse reactions and developed male mammary gland development. Previous studies have shown that targeted anti-tumor drugs such as imatinib could induce male mammary gland development,specifically, 18% (7/38) of patients who received imatinib treatment had developed male mammary gland development in that study.Moreover, it was suggested the mechanism of imatinib inducing male mammary gland development was blocking testosterone synthesis involved in tyrosine kinases.Another case reported that kidney cancer patients who developed male gynecomastia under treatment of sunitinib, which is a multi-target receptor tyrosine kinase inhibitor, including VEGF.An experimental study used mice suggested that neutralizing VEGF bioactivity also reduced the testosterone- stimulated growth of the glandular epithelial tissue. This effect, as the previously demonstrated effect of anti-VEGF treatment on corpus luteum growth, is secondary to inhibition of vascular growth.However, we did not check the patient's testosterone levels which may provide more clues to our study. Since apatinib mesylate is also a tyrosine kinase inhibitor, it is speculated the use of this drug could be the cause of male mammary gland development in our study. As there has been few study reported in this area yet, more research would be need to verify.
## Consent
We have obtained written informed consent from the patient's daughter for publication of the patient's details.
|
Microbial generalists and specialists differently contribute to the community diversity in farmland soils
# Introduction
Microorganisms are ubiquitous and perform crucial functions in various ecosystems, such as the Earth's biogeochemical cycles [bib_ref] The microbial engines that drive Earth's biogeochemical cycles, Falkowski [/bib_ref] , human metabolism [bib_ref] Gut microorganisms, mammalian metabolism and personalized health care, Nicholson [/bib_ref] and biotechnological processes [bib_ref] Microbial ecology to manage processes in environmental biotechnology, Rittmann [/bib_ref]. Broad interest has arisen in terms of understanding and modeling how microbes influence ecosystem functions [bib_ref] Microbial diversity drives multifunctionality in terrestrial ecosystems, Delgado-Baquerizo [/bib_ref]. It is thought that microbial communities that exhibit higher diversity are more functionally stable. The broad assumption in these studies, which use microbial diversity as a predictor of ecosystem service and function stability, is that distant taxa are functionally equivalent [bib_ref] Does adding microbial mechanisms of decomposition improve soil organic matter models? A..., Lawrence [/bib_ref]. However, there is mounting evidence that bias may be introduced by using this assumption [bib_ref] Diversity is the question, not the answer, Shade [/bib_ref]. Microbial generalists and specialists impart different impacts on microbial community dynamics. Generalists exhibit broad environmental tolerances, while specialists have narrower range of habitats and specific environment fitness [bib_ref] The importance of species sorting differs between habitat generalists and specialists in..., Székely [/bib_ref]. Therefore, there is a need to parse these broad microbial groups to estimate their contribution to diversity (e.g., structure and population size), and further to better understand and predict how ecosystem functions may respond to changes in global climate.
Rare taxa have been found to greatly contribute to community stability and play a more important role in species interactions as compared to abundant taxa. This was previously indicated by the assignment of the majority of microbial network hubs to rare taxa [bib_ref] Rare taxa maintain the stability of crop mycobiomes and ecosystem functions, Xiong [/bib_ref]. This is counter-intuitive, considering that abundant taxa with more individuals have a higher probability of interaction with others and appear to impart stronger impacts on population dynamics. The contribution of generalists/specialists may be analogous with that of abundant/rare microbes in terms of species interactions. Herein, we hypothesized that (i) although existing in limited habitats, specialists may play a broader role in species interactions. This would be measurable in co-occurrence network topologies by higher degree of connexions (i.e., more links with other species in the network) and a larger number of keystone species than that exhibited by generalists. This is based on the idea that specialists are more dependent on species interactions (e.g., auxotrophy) [bib_ref] The social network of microorganisms-How auxotrophies shape complex communities, Zengler [/bib_ref]. Specialists are also expected to have stricter growth conditions that may include specific metabolic requirements while generalists are less affected or better buffered to environmental filtering [bib_ref] Microbial generalist or specialist: Intraspecific variation and dormancy potential matter, Xu [/bib_ref]. Thus, we also hypothesized that (ii) in the microbial assembly, the specialists are expected to contribute more to deterministic processes (e.g., environmental filtering), while the generalists are expected to mainly contribute to stochastic processes. In addition to these expected distinct species interactions and contrasting impacts on microbial community assembly, we also aimed to decipher the spatial biodiversity patterns impacted by generalists and specialists. Within the biogeography context, the distance-decay relationship allows to assess the community similarity changes with increasing geographical distance [bib_ref] The distance decay of similarity in biogeography and ecology, Nekola [/bib_ref] and indicates the impacts of both dispersal limitation and spatial autocorrelation [bib_ref] Towards a unification of unified theories of biodiversity, Mcgill [/bib_ref]. Due to a limited impact of dispersal limitation and environmental filtering, generalists are expected to minimize species turnover, the b-diversity, which reflects the loss or replacement of species across space or time. Hence, we further hypothesized that (iii) the presence of microbial generalists will result in a flatter slope of the distance-decay relationship while microbial specialists are expected to increase the slope.
To test these hypotheses, we used high-resolution community profiling to determine the spatial distribution of bacterial communities across farmland soil ecosystems. While there are various concepts and definitions to describe ecological specialization [bib_ref] Defining and measuring ecological specialization, Devictor [/bib_ref] [bib_ref] Many roads to bacterial generalism, Bell [/bib_ref] , herein, species were defined as 'habitat generalists' and 'habitat specialists' based on their spatial distribution [bib_ref] Microbial generalist or specialist: Intraspecific variation and dormancy potential matter, Xu [/bib_ref] [bib_ref] Metabolic flexibility allows bacterial habitat generalists to become dominant in a frequently..., Chen [/bib_ref] , rather than by functional traits. >2,000 samples from eleven countries across seven climate types were collected to determine the characteristics (e.g., roles in species interactions, community assemblies and biogeographical patterns) of microbial generalists and specialists in community dynamics. This allowed us to better understand the ecological roles of microbial generalists and specialists, laying a foundation for imparting systematic distinctions among different categories of species when modelling and predicting the fate of ecosystems.
# Materials and methods
## Data collection and species table generation
Bacterial sequencing datasets related to farmland soils were collected from 66 studies, published within 2015.01 to 2018.12, including a total of 3,086 samples (Supplementary Data) by searching terms ''microbe" and ''farmland soil" in the Web of Science database. The locations of farmlands in each study were also collected for analysis related to biogeographic patterns. Publicly accessible bacterial sequencing data, primarily from the NCBI SRA database, for the studies were downloaded. Nineteen primer pairs (8F:556R; 8F:533R; 8F:343R; 9F:530R; 28F:219R; 27F:533R; 784F:1046R; 577F:926R; 519F:926R; 515F:907R; 515F:806R; 502F:802R; 479F:888R; 341F:907R; 341F:805R; 341F:785R; 341F:534R; 338F:806R; 1106F:1378R) were identified from the datasets to perform sequencing based on Roche 454 (30.8%) or Illumina technologies (69.2%). All the data were obtained from published articles in public database (NCBI) using Sratoolkit tool. We listed all the data sources and detailed environment information in the metadata.xlsx file on Github (https:// github.com/Qicheng-Xu/Metadata-for-Generalists-and-Specialists-in-Global-Farmland-Soils).
Raw paired-end (forward and reverse) sequences in each study were merged with the ''fastq_mergepairs" function and lowquality sequences (length < 150 or quality score < 20) were filtered with ''fastq_filter" function in USEARCH software [bib_ref] UPARSE: Highly accurate OTU sequences from microbial amplicon reads, Edgar [/bib_ref]. The ''fas-tx_uniques" and ''unosie3" functions were then used to perform the dereplication and denoising (error-correction) of sequences. Singletons were discarded, as they may result from erroneous sequencing or prediction.
As the datasets include sequences targeted at different regions of the 16S rRNA gene, it's impossible to perform the analysis at the single nucleotide difference level, such as zOTU nor ASV level, and the sequences after quality control cannot be compared directly. Hence, a closed-reference workflow was employed to map these fragments to full-length 16S rRNA gene sequences.
The ''closed_ref" function was used to perform the mapping processes at a 97% identify threshold with RDP database (http:// www.drive5.com/usearch/manual/sintax_downloads.html), whereas sequences that could not be mapped were assigned to unknown parts. This RDP training set containing high quality 16S rRNA gene sequences is recommended for USEARCH by the recent study [bib_ref] Taxonomy annotation and guide tree errors in 16S rRNA databases, Edgar [/bib_ref]. When a fragment could be matched to multiple fulllength 16S rRNA sequences, the Occum's Razor principle was followed and the fragment was assigned to the candidate fulllength sequence that contained the most fragment hits. The corresponding full-length 16S rRNA gene sequence and its annotation were taken as representative sequence and taxonomic classification for further study. Mapped results of each study were merged into an integrated table with the ''merge" function in R software. This workflow [fig_ref] Figure 1: The composition of microbial communities in farmland soils and the definition of... [/fig_ref] shared similar idea with a previous study [bib_ref] Generalist species drive microbial dispersion and evolution, Sriswasdi [/bib_ref] and can solve the problem that sequences with different primers cannot be compared directly.
Samples with > 90% unassigned sequences or < 1000 sequencing depth were discarded. The remaining 2,079 samples from eleven countries were further classified into seven climate types according to their locations, namely temperate maritime climate, temperate continental climate, subtropical monsoon climate, temperate monsoon climate, tropical monsoon climate, savanna climate and tropical desert climate [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref].
## The identification of microbial generalists and specialists
The generalists and specialists were classified according to a recent study [bib_ref] Long-term chemical-only fertilization induces a diversity decline and deep selection on the..., Xu [/bib_ref]. The ANOSIM showed that the climate (global R = 0.21) was the best factor to divide the samples, compared with country (global R = 0.13) and continent (global R = 0.14) . Similar results were obtained when only considering the absence/ presence of species. Hence, each climate was considered as a unique environment to explore the distribution of species. The species table was randomly shuffled 10,000 times by preserving the richness in each sample to obtain a random background distribution. An enrichment of the observed species compared with the random distribution indicated that the species were selected by the environment or had higher ability to adapt to the environment than expected. If the species enrichment occurred in narrow environments, these species were defined as specialists (e.g., enriched in limited environments). Accordingly, generalists were species enriched in wide environments. The niche breadth of species was calculated by the ''niche.width" function in the ''spaa" package in R.
## Construction of co-occurrence networks and detection of keystone species
Microbial co-occurrence networks were constructed based on a Spearman correlation [bib_ref] Determinants of community structure in the global plankton interactome, Lima-Mendez [/bib_ref]. The P-values were multiple-testingcorrected with the Benjamini-Hochberg FDR controlling procedure [bib_ref] The control of the false discovery rate in multiple testing under dependency, Benjamini [/bib_ref] \. Direct correlation dependencies were determined using the network enhancement method [bib_ref] Network deconvolution as a general method to distinguish direct dependencies in networks, Feizi [/bib_ref]. Edges with an adjusted Pvalue < 0.05 and a score above the threshold determined by the random matrix theory (RMT) method [bib_ref] Application of random matrix theory to biological networks, Luo [/bib_ref] were retained. Networks were graphed in Gephi.
The modules, within which nodes were highly connected while with few connections outside [bib_ref] Modularity and community structure in networks, Newman [/bib_ref] , were defined with the greedy modularity optimization [bib_ref] Molecular ecological network analyses, Deng [/bib_ref]. To assess the topological roles of taxa in the networks, the nodes were classified into four categories based on the within-module connectivity (Zi) and among-module connectivity (Pi), including module hubs (Zi > 2.5), network hubs (Zi > 2.5 and Pi > 0.62), connectors (Pi > 0.62) and peripherals (Zi < 2.5 and Pi < 0.62) [bib_ref] Functional molecular ecological networks, Zhou [/bib_ref].
The structural robustness (i.e., natural connectivity) of each network was calculated to compare the stability of the networks. It is an average eigenvalue derived from network spectrum, which describes the redundancy of alternative paths [bib_ref] Optimal network topology for structural robustness based on natural connectivity, Peng [/bib_ref]. A higher robustness indicates a more stable network structure.
Bacterial assembly processes based on the null-model-framework
The estimation of assembly processes based on the null model and the calculation of nearest taxon index (NTI) were based on a subset of total samples. The samples with top 10 mapping ratios (i.e., samples with lower proportion of unknown parts) in each climate type were selected and a total of 70 samples was obtained. The NTI was calculated by the ''ses.mntd" function in ''picante" package in R.
Significant positive correlations (P < 0.05) across short phylogenetic distances (i.e., phylogenetic signals) were detected in all groups, ensuring the further use of null-model-framework [bib_ref] Quantifying community assembly processes and identifying features that impose them, Stegen [/bib_ref]. The null model was conducted as described by recent studies [bib_ref] Stochastic community assembly: Does it matter in microbial ecology?, Zhou [/bib_ref] [bib_ref] Community assembly processes of the microbial rare biosphere, Jia [/bib_ref]. This framework is based on the phylogenetic turnover, which is the evolutionary distance separating species in one community from species in another community. A null distribution of b-mean nearest taxon distance (bMNTD null ) was obtained by shuffling the species across the tips of the phylogeny. This randomization was repeated 999 times. The b-nearest taxon index (bNTI) indicates the difference between observed bMNTD and the bMNTD null . |bNTI| > 2 indicates that the observed bMNTD deviated significantly from the bMNTD null distribution (i.e., the control of deterministic processes). In this case, bNTI > +2 and bNTI < -2 represent heterogeneous selection and homogeneous selection, respectively. When |bNTI| < 2, the community diversity results from a stochastic assembly process. In this case, the Raup-Crick (RCvalue) index was calculated based on the contingency matrix for further assessment in stochastic processes. The contingency matrices were randomly shuffled by maintaining the observed species richness and the number of sequences for each sample. This random shuffle was performed for 999 times to generate a null distribution. RCvalue > 0.95 and RCvalue < À 0.95 indicate significant departures from the null distribution, and represent dispersal limitation and homogenizing dispersal, respectively. |RCvalue| < 0.95 represents the undominated scenario, which indicates that community turnover is dominated by multiple processes consisting of weak selection and dispersal, diversification, and/or drift (see [bib_ref] Estimating and mapping ecological processes influencing microbial community assembly, Stegen [/bib_ref] for details).
## Estimation of the stochastic ratio in community assembly
The modified stochasticity ratio (MST) was also one of null model based indexes and assumes that deterministic processes drive the community to be more similar or dissimilar than the null expectation. Different from the null-model-based framework mentioned above, MST reflects the contribution of stochastic processes based on relative differences between the observed situation and the null expectation, rather than the significance of the difference, and therefore can better quantitatively measure the stochasticity in assembly [bib_ref] Climate warming leads to divergent succession of grassland microbial communities, Guo [/bib_ref] [bib_ref] A general framework for quantitatively assessing ecological stochasticity, Ning [/bib_ref]. The MST index defines 0.5 as the boundary point to determine whether the community assembly is more deterministic (<0.5) or more stochastic (>0.5). However, the phylogenetic diversity was not considered as the above-mentioned null model approach. The MST was calculated based on both Bray-Curtis and Jaccard distance by using the ''NST" package in the R software environment.
## Estimated contribution of the neutral processes based on sloan neutral model
The Sloan neutral model is a neutral-theory-based approach and assumes that species are ecologically functionally equivalent and could be randomly lost and then replaced by other members in the local community and/or supplemented from the metacommunity via dispersal [bib_ref] Quantifying the roles of immigration and chance in shaping prokaryote community structure, Sloan [/bib_ref] [bib_ref] Contribution of neutral processes to the assembly of gut microbial communities in..., Burns [/bib_ref]. The relationship between the occurrence frequency of species in local communities and their relative abundance in the metacommunity was examined to estimate the potential contribution of neutral processes in assembly. Herein, the species associated with independent samples were taken to be local communities that are a part of a broader metacommunity consisting of the species associated with all of the samples in datasets. The parameters R 2 and m represent the goodness of fit and migration rate, respectively. The species with an occurrence frequency that deviated from the neutral distribution are considered to be selected for or against by environment, or have a distinct dispersal ability. The model fitting was performed in R software as previously described [bib_ref] Contribution of neutral processes to the assembly of gut microbial communities in..., Burns [/bib_ref].
## Evolutionary trends of specialists and generalists
The binary-state speciation and extinction (BiSSE) model was performed to explore the ecological roles of generalists and specialists as described [bib_ref] Generalist species drive microbial dispersion and evolution, Sriswasdi [/bib_ref]. The model considers generalists and specialists as distinct evolutionary states and calculates evolutionary rate parameters (the speciation and extinction rates), allowing for the estimation of the transformation from one ecological state to another (the transition rate between generalists and specialists). The BiSSE method creates a model in which specialist or generalist is able to become more or less abundant in one of two ways: either by differences in character state transition (i.e., specialist species evolving to become generalists or vice-versa), or through relative diversification (specialist species give rise to more or fewer descendent species than generalist species do). BiSSE then selects a combination of rate parameters that with the highest likelihood to generate the data that it is given. The phylogenetic tree is the most critical input for BiSSE model. The Living Tree Project datasets provide curated entries and the best quality sequences with manually checked alignment. The identified generalists and specialists were mapped to the All-species Living Tree (LTP tree from this wellestablished Living Tree Project) according to their accession number. A subtree was obtained by retaining the branches with mapped taxa. The subtree was linearized and reconstructed to be an ultrametric tree using the ''ape" package in R. BiSSE model was performed with the ''diversitree" package in R software [bib_ref] Comparative phylogenetic analyses of diversification in R, Fitzjohn [/bib_ref]. The ''starting.point.bisse" function was used to determine the starting point for the simulation by setting identical speciation and extinction rates. Then, a maximum likelihood method was used to estimate the evolutionary rate parameters. A Markov chain Monte Carlo (MCMC) simulation with 10,000 permutations was performed to ensure the stability of the final estimate. MCMC simulations with 1,000 and 5,000 permutations obtained similar results.
## Other statistical analyses
The a-diversity (including the Shannon and Chao index) was calculated using ''vegan" package. The ecological community thresholds were identified using threshold indicator species analyses (TITAN) in the ''TITAN2" package. TITAN divides the community into two groups: Z-taxa negatively respond to the increased gradient, and Z + taxa positively respond to the increased gradient. The taxa with no response to the environment gradient were not considered. Then, TITAN tracks cumulative responses of declining [sum(Z-)] and increasing [sum(Z + )] taxa in the community. The ecological thresholds were defined as points where the maximum aggregate change in the frequency and relative abundance of responding taxa occurs. When the environmental values reach by and exceed the ecological threshold, the abundance and occurrence frequency of species will decrease in group Z-, while increase in group Z+. Hence, the range of niche optima of the community is defined as the gradient below sum(Z-) and above sum (Z + ). The curve fitting (linear and quadratic) was performed with ''lm" function in ''vegan" package. The heatmap was plotted with ''pheatmap" package. The phylogenetic tree was constructed using Neighbor-Joining method in R based on default parameters with the multiple alignment matrix produced in Muscle, and plotted with iTOL (http://itol.embl.de/).
# Results
## The composition of generalist and specialist groups
The merged datasets provided a contingency table matching to 5,266 annotated bacterial species from 2,079 samples of satisfying sequencing depth. These samples were from eleven countries and across seven climate types [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. The bacterial communities in farmland soils were dominated by Actinobacteria (13%), Proteobacteria (11%), Acidobacteria (6%), Firmicutes (3%), Bacteroidetes (2%) and Verrucomicrobia (1%) [fig_ref] Figure 1: The composition of microbial communities in farmland soils and the definition of... [/fig_ref].
By comparing the random and observed species distribution patterns (see methods for details), species existing in only one climate type were defined as specialists, species existing in more than five climate types were defined as generalists, and the remaining were considered as opportunists [fig_ref] Figure 1: The composition of microbial communities in farmland soils and the definition of... [/fig_ref]. In total, 1,396 species ($27%) were classified as specialists and 495 species ($9%) were classified as generalists [fig_ref] Figure 1: The composition of microbial communities in farmland soils and the definition of... [/fig_ref]. Specialists and generalists accounted for $ 4% and 56% of the total relative abundance, respectively. While the richness of the generalists was approximately one-third that of specialists, the generalists were $ 16 times higher in relative abundance. In general, generalists were distributed across a wider range of samples and exhibited a higher relative abundance than the specialists. However, some specialists were found to have a high relative abundance and were identified in a wide range of samples (across one climate type), while some generalists were only found in a limited number of samples (from various climate types). Hence, the frequency of occurrence across samples and relative abundances do not necessarily dictate the ecotype. As such, the environmental heterogeneity of the species should be considered. The proportions of generalists and specialists varied across different climate types .
Whether considering species richness or relative abundance, both the generalist and specialist groups were principally within the Actinobacteria, Proteobacteria, Acidobacteria, Firmicutes, Bacteroidetes, Verrucomicrobia and Cyanobacteria [fig_ref] Figure 1: The composition of microbial communities in farmland soils and the definition of... [/fig_ref]. More generalists were identified within the Proteobacteria, Acidobacteria and Verrucomicrobia, while more specialists were related to the Actinobacteria, Firmicutes, Bacteroidetes and Cyanobacteria. Dominant generalists (species within the top 15, based on relative abundance) were classified as Arthrobacter, Blastococcus, Nocardioides, Solirubrobacter, Gp16, Gp4, Gp6, Spartobacteria genera incertae sedis, Sphingomonas, Povalibacter, Lysobacter and Bacillus [fig_ref] Figure 1: The composition of microbial communities in farmland soils and the definition of... [/fig_ref]. Dominant specialists were Sanguibacter, Conyzicola, Promicromonospora, Knoellia, Streptomyces, Lentzea, Mycobacterium, Pseudonocardia, Flavobacterium, Sediminibacterium, Fusobacterium, Acinetobacter, Klebsiella, Massilia and Sporosarcina [fig_ref] Figure 1: The composition of microbial communities in farmland soils and the definition of... [/fig_ref]. Although there may be preferences within the generalist or specialist groups to a specific phylum, it was not possible to determine ecotypes at this coarse phylogenetic level due to the high similarity of dominant phylum. However, at the genus level, there was no overlap in the dominant generalists and specialists.
The contribution of generalist and specialist groups to the cooccurrence networks
The co-occurrence network consisted of 3,619 nodes and 51,594 edges with the average degree of connexions (i.e., average links per node) at 28.51. Actinobacteria (35%), Proteobacteria (33%), Firmicutes (13%), Bacteroidetes (8%) and Acidobacteria (7%) harboured the majority (>96% totally) of links in the network [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. Compared to the mixed distribution of phyla across the network topology, there was greater separation between generalists and specialists [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. With the exception of module I, in which 27% generalist nodes and 17% specialist nodes were detected, gen-eralists and specialists were more biased to exist in different modules [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. Modules II and III were comprised of 49% and 11% generalists, respectively. In contrast, modules IV, V, and VI contained 10%, 16% and 49% specialists, respectively. When constructing a new network only containing generalists and specialists, this phenomenon of topological separation was confirmed [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. In addition, generalists and specialists simultaneously contributed to the overall network topology with similar contributions (13%, [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref].
Specialists were more likely module hubs, with a six times larger number over generalists. In contrast, generalists were more likely to be connectors . There were no network hubs detected in this network [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. These keystone species (module hubs and connectors) principally belonged to the Actinobacteria, Proteobacteria and Firmicutes [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. The assigned species were further classified into generalist, specialist and opportunist groups. The richness and relative abundance composition of different groups were provided. (b) Classification of generalists and specialists. By comparing the expected and observed distribution, species observed in one environment and more than five environments were defined as specialists and generalists, respectively. The remaining were opportunists. The niche breadth, which indicates the occurrence frequency of species, and the relative abundance of the specialists and generalists are shown. Relative abundance is shown on a logarithmic (lg) scale on the x axis. (c) Taxonomic distribution of generalists and specialists at the phylum level, based on richness (left) and relative abundance (right). The thickness of each ribbon (i.e., the links in the middle area) in the circos-plot represents the richness or relative abundance of generalist and specialist groups assigned to different phyla. (d) Phylogenetic tree of the most abundant generalists and specialists (i.e., taxa with top 15 relative abundance in generalist and specialist groups) with their occurrence in different samples. Red and grey represents the existence and absence of species, respectively. The sample climate type, taxonomic information and species ecotypes are indicated by different colours.
Generalists exhibited higher distribution at medium [bib_ref] Determinants of community structure in the global plankton interactome, Lima-Mendez [/bib_ref] [bib_ref] The control of the false discovery rate in multiple testing under dependency, Benjamini [/bib_ref] [bib_ref] Network deconvolution as a general method to distinguish direct dependencies in networks, Feizi [/bib_ref] [bib_ref] Application of random matrix theory to biological networks, Luo [/bib_ref] [bib_ref] Modularity and community structure in networks, Newman [/bib_ref] [bib_ref] Molecular ecological network analyses, Deng [/bib_ref] [bib_ref] Functional molecular ecological networks, Zhou [/bib_ref] [bib_ref] Optimal network topology for structural robustness based on natural connectivity, Peng [/bib_ref] [bib_ref] Quantifying community assembly processes and identifying features that impose them, Stegen [/bib_ref] [bib_ref] Stochastic community assembly: Does it matter in microbial ecology?, Zhou [/bib_ref] [bib_ref] Community assembly processes of the microbial rare biosphere, Jia [/bib_ref] and high (130-150) degree of connexions than the specialists [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. Generalists were also in higher distribution in low (0-4,000) and medium (8,000-12,000) betweenness values [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. These differences in degree of connexions and betweenness distributions between the generalists and specialists indicated their different contributions to network topology.
The removal of generalist or specialist groups from the original dataset caused a decline in network robustness. The decrease in the generalists truncated dataset (from 40.5 to 36.4) is slightly higher than when the specialists were removed (from 40.5 to 37.1) [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. This indicated that the removal of generalists resulted in a more sensitive, less robust network compared with the removal of specialists.
Overall, microbial generalists and specialists simultaneously contributed to the species interactions, and the absence of these two groups would decrease the robustness of the co-occurrence networks.
## The roles of generalist and specialist groups in assembly processes
The null-model-based framework showed that both deterministic and stochastic processes were important in microbial community assembly in these farmland soils. Stochastic processes (dispersal limitation: $15%, homogenizing dispersal: $28%, undominated: $13%, in total $55%) contributed more than deterministic processes (heterogeneous selection: $32%, homogeneous selection: $12%, in total $46%) at the continental scale. The removal of generalists from the original dataset resulted in an increased contribution of deterministic processes (from 45% to 53%) and dispersal limitation (from 15% to 20%) that was accompanied by a decrease in homogenizing dispersal (from 28% to 16%). Together, this indicates that the removal of generalists increased the environmental selection but limited the microbial dispersal in the community. The removal of specialists only imparted minor changes. Furthermore, compared with specialists (deterministic processes: $47%, dispersal limitation: $24%), the generalists (deterministic processes: $28%, dispersal limitation: $9%) were less controlled by deterministic processes and exhibited a higher dispersal ability (less controlled by dispersal limitation). In addition, the nearest taxon index (NTI) suggested that the specialists were more phylogenetically clustered than the generalists. The removal of specialists caused a larger decrease in NTI, leading the original community to become more phylogenetically random than with the removal of the generalists .
The relationship between the distribution and relative abundances of microbial taxa was well described by the Sloan neutral model . As such, neutral processes played an important role in microbial community assembly (R 2 = 0.51). The removal of spe-cialists slightly decreased the impacts of neutral processes (R 2 decreased from 0.51 to 0.49), whereas the removal of generalists led to a larger decrease in R 2 (from 0.51 to 0.33) and a slight decrease in estimated migration rates.
When considering the relative abundance of taxa, the modified stochasticity ratio (MST) showed that the generalists (0.43) exhibited higher stochasticity than specialists (0.13). Removal of generalists caused a decrease (from 0.45 to 0.30) in stochasticity, whereas, the removal of specialists resulted in a small increase. Similar trends among groups was observed when only considering the presence/absence of taxa, while stochasticity indexes were higher overall.
Compared with specialists (speciation rate: 3.18, extinction rate: 0.03), the generalists (speciation rate: 5.42, extinction rate: 0.04) were characterized by an approximately two-fold higher speciation rate and a slightly higher extinction rate. The transition rate from generalists to specialists (3.12) was approximately 15-fold higher than that from specialists to generalists (0.20).
Overall, three different models, the null-model-based framework, Sloan neutral model, and MST, were employed to investigate the roles of generalists and specialists in microbial assembly by assessing the relative contribution of deterministic and stochastic processes . Generalists contributed more to stochastic processes while the specialists contributed more to deterministic processes of community assembly. Generalists . The assembly and evolutionary roles of generalist and specialist groups. (a) Assembly of different groups based on the null-model-based framework. The inner circle represents the contribution of stochastic and deterministic processes to community assembly. The outer circle represents the percentage of detailed ecological processes statistically assigned to stochastic or deterministic processes. The original dataset represents the entire community, the specialist truncated represents the dataset after removing specialists, and the generalist truncated represents the dataset after removing generalists. (b) The nearest taxon index (NTI) to assess the phylogenetic structure of different groups. A higher NTI indicates that the community exhibits more phylogenetic clustering and a NTI close to 0 indicates the community is phylogenetically random. (c) Estimation of the neutral processes based on the Sloan neutral model. The parameters R 2 and m represents the goodness of fitting and migration rate, respectively. The species that occur more and less frequently than predicted are shown in yellow and green, respectively. Dashed lines represent 95% confidence intervals and the species falling within the confidence intervals are considered neutrally distributed. (d) The modified stochasticity ratio (MST) of different groups based on Bray-Curtis and Jaccard distance. The higher MST indicated the community assembly was more stochastic. (e) Evolutionary characteristics of the generalists and specialists based on the binary-state speciation and extinction (BiSSE) model. The distribution of speciation rate, extinction rate and transition rate are shown, and their average values among the specialist and generalist groups were provided. imparted greater impacts on assembly than specialists. In addition, the higher diversification rate (speciation rate minus extinction rate) and the higher transition rate of generalists indicated their important roles in community assembly, for example in the maintenance of diversity, from an evolutionary perspective .
## Geographic patterns of generalist and specialist groups
Alpha diversity indices (Shannon and Chao) showed an increasing then decreasing trend (quadratic distribution) with latitude across all groups . Diversity was highest at $ 35°latitude for the generalist group and the entire community and at $ 40°f or the specialist group in our datasets. The removal of generalists or specialists did not alter the position at which the maximum points occurred. Quadratic coefficients were used to compare the trend strength. For the Shannon index, the specialists (quadratic coefficient: 14.08) had a stronger trend compared to the generalists (quadratic coefficient: 7.33). The removal of generalists increased the trend strength with an increase in the quadratic coefficient from 11.88 to 18.55, while the removal of specialists has only a slight effect of decrease on these quadratic coefficients. Results based on the Chao index were similar to Shannon.
A distance-decay relationship in community similarity was detected across all groups . Slopes were used to compare the strength of distance decay relationship. The specialists (slope: 5.98 Â 10 -2 ) had stronger distance-decay compared with general-ists (slope: 1.50 Â 10 -2 ), and the removal of specialists slightly weakened the distance-decay of the entire community. In contrast, the removal of generalists strengthened the distance-decay with an increase in slope from 1.56 Â 10 -2 to 1.67 Â 10 -2 . The distance decay based on the presence/absence of species showed similar results.
The environmental points, where the maximum aggregate change in the species frequency and relative abundance occurs (i.e., the ecological community thresholds), were identified using threshold indicator species analyses (TITAN, see methods for details). The communities were divided into two groups: Z-taxa negatively respond to the latitude, and Z + taxa positively respond to the latitude. The range of niche optima of the community is defined as the latitude gradient below sum(Z-) and above sum (Z + ). The sum(Z-) ecological thresholds were 31.40°, 42.49°and 39.60°for specialists, generalists and the entire community, respectively. The sum(Z+) ecological thresholds were 49.15°, 43.20°and 43.31°for specialists, generalists and the entire community, respectively. Hence, compared with specialists, the generalists had a wider range of niche optima, where the accumulated relative abundance was higher.
Overall, compared with specialists, the generalists demonstrated weaker biogeographic patterns, including the relationship between latitude and a-diversity and distance-decay in community similarity, and broader environment adaptation. In addition, generalists appear to dampen the biogeographical patterns of the . The biogeographical patterns and ecological thresholds of generalist and specialist groups. (a) The relationship between latitude and alpha diversity (i.e., Shannon and Chao) in different groups. Quadratic coefficients (absolute value) are provided to compare the strength of trend. A higher quadratic coefficient indicates a stronger relationship. R 2 represents the goodness of fitting. ***P < 0.001. The original dataset represents the entire community, the specialist truncated represents the dataset after removing specialists, and the generalist truncated represents the dataset after removing generalists. (b) The distance decay in community similarity based on Bray-Curtis distance (considering relative abundance) and Sorensen distance (considering presence/absence). Slopes (absolute value) are provided to compare the strength of distance decay. A higher slope indicates a stronger distance decay. (c) The ecological community thresholds, indicating the environmental point where the maximum aggregate change in the species frequency and relative abundance occurs, for different groups identified using threshold indicator species analyses (TITAN). Blue and orange symbols represent sum values of positive (Z+) and negative (Z-) indicator taxa along with latitude. The range of niche optima is the gradient below sum(Z-) and above sum(Z+). entire community, while the specialists strengthen the biogeographical patterns.
The species interactions, assembly processes and biogeographic patterns in different climate types and the related effects of microbial generalists and specialists were also explored . Similar results were obtained.
In addition, the generalists were predicated to have larger genomes and proteomes , which is in agreement with the expectation that generalists require larger gene and protein repertoires to survive in multiple environment conditions. We also found that the G + C contents and coding density were significantly different between generalists and specialists .
# Discussion
When predicting the ecosystem service and functional stability, a critical assumption underlying many models is that distant taxa are functionally equivalent. If taxa are functionally equivalent, microbial communities with varying compositions will function in an identical manner when they are placed in the same environment. Given high species diversity and the rapid adaptability of microbes to new conditions, this functional redundancy seems plausible. However, this may not be the truth. For example, microbial communities that shared a common history with a given habitat were found to exhibit higher functional performance compared to communities foreign to that habitat, due to local specialization [bib_ref] Testing the functional significance of microbial community composition, Strickland [/bib_ref]. Hence, an improved predictive ability may be achieved by imparting systematic distinctions among different categories of microbial species. The most common classification schemes of microbial groups include the separation of fungal and bacterial taxa [bib_ref] Differences in fungal and bacterial physiology alter soil carbon and nitrogen cycling:..., Waring [/bib_ref] , active and dormant pools [bib_ref] Representation of dormant and active microbial dynamics for ecosystem modeling, Wang [/bib_ref] and generalists and specialists [bib_ref] A theoretical model of litter decay and microbial interaction, Moorhead [/bib_ref]. Among these categories, the contribution of microbial generalists and specialists to diversity has been found to be significant, though it is rarely investigated. Herein, we explored the characteristics of microbial generalists and specialists in terms of species interactions, assembly rules and biogeographic patterns.
We found that (i) generalists and specialists simultaneously contributed to interactions between species through different mechanisms [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref] ; (ii) generalists contribute more to stochastic processes in community assembly while specialists contribute more to the deterministic processes ; (iii) the existence of microbial generalists dampens microbial biogeographic patterns, with contrasting impacts by specialists . These results give insights into how these two groups contribute to community diversity and enrich the knowledge on better understanding and predicting ecosystem functions and diversity in the future.
## The implications of generalists and specialists on occurrence networks
Species interactions may affect community composition and drive the stability and distribution patterns in microbial ecology. Patterns of species interactions were highly dynamic and contingent on community composition, species density and the environmental condition [bib_ref] Metabolite cross-feeding enhances virulence in a model polymicrobial infection, Ramsey [/bib_ref]. Compared with specialists, generalists are more available due to wider distribution and higher population densities, which is thought to lead to a higher probability of interaction with others. However, recent studies in microbial ecology have shown that rare taxa may play more important role in species interactions [bib_ref] Rare taxa maintain the stability of crop mycobiomes and ecosystem functions, Xiong [/bib_ref]. As such, high availability may not determine the strength of interactions. Based on existing knowledge, it remains difficult to determine whether generalists or specialists contribute more to microbial interactions.
Our results, based on co-occurrence network analyses, showed that both the generalists and specialists had similar degree of connexions and that the removal of generalist or specialist nodes from the network resulted in a clear decrease in network robustness [fig_ref] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist... [/fig_ref]. Thus, both specialists and generalists were important in shaping interactions among species. However, the significance of species interactions may be different between the two groups. Microbial species specialize towards the consumption of common resources in order to escape competition and offset the energetic survival costs [bib_ref] The social network of microorganisms-How auxotrophies shape complex communities, Zengler [/bib_ref] [bib_ref] Beyond the black queen hypothesis, Mas [/bib_ref]. This is explained by the Black Queen Hypothesis [bib_ref] Beyond the black queen hypothesis, Mas [/bib_ref] [bib_ref] Microbial syntrophy: Interaction for the common good, Morris [/bib_ref] that adaptive genome reduction allows free-living organisms to become beneficiary of a common good produced by a helper (i.e., dependency interaction) [bib_ref] The social network of microorganisms-How auxotrophies shape complex communities, Zengler [/bib_ref]. Generalists are likely to be helpers (common good producers), and from the Black Queen Hypothesis corollary, their ecological status is supposed to be paired with higher genome sizes and possibly with genome expansion [bib_ref] Beyond the black queen hypothesis, Mas [/bib_ref]. Thus, in the face of environmental stress, generalists are supposed to be less dependent of the presence of others to grow and may pass the environmental filter more efficiently than nongeneralists. That means, species interactions seem not to be obligatory for generalists while are more likely to stabilize the community. In contrast, specialists are expected to more rely on external nutrients or metabolites produced by a different microorganism for growth, the absence of these interactions would lead to cell death.
## The implications of generalists and specialists on assembly processes
Understanding the mechanisms that underlie the assembly of microbial communities are essential to unravelling the sustainability of ecological systems [bib_ref] Patterns and processes of microbial community assembly, Nemergut [/bib_ref]. Recent studies have revealed that niche and neutral processes are not mutually exclusive; in contrast, they are complementary and work together in structuring microbial communities [bib_ref] Stochastic assembly leads to alternative communities with distinct functions in a bioreactor..., Zhou [/bib_ref]. However, the assembly processes of microbial generalists and specialists remains unexplored. Generalists, inhabiting a wide range of environments, and specialists, having a narrower habitat range, are thought to be dominated by contrasting assembly processes and thus contribute differently to the overall diversity. Generalists are thought to be favoured by environments that vary frequently, while specialists evolve in environments that remain constant in space and time [bib_ref] The experimental evolution of specialists, generalists, and the maintenance of diversity, Kassen [/bib_ref]. As such, in constantly varying natural systems, generalists appear to be more impactful on the overall community structure and function, even though they have a lower species richness.
We found that generalists contribute more to stochastic processes in community assembly while specialists contribute more to the deterministic processes . This is in agreement with a previous study that showed that the distribution of generalists was primarily determined by neutral processes due to their general indifference to variations in habitat conditions, while habitat specialists are more affected by species sorting (deterministic processes) due to their ''preferences" for certain environmental conditions [bib_ref] The importance of neutral and niche processes for bacterial community assembly differs..., Liao [/bib_ref]. At the community level, these results indicated that specialists may contribute more to the function of the ecosystem in stable environments, considering that the deterministically assembled community was found to have an increased function compared with the stochastically assembled one [bib_ref] Dispersal-based microbial community assembly decreases biogeochemical function, Graham [/bib_ref]. However, there may be another scenario that occurs within fluctuating environments. Community-level functional acclimatisation to environment changes could be obtained through a greater generalist physiological breadth through adjustments to new conditions or through the multitude of specialist physiologies as they become more abundant/active from dormant pools [bib_ref] Resilience vs. historical contingency in microbial responses to environmental change, Hawkes [/bib_ref]. A generalistspecialist trade-off would affect the potential of the community to acclimatise to new environments. For generalists, the stochastic assembly process may be an acclimatization to future possible environment changes. The large pool of generalists may expand the range of optimal niches and lead to a greater chance of a compensatory response to alterations in the environment. However, there may be a tipping point in this wide niche breadth effect, where the negative effect of containing too high a proportion of generalists begins to reduce species richness and the resource use efficiency in the ecosystem.
## The implications of generalists and specialists on biogeographic patterns
Biogeographical patterns describe the distribution of species across space or time and reveal the mechanisms maintaining and generating species diversity. Although crucial for understanding population dynamics, microbial generalists or specialists are often neglected when attempting to predict and explain spatial patterns of biodiversity. There are similarities in biogeographical patterns in macro-and microorganisms [bib_ref] Patterns and processes of microbial community assembly, Nemergut [/bib_ref] , whereas, it is important to emphasize that some classical biogeographical patterns observed for macro-organisms, such as the links between latitude and diversity, are generally not observed in microorganisms [bib_ref] Patterns and processes of microbial community assembly, Nemergut [/bib_ref] [bib_ref] The diversity and biogeography of soil bacterial communities, Fierer [/bib_ref]. This observation suggests differences in biogeography between macro-and microorganisms. The large populations of microbial generalists may distort the linkages between latitude and diversity of community members. We found that, when the generalists were absent from the entire community, the linkages between latitude and diversity as well as the distance decay relationship were weaker , which supports this assumption. Dispersal limitation and environmental filtering are two of the main forces shaping biogeographical patterns. For generalists, the higher tolerance, which was indicated by a wider ecological range of fitness , and more individuals within groups [fig_ref] Figure 1: The composition of microbial communities in farmland soils and the definition of... [/fig_ref] may result in a higher probability of dispersal and successful survival in new environments which would lead to weaker biogeographic patterns. Furthermore, it may be possible to quantify the dampening effect of microbial generalists on biogeographical patterns. To accomplish this, differences in biogeographical patterns for total communities and their generalist and specialist fractions could be compared and the differences quantified using a measure of effect size. This approach would bring new insights into predicting microbial biogeographic community patterns, where systematic distinctions among different categories of taxa were neglected before.
In addition, a subset was created to investigate the effects of environmental factors. We found that planted crops, climates, soil depths and soil properties had significant impacts on the bacterial communities . Hence, future work should take other biotic and abiotic factors into consideration when explore the microbial generalists and specialists, such as plant physiological activity. It is well known that, under stress, plants create and excrete a broad suite of secondary metabolites which microbiomes can use for signalling and food [bib_ref] Development of fungal-mediated soil suppressiveness against Fusarium wilt disease via plant residue..., Yuan [/bib_ref] [bib_ref] Harnessing rhizosphere microbiomes for drought-resilient crop production, De Vries [/bib_ref] [bib_ref] A specialized metabolic network selectively modulates Arabidopsis root microbiota, Huang [/bib_ref] [bib_ref] Pathogen-induced activation of disease-suppressive functions in the endophytic root microbiome, Carrión [/bib_ref] [bib_ref] Root exudate metabolites drive plant-soil feedbacks on growth and defense by shaping..., Hu [/bib_ref]. Some proxy (e.g., LICOR measurements, photosynthetic activity) can be used to identify possible differences of plant activity that can ultimately shape the associated microbial generalists and specialists.
# Conclusion
This global survey of microbiomes originating from farmland soils revealed the distinct contributions of generalists and specialists to microbial diversity from the perspective of species interactions, community assembly and biogeographical patterns. In stable environments, specialists contribute more to community diversity and function through their more robust deterministic processes and higher richness. However, in constantly fluctuating natural environments, generalists that confer a high degree of ecological resistance to altered environmental conditions and exhibit higher diversification and transition rates, potentially play a more important role in maintaining community and functional stability. Generalists and specialists simultaneously contributed to interactions between species through different mechanisms. Considering that specialists appear to have a dependency of species interactions in terms of survival, when facing an environment disturbance, conservation strategies should focus on microbial specialists to avoid a decrease in overall diversity. Lastly, our results support that the existence of microbial generalists dampens microbial biogeographic patterns and indicate that the contribution of both generalists and specialists should be taken into consideration when predicting global patterns of microbial diversity to likely increase the predictive power of the data analysis.
## 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: The composition of microbial communities in farmland soils and the definition of generalist and specialist groups. (a) The distribution of communities at phylum level. [/fig]
[fig] Figure 2: The co-occurrence networks, network properties and keystone species of generalist and specialist groups. The networks are based on entire communities with different colors indicating (a) phyla distribution, (b) ecotype distribution and (c) module distribution of species. (d) The network constructed with the identified generalist and specialist species (i.e., opportunists were not included). (e) Classification of keystone species. The nodes were classified into four categories including module hubs (Zi > 2.5), network hubs (Zi > 2.5 and Pi > 0.62), connectors (Pi > 0.62) and peripherals (Zi < 2.5 and Pi < 0.62). Shapes and colours indicate the node types and ecotypes, respectively. (f) The phylogenetic tree of keystone species. Branch colours indicate the phyla distribution. Shapes and colours outside the tree indicate the node types and ecotypes, respectively. (g) The distribution of degree of connexions and (h) betweenness in generalist and specialist groups. The ecotypes are represented by colours. (i) The robustness of network before or after the removal of generalists/specialists to measure the importance of the contribution of generalists/specialists to the network structure. The original dataset represented the entire community, the specialist truncated represented the dataset after removing specialists, and the generalist truncated represented the dataset after removing generalists. [/fig]
|
Complete regression of renal tumour following ligation of an accessory renal artery during repair of an abdominal aortic aneurysm
The existence of concomitant intra-abdominal pathology with abdominal aortic aneurysms is not uncommon. The optimal management is often controversial. We describe the successful treatment of a case of an abdominal aortic aneurysm (AAA) associated with a renal tumour without performing a nephrectomy. An accessory lower pole renal artery supplying the tumour was ligated at the time of open AAA repair. The lower pole renal tumour (suspected renal cell carcinoma) reduced in size dramatically and progressively on follow-up computed tomography and the patient remains well at over two years after surgery. The successful treatment of the two conditions in such a manner represents an alternative management strategy and adds to the options available in selected patients who present with challenging and unusual pathology.
## Online case report
An abdominal aortic aneurysm (AAA) is sometimes discovered incidentally when investigations are performed for other reasons, most commonly for known or suspected intra-abdominal pathology. Alternatively, other intra-abdominal pathology may be discovered incidentally during the diagnostic workup for the AAA. In any case, the clinician is faced with a dilemma when two life threatening conditions are discovered together: how best to treat the patient. A common scenario is an intra-abdominal malignancy in association with an AAA. When both conditions require operative intervention, the order in which to do so is dictated by which condition poses the greatest short-term risk to life although this may not always be clear. Repair of the AAA often takes priority as the risk of rupture in the intervening period is considerable. [bib_ref] Abdominal aortic aneurysm and associated colorectal carcinoma: a management problem, Robinson [/bib_ref] Simultaneous treatment of both conditions is also an option. However, when considering bowel resections, this is generally considered unattractive, given the enhanced risk of graft contamination and subsequent infection from bowel pathogens at the time of surgery. On the other hand, the upper urinary tract is usually sterile. A simultaneous nephrectomy at the time of AAA repair is therefore more attractive. This has been described by several authors and is an accepted treatment option for patients diagnosed with a renal tumour and AAA. 2,3 Some advocate simultaneous treatment whenever possible when dealing with renal tumours in association with AAA although clear guidance on optimal management does not exist. [bib_ref] Concurrent abdominal aortic aneurysm and urologic neoplasm: an argument for simultaneous intervention, Ginsberg [/bib_ref] Nephron sparing or partial nephrectomy is also an option in selected patients. [bib_ref] Partial nephrectomy for renal cancer: part i, Russo [/bib_ref] We describe a case of AAA repair when ligation of an accessory renal artery at the time of surgery resulted in the resolution of a lower pole renal tumour. We believe this is the first published report describing the successful treatment of the two conditions in such a manner.
## Case history
A 63-year-old man was referred for vascular surgical assessment following discovery of an AAA on ultrasonography of the abdomen. The ultrasonography had been requested as an investigation for abdominal symptoms. The 7.4cm AAA was an incidental finding. The patient had a history of chronic lower back pain (having had surgery for spondylolisthesis many years previously) and hypertension. More recently, he had been diagnosed with inflammatory bowel disease. Computed tomography (CT) was requested to define the aneurysm more accurately prior to surgical intervention. This revealed an 8.2cm infrarenal AAA with a 2cm long conical neck and an angle of 50º in the sagittal plane. There was an incidental finding of a 3cm soft tissue mass involving the lower pole of the left kidney, highly suspicious of a renal cell carcinoma [fig_ref] figure 1: Computed tomography of abdomen demonstrating left-sided lower pole renal tumour [/fig_ref].
The repair the aneurysm. A high accessory renal artery to the left lower pole was ligated during surgery. This vessel could be seen on the pre-operative CT [fig_ref] figure 2: more cranial computed tomography image demonstrating left accessory renal artery arising from... [/fig_ref]. He made an uneventful post-operative recovery and was discharged on day 7. Follow-up CT revealed initial cystic change and marked reduction in renal tumour size at three months [fig_ref] figure 3: Computed tomography of renal tumour at three months after surgery with the... [/fig_ref]. There was a further decrease in the size of the residual lesion at eight months post-operatively and the lesion was no longer identifiable on the most recent CT at 17 months following surgery [fig_ref] figure 4: Computed tomography of tumour at 17 months after surgery [/fig_ref]. The patient remains well at over two years after surgery.
# Discussion
Generally, where nephrectomy is considered the only potentially curative option, this should be offered at the time of AAA repair. [bib_ref] Concurrent abdominal aortic aneurysm and urologic neoplasm: an argument for simultaneous intervention, Ginsberg [/bib_ref] When dealing with left-sided tumours confined radiologically to the kidney, a retroperitoneal approach for both nephrectomy and AAA repair has previously been suggested as optimal whereas a transperitoneal approach was recommended for right-sided tumours or those patients with suspected intraperitoneal tumour spread. [bib_ref] Coexistent abdominal aortic aneurysm and renal carcinoma: management options, Demasi [/bib_ref] However, with the advent of laparoscopic nephrectomies and endovascular aneurysm repair (EVAR), treatment options have widened. For example, EVAR followed by a staged laparoscopic nephrectomy during the same hospital stay was described in 2009 with a good result. [bib_ref] Staged minimally invasive treatment of inflammatory abdominal aortic aneurysm and renal cell..., Pattaras [/bib_ref] Partial nephrectomy is an attractive option when anatomically possible as some kidney function will be preserved in the operated kidney. [bib_ref] Partial nephrectomy for renal cancer: part i, Russo [/bib_ref] The risks and benefits of the procedure(s) and the final management will depend on the particular features of the case and discussion with the patient. Our case is unusual in that the renal tumour was relatively small (3cm) and was confined to a lower pole that appeared to derive its blood supply from a prominent accessory renal artery that was coming off the neck of the AAA [fig_ref] figure 2: more cranial computed tomography image demonstrating left accessory renal artery arising from... [/fig_ref]. The management was discussed prior to performing surgery. There appeared to be three principle management options:
1. radical nephrectomy at the time of AAA repair 2. partial nephrectomy at the time of AAA repair or 3. ligation of accessory renal artery supplying tumour at the time of AAA repair with close monitoring of the renal mass post-operatively It was felt that the AAA posed the greatest short-term risk to the patient's health (ie management of renal tumour should not precede AAA repair). Loss of the entire kidney (ie option 1) was not essential anatomically so it was felt that the possibility of preserving renal function should be offered. However, performing a partial nephrectomy would have added considerable time to the procedure and, given the patient's general health, there were concerns about the risks of such a long anaesthetic. The decision was therefore made to ligate the lower pole vessel supplying the tumour during AAA repair with post-operative CT surveillance of the renal mass. If the renal tumour persisted, the patient would be offered a nephron sparing nephrectomy at a later date.
# Conclusions
To our knowledge, this is the first case describing the successful treatment of a renal tumour associated with an AAA by ligating an accessory renal artery at the time of AAA repair. Ligation of the vessel resulted in infarction of the lower pole with complete regression of the renal tumour and preservation of renal function. It also obviated the need for a nephrectomy and its associated risks. Detailed pre-operative anatomical imaging employing CT angiography can identify patients suitable for bespoke interventions when presenting with unusual pathology.
[fig] figure 1: Computed tomography of abdomen demonstrating left-sided lower pole renal tumour (highly suspicious of renal cell carcinoma) together with large abdominal aortic aneurysm. [/fig]
[fig] figure 2: more cranial computed tomography image demonstrating left accessory renal artery arising from origin of abdominal aortic aneurysm that was supplying the lower pole tumour, the superior aspect of which can also be visualised. [/fig]
[fig] figure 3: Computed tomography of renal tumour at three months after surgery with the vascular graft in situ and the lesion in lower pole of left kidney, showing cystic change and reduction in size (as a result of tumour necrosis following ligation of accessory renal artery). [/fig]
[fig] figure 4: Computed tomography of tumour at 17 months after surgery: The upper aspect of the tube graft with collapsed aneurismal sac surrounding it can be seen; the left lower pole renal lesion is no longer demonstrable. [/fig]
|
The osa‐miR164 target OsCUC1 functions redundantly with OsCUC3 in controlling rice meristem/organ boundary specification
The specification of the meristem/organ boundary is critical for plant development. Here, we investigate two previously uncharacterized NAC transcription factors: the first, OsCUC1, which is negatively regulated by osa-miR164c, dimerizes with the second, OsCUC3, and functions partially redundantly in meristem/organ boundary specification in rice (Oryza sativa).We produced knockout lines for rice OsCUC1 (the homolog of Arabidopsis CUC1 and CUC2) and OsCUC3 (the homolog of Arabidopsis CUC3), as well as an overexpression line for osa-miR164c, to study the molecular mechanism of boundary specification in rice.A single mutation in either OsCUC1 or OsCUC3 leads to defects in the establishment of the meristem/organ boundary, resulting in reduced stamen numbers and the fusion of leaves and filaments, and the defects are greatly enhanced in the double mutant. Transgenic plants overexpressing osa-miR164c showed a phenotype similar to that of the OsCUC1 knockout line. In addition, knockout of OsCUC1 leads to multiple defects, including dwarf plant architecture, male sterility and twisted-rolling leaves. Further study indicated that OsCUC1 physically interacts with leaf-rolling related protein CURLED LEAF AND DWARF 1 (CLD1) and stabilizes it in the nucleus to control leaf morphology.This work demonstrated that the interplay of osa-miR164c, OsCUC1 and OsCUC3 controls boundary specification in rice.
# Introduction
Most of the aerial parts of higher plants are derived from the shoot apical meristem (SAM). In the SAM, the creation of a boundary that separates a meristem/organ primordium from its surroundings is critical for organ formation. In Arabidopsis, the CUP-SHAPED COTYLEDON (CUC) genes CUC1 and CUC2, which encode a paralogous pair of NAC transcription factors and are negatively regulated by the microRNA miR164, have been shown to be involved in embryonic SAM formation and boundary specification . Since CUC1 and CUC2 are functionally redundant, neither the cuc1 nor cuc2 single mutant displays a severe seedling phenotype, while the cuc1 cuc2 double mutant shows severe cotyledon fusion and the absence of a SAM [bib_ref] Genes involved in organ separation in Arabidopsis: an analysis of the cup-shaped..., Aida [/bib_ref] [bib_ref] Shoot apical meristem and cotyledon formation during Arabidopsis embryogenesis: interaction among the..., Aida [/bib_ref] [bib_ref] The CUP-SHAPED COTYLEDON1 gene of Arabidopsis regulates shoot apical meristem formation, Takada [/bib_ref]. Another NAC gene family member, CUC3, which is a homolog of CUC1 and CUC2, also participates in the establishment of the cotyledon boundary and the shoot meristem redundantly with CUC1 and CUC2 [bib_ref] The CUP-SHAPED COTYLEDON3 gene is required for boundary and shoot meristem formation..., Vroemen [/bib_ref] [bib_ref] Arabidopsis CUP-SHAPED COTYLEDON3 regulates postembryonic shoot meristem and organ boundary formation, Hibara [/bib_ref]. Systematic sequence analysis revealed 151 putative NAC or NAC-like genes in rice (Oryza sativa L.) [bib_ref] Genome-wide analysis of NAC transcription factor family in rice, Nuruzzaman [/bib_ref]. Some of these genes have been demonstrated to be involved in different rice developmental processes. For example, NAC29/30 regulates cellulose synthesis [bib_ref] A gibberellin-mediated DELLA-NAC signaling cascade regulates cellulose synthesis in rice, Huang [/bib_ref] ; O. sativa NAC-like activated by apetala3/pistillata (OsNAP) positively regulates leaf senescence and serves as a link between abscisic acid (ABA) and leaf senescence [bib_ref] The NAC family transcription factor OsNAP confers abiotic stress response through the..., Chen [/bib_ref] [bib_ref] OsNAP connects abscisic acid and leaf senescence by fine-tuning abscisic acid biosynthesis..., Liang [/bib_ref] ; OsNAC2 not only affects plant height, shoot branching, and thus, yield [bib_ref] Overexpression of a NAC-domain protein promotes shoot branching in rice, Mao [/bib_ref] [bib_ref] OsNAC2 encoding a NAC transcription factor that affects plant height through mediating..., Chen [/bib_ref] [bib_ref] Overexpression of miR164b-resistant OsNAC2 improves plant architecture and grain yield in rice, Jiang [/bib_ref] , but also promotes leaf senescence via ABA biosynthesis [bib_ref] A rice NAC transcription factor promotes leaf senescence via ABA biosynthesis, Mao [/bib_ref] ; ONAC020, ONAC023 and ONAC026 heteromerize and play important roles in seed development [bib_ref] Three rice NAC transcription factors heteromerize and are associated with seed size, Mathew [/bib_ref]. In addition, many rice NAC genes have been shown to be involved in responses to various abiotic and biotic stresses, including ONAC048 (OsNAC6), ONAC017 (OsNAC111), ONAC122, ONAC131, ONAC054 (RICE DWARF VIRUS MULTIPLICATION 1) and ONAC068 (OsNAC4) in defense against pathogen infection [bib_ref] Functional analysis of a NAC-type transcription factor OsNAC6 involved in abiotic and..., Nakashima [/bib_ref] [bib_ref] The transcription factor OsNAC4 is a key positive regulator of plant hypersensitive..., Kaneda [/bib_ref] [bib_ref] Disruption of a novel gene for a NAC-domain protein in rice confers..., Yoshii [/bib_ref] [bib_ref] Functions of rice NAC transcriptional factors, ONAC122 and ONAC131, in defense responses..., Sun [/bib_ref] [bib_ref] OsNAC111, a blast diseaseresponsive transcription factor in rice, positively regulates the expression..., Yokotani [/bib_ref] , and ONAC022, ONAC002 (STRESS-RESPONSIVE NAC 1/OsNAC9), ONAC048 (SNAC2/OsNAC6), ONAC009 (OsNAC5), ONAC122 (OsNAC10), ONAC045, ONAC058 (OsNAP), ONAC004, ONAC060, ONAC011 and ONAC104 in abiotic stress tolerance [bib_ref] Overexpressing a NAM, ATAF, and CUC (NAC) transcription factor enhances drought resistance..., Hu [/bib_ref] [bib_ref] Characterization of transcription factor gene SNAC2 conferring cold and salt tolerance in..., Hu [/bib_ref] [bib_ref] Functional analysis of a NAC-type transcription factor OsNAC6 involved in abiotic and..., Nakashima [/bib_ref] [bib_ref] Overexpression of a NAC transcription factor enhances rice drought and salt tolerance, Zheng [/bib_ref] [bib_ref] Root-specific expression of OsNAC10 improves drought tolerance and grain yield in rice..., Jeong [/bib_ref] [bib_ref] OsNAC5 overexpression enlarges root diameter in rice plants leading to enhanced drought..., Jeong [/bib_ref] [bib_ref] Physiological mechanisms underlying OsNAC5-dependent tolerance of rice plants to abiotic stress, Song [/bib_ref] [bib_ref] The overexpression of OsNAC9 alters the root architecture of rice plants enhancing..., Redillas [/bib_ref] [bib_ref] The NAC family transcription factor OsNAP confers abiotic stress response through the..., Chen [/bib_ref] [bib_ref] Conserved miR164-targeted NAC genes negatively regulate drought resistance in rice, Fang [/bib_ref] [bib_ref] OsNAP connects abscisic acid and leaf senescence by fine-tuning abscisic acid biosynthesis..., Liang [/bib_ref] [bib_ref] Overexpression of a stressresponsive NAC transcription factor gene ONAC022 improves drought and..., Hong [/bib_ref].
Although many NAC genes have been characterized in rice, whether they are involved in meristem or organ boundary specification remains unknown. To clarify this question, we knocked out the homologues of CUC genes in rice using CRISPR/Cas9. Our results indicated that OsCUC1, which is negatively regulated by osa-miR164c, together with OsCUC3, is involved in rice meristem/organ boundary specification in a partially redundant manner. In addition, OsCUC1 physically interacts with a previously known leaf rolling related protein CURLED LEAF AND DWARF 1 (CLD1) and stabilizes it in the nucleus to control leaf morphology.
# Materials and methods
## Plant materials and growth conditions
Rice (Oryza sativa L.) plants were grown in a glasshouse at 30°C (for daytime)/25°C (for nighttime) under a 13.5 h : 8.5 h, light : dark photoperiod (>3000 lux) with 60% humidity. The genetic background of all transgenic plants used was Nipponbare. omtn4 ZH11 and omtn6 ZH11 were obtained from Biogle GeneTech Phylogenetic tree and sequence alignment Rice OsCUC1, OsCUC3 and 70 homologues from 13 species were chosen for phylogenetic analysis. Protein sequences were obtained from NCBI (https://www.ncbi.nlm.nih.gov/) according to their accession number. The sequence alignment was performed with CLUSTALX and the phylogenetic tree was constructed using MEGA7 with the neighbor-joining method and bootstrap analysis (1000 replicates). The accession numbers are listed in Supporting Information .
## Vector construction
To knock out OsCUC1 and OsCUC3, target sites were designed online (http://cbi.hzau.edu.cn/crispr/), and the optimal target sites with low off-target scores and high sgRNA scores were selected. We prepared the CRISPR/Cas9 binary constructs as described previously . To overexpress osa-miR164c, a polymerase chain reaction (PCR) fragment amplified from Nipponbare genomic DNA using primers OE-miR164cF and OE-miR164cR was cloned into the PstI/SpeI sites of binary vector pOX between the maize Ubi1 promoter and Nos terminators.
Generation of pOsCUC1::GUS was achieved by amplifying a 2 kb OsCUC1 promoter from Nipponbare genomic DNA using primers OsCUC1pgus-F and OsCUC1pgus-R, which was then inserted into pCAMBIA1300-GN at the HindIII/XbaI site. To generate pOsCUC1::OsCUC1-GUS, a PCR fragment harboring the 2 kb upstream promoter and gDNA (without a stop codon) of OsCUC1 was amplified from Nipponbare genomic DNA using primers pro-cds-gus-F and pro-cds-gus-R, and was then inserted into pCAMBIA1300-GN at the SalI/XbaI site. To generate pOsCUC1::mOsCUC1-GUS, two overlapped fragments harboring the 2 kb upstream promoter and gDNA (without a stop codon) of OsCUC1 were amplified using two primer sets: procds-gus-F/mOsCUC1-R (carrying the 7 bp mutations) and mOsCUC1-F (carrying the 7 bp mutations)/pro-cds-gus-R, respectively. After purification, the two fragments were mixed in equal molar ratios and used as the PCR templates to amplify a fragment containing the 2 kb upstream promoter and gDNA (without a stop codon) of OsCUC1, into which the osa-miR164cresistant mutation was introduced, using pro-cds-gus-F and procds-gus-R. After sequencing, the fragment was inserted into pCAMBIA1300-GN at the SalI/XbaI site.
Primer sequences for the constructions are listed in .
## Pollen viability test
Mature pollen grains from the unopened flowers were collected and immediately put on a microscope slide. A drop of IKI (iodine potassium iodide) solution was deposited onto the pollen, and the slide was covered with a coverslip. Pollen viability counts were made 5 min after the IKI solution was added to the pollen. Pollen grains exhibiting dark staining (dark red or black color) were counted as viable.
## Histochemical analysis and b-glucuronidase (gus) assay
Samples of the transgenic plants were incubated with GUS staining solution (50 mM sodium phosphate at pH 7.2, 10 mM ethylenediaminetetraacetic acid (EDTA), 0.1% Triton X-100, 2 mM of X-Gluc, 2 mM potassium ferricyanide, 2 mM potassium ferrocyanide) overnight at 37°C. After staining, the tissues were rinsed several times with ethanol, then mounted on slides and photographed. For histochemical analysis, the samples were fixed with formaldehyde alcohol acetic acid (FAA) fixation solution at 4°C overnight, followed by dehydration and embedding in paraffin (Paraplast Plus, Sigma). They were then cut into 7 lm sections with a microtome, and stained with safranine and fast green FCF. Sections were observed under bright field with a microscope (CX31; Olympus, Tokyo, Japan).
## Yeast-two-hybrid assay
The Matchmaker yeast-two-hybrid system (Clontech, Kusatsu, Japan) was used to study the interaction of OsCUC1, OsCUC3, OMTN4 and OMTN6 with CLD1. The deduced amino acid sequences of OsCUC1, OsCUC3, OMTN4 and OMTN6 were separately cloned into pGADT7 (AD) vectors, while CLD1 was inserted into pGBKT7 (BD). Yeast strain Y2HGold was transformed with bait plasmid and strain Y187 was transformed with prey plasmid. Co-transformants were plated on synthetic defined (SD)/-Leu/-Trp/-His/-Ade medium plates and SD/-Leu/-Trp/ -His/-Ade/X-a-Gal medium plates for examination of growth. Primer sequences for the constructions are listed in .
## Transactivation assay
The deduced amino acid sequences of OsCUC1 and OsCUC3 were separately cloned into pGBKT7 (BD), resulting in fusions with the GAL4 binding domain. The fusion plasmids pBD-OsCUC1 and pBD-OsCUC3 were transformed into yeast strain AH109 and plated on SD/-Trp/-His/-Ade and SD/-Trp/-His/ -Ade/X-a-Gal medium plates for examination of growth. Primer sequences for the constructions are listed in .
## Scanning electron microscopy (sem)
The samples were fixed in FAA (50% ethanol/acetic acid/ formaldehyde, 9 : 0.5 : 0.5), and then dehydrated in an ethanol series and dried by supercritical fluid drying with CO 2 . The dried samples were mounted on copper supports and sputter-coated with gold, and then observed under SEM (S-3400N; Hitachi, Tokyo, Japan).
## Bi-molecular fluorescence complementation (bifc) assay
To study the interactions of OsCUC1, OsCUC3, OMTN4, OMTN6, CLD1, CUC1 CUC2, CUC3, mCUC1 and mCUC2, the coding sequences of these genes were correspondingly cloned into p35S-Vn and p35S-Vc. The plasmids were coexpressed in rice/Arabidopsis protoplasts. The protein-protein interaction was evaluated using a confocal laser scanning microscope (TCS-SP8MP; Leica, Wetzlar, Germany). Primer sequences for the constructions are listed in .
## Subcellular localization assay
The coding sequences of OsCUC1, OsCUC3 and CLD1 (without stop codons) from Nipponbare were fused with green fluorescent protein (GFP) to generate the fusion protein. The fusion protein driven by the cauliflower mosaic virus (CaMV) 35S promoter was transcribed in rice protoplasts. The subcellular localization was evaluated using a confocal laser-scanning microscope (TCS-SP8MP; Leica). OsRac3-mCherry and Ghd7-mCherry were used as the membrane localization marker [bib_ref] Analysis of the Rac/Rop Small GTPase family in rice: expression, subcellular localization..., Chen [/bib_ref] and nuclear localization marker [bib_ref] Natural variation in Ghd7 is an important regulator of heading date and..., Xue [/bib_ref] , respectively. Primer sequences for the constructions are listed in .
## Rna extraction and quantitative real time (rt) polymerase chain reaction assay
To evaluate gene expression, total RNA was extracted using the RNeasy Plant Mini Kit (cat. no. 74904; Qiagen, Germany) following the manufacturer's instructions. First-strand cDNA was synthesized from 1 lg of total RNA using the PrimeScript RT Reagent Kit (cat. no. RR047A; Takara, Kusatsu, Japan) according to the manufacturer's instructions. ChamQ SYBR qPCR Master Mix (cat. no. Q311-01; Vazyme, Nanjing, Jiangsu, China) and a LightCycler 480II (Roche) were used for qRT-PCR, according to the manufacturers' instructions. Rice Ubiquitin (Os03g0234200) was used as the internal reference, and the level of gene expression was normalized to the Ubiquitin level. To evaluate the expression level of osa-miR164, RNA extraction and stem-loop qRT-PCR were performed as described previously . Rice miRNA U6 was used as the internal reference and the osa-miR164 level was normalized to U6. Primer sequences for the qRT-PCR are listed in .
## In situ hybridization
To prepare the probe, a 500-bp fragment of OsCUC1-specific cDNA and a 500-bp fragment of OsCUC3-specific cDNA were amplified. The probes were labeled using a DIG RNA Labeling Kit (Roche). The in situ hybridization experiments were carried out as described previously [bib_ref] In situ hybridization for mRNA detection in Arabidopsis tissue sections, Brewer [/bib_ref]. Primer sequences for the in situ hybridization are listed in .
## Co-immunoprecipitation (co-ip) assay
To confirm the protein-protein interactions of OsCUC1, OsCUC3 and CLD1, OsCUC1-GFP, OsCUC1-MYC, OsCUC3-GFP and CLD1-Flag were artificially synthesized (GenScript Biotech Corp., Nanjing, Jiangsu, China) and transcribed under the control of the cauliflower mosaic virus (CaMV) 35S promoter. Twelve hours after co-transformation of the plasmids, the rice protoplasts were harvested by centrifugation at 400 g for 5 min. The collected protoplasts were homogenized in CelLytic immunoprecipitation (IP) buffer (B7345; Sigma-Aldrich), and incubated for 30 min on ice and then centrifuged at 15 000 g for 10 min at 4°C to remove aggregates. Next, 40 ll of Protein A/G Agarose Beads (nProtein A Sepharose 4 Fast Flow; GE Healthcare, Uppsala, Sweden) were added into the protein extract, which was diluted by IP buffer. The mixture was incubated for 3 h at 4°C with gentle shaking (40-50 rpm). After centrifugation at 14 000 g for 5 min at 4°C, the supernatant was separated from the beads, and the antibodies were added to the supernatant. After overnight incubation at 4°C, 40 ll of protein A-Sepharose was added and incubated for a further 2-3 h at 4°C. The beads were collected by centrifugation at 100 g for 3 min at 4°C, and then washed five times with ice-cold washing buffer (25 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5% Triton X-100, 1 mM EDTA, and 1% protease inhibitor). The proteins were eluted from the beads by boiling in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer for 5 min and analysed by Western blotting. The original blot images are shown in [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. Primer sequences for the constructions are listed in .
## In vitro pull-down assay
To confirm the protein-protein interaction of OsCUC1 and OsCUC3, the coding sequences of OsCUC1 and OsCUC3 were New Phytologist (2021) 229: 1566-1581 Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation www.newphytologist.com
## Research
New Phytologist amplified and inserted into pGEX-6P-1 (in frame fusion with a glutathione-S-transferase (GST) tag) and pRSFDuet-MBP (in frame fusion with a maltose binding protein (MBP) tag), respectively, and then GST-OsCUC1 and MBP-OsCUC3 were amplified and cloned into pT N T (Promega). The recombinant GST-OsCUC1 and MBP-OsCUC3 were synthesized using the TNT SP6 High-Yield Wheat Germ Protein Expression System (Promega) following the manufacturer's instructions. The recombinant GST-OsCUC1 fusion protein was immobilized on GST-Binding Resin (MagneGST Glutathione Particles, Promega) and incubated with MBP-OsCUC3 for 4 h at 4°C. After incubation, the beads were washed five times in washing buffer (4.2 mM Na 2 HPO 4 , 2 mM KH 2 PO 4 , 140 mM NaCl and 10 mM KCl), and subsequently eluted using elution buffer (50 mM Tris-HCl, 50 mM glutathione). The supernatant was subjected to immunoblotting analysis with anti-MBP (TransGen, Beijing, China) and anti-GST (TransGen) antibodies.
To confirm the protein-protein interaction of OsCUC1 and CLD1, the coding sequence of CLD1 was amplified and inserted into pRSFDuet-His (in frame fusion with a His tag), and then His-CLD1 was amplified and cloned into pT N T (Promega). The recombinant His-CLD1 was synthesized using the TNT SP6 High-Yield Wheat Germ Protein Expression System (Promega) following the manufacturer's instructions. The recombinant GST-OsCUC1 fusion protein was immobilized on GST-Binding Resin (MagneGST TM Glutathione Particles; Promega) and incubated with His-CLD1 for 4 h at 4°C. After incubation, the beads were washed five times in washing buffer (4.2 mM Na 2 HPO 4 , 2 mM KH 2 PO 4 , 140 mM NaCl and 10 mM KCl), and subsequently eluted using elution buffer (50 mM Tris-HCl, 50 mM glutathione). The supernatant was subjected to immunoblotting analysis with anti-His (TransGen) and anti-GST (TransGen) antibodies.
The original blot images are shown in [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. Primer sequences for the constructions are listed in .
## Protein extraction
For an in planta CLD1 accumulation assay, total proteins were extracted from 0.2 g of 12-d-old fresh leaves of non-KO1 OsCUC1 and oscuc1-KO1. The leaves were collected and ground in liquid nitrogen, then homogenized in protein extraction buffer (50 mM Tris-HCl, pH7.5, 150 mM NaCl, 4 M urea and 1 mM phenylmethylsulfonyl fluoride). Extracts were incubated at 4°C for 1 h, then centrifuged for 30 min at maximum speed. Supernatants were collected for further analysis. To obtain the nuclear-enriched fraction, samples were collected in the same quantity (0.2 g of 12-d-old fresh leaves), and the nuclear proteins were extracted as described previously [bib_ref] Orchestration of floral initiation by APETALA1, Kaufmann [/bib_ref].
For the CLD1-GFP accumulation assay, to obtain total proteins, rice protoplasts were collected and treated with lysis buffer (cat. no. B7345; Sigma-Aldrich). After incubation at 4°C for 1 h, the extracts were centrifuged for 30 min at maximum speed. Supernatants were collected for further analysis. The nuclear proteins of the rice protoplasts were extracted using a Minute Cytoplasmic and Nuclear Kit (Invent Biotechnologies, Eden Prairie, MN, USA) according to the manufacturer's instructions.
## Protein degradation assay
The recombinant His-CLD1 was synthesized using the TNT SP6 High-Yield Wheat Germ Protein Expression System (Promega) according to the manufacturer's instructions. The equivalent amount of His-CLD1 was added to 120 ll crude total proteins (20 µg µl -1 ) extracted from indicated plants (non-KO1 OsCUC1 or oscuc1-KO1) using reaction buffer (50 mM Tris-HCl, pH 7.8, 100 mM NaCl, 0.1% (v/v) Tween 20, 10% (v/v) glycerol and 20 mM b-mercaptoethanol). The mixture was incubated at 25°C and sampled at indicated time points (0, 3, 6 and 9 h). The abundance of remaining His-CLD1 in these samples was detected by immunoblotting using an anti-His (TransGen) antibody. The original blot images are shown in [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref].
# Statistical analysis
Statistical analysis was performed by Student's t-test. P-values of < 0.05 were considered to indicate statistical significance. P-values of < 0.01 were considered to indicate statistically high significance. Statistical calculations were performed using Microsoft Excel 2010.
# Results
Knocking out OsCUC1 and OsCUC3 in rice leads to defects in meristem/organ boundary specification
To identify the OsCUC genes, we searched for homologs of CUC1, CUC2 and CUC3 in rice. Sequence alignment indicated that LOC_Os06g23650 (Os06g0344900) is the closest homolog of both CUC1 and CUC2, hereafter referred to as OsCUC1; LOC_Os08g40030 (Os08g0511200) shares the greatest similarity with CUC3, and we therefore named it OsCUC3. Both genes were previously uncharacterized. OsCUC1 encodes a 373 amino acid (aa) protein with a typical NAC domain; while OsCUC3 encodes a 340-aa NAC protein [fig_ref] Figure 2: Appearance of rice oscuc3 knockout mutants [/fig_ref]. Both OsCUC1 and OsCUC3 belong to the NAM subgroup of the NAC gene family [bib_ref] Genome-wide analysis of NAC transcription factor family in rice, Nuruzzaman [/bib_ref]. A phylogenetic tree analysis showed that OsCUC1 is the homologue of Arabidopsis CUC1, and it is also very similar to Arabidopsis CUC2, petunia NAM, tomato GOBLET, strawberry FveCUC2a, FveCUC2b, and FveCUC2c. OsCUC3 is the homologue of CUC3 in Arabidopsis and FvH4_5g12090 in strawberry [fig_ref] Figure 2: Appearance of rice oscuc3 knockout mutants [/fig_ref].
To understand the biological function of OsCUC1 and OsCUC3, we employed CRISPR/Cas9 to produce two individual knockout lines for each of these two genes [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. For OsCUC1, a 1 bp deletion for oscuc1-KO1 and a 1 bp insertion for oscuc1-KO2 (at different sites) lead to a frame-shift with premature transcription termination [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref] , while both oscuc3-KO1 and oscuc3-KO2 harbor a 1 bp insertion at different sites of the coding region resulting in a premature stop codon [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref]. In comparison with the non-KO control line, which harbored the intact OsCUC1, both homozygous oscuc1-KO lines showed multiple defects in rice development, including the fusion of leaves and filaments, decreased stamen numbers, reduced plant height and twisted-rolling leaves [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. During the vegetative growth stage, fusion could be observed in the leaf blade [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref] as well as in the whole leaf sheath (Figs 1i-l, S4c,d) of the oscuc1-KO plants. The tube-like leaf
[formula] (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (l) (m) (n) (o) (p) (q) (r) (s) (t) (u) (v) [/formula]
## Research
New Phytologist sheath physically prevented the outgrowth of the new leaf or the panicle. During the reproductive growth stage, we also observed defects in floral development in oscuc1-KO, in which the florets displayed fusions of two or more filaments [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. This fusion in the mutants indicated the function of OsCUC1 in organ boundary specification. At the same time, most of the florets showed reduced stamen numbers [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. An SEM assay showed that not all of the stamen primordia correctly formed in the early floral developmental stage in the oscuc1-KO plants, indicating that OsCUC1 also functions in meristem boundary specification [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. We did not obtain any seeds from any of the individual T 0 homozygous oscuc1-KO plants, while the non-KO plants displayed normal fertility. This result may be attributable to the development of aberrant stamens, which may lead to pollen defects in the mutant plants. The pollen viability test indicated that almost no pollen was produced in the KO plants [fig_ref] Figure 5: osa-miR164c is involved in boundary specification and leaf development by negatively regulating... [/fig_ref]. Furthermore, pollen from wild-type plants (Nipponbare) was used to pollinate the gynoecia of oscuc1-KO1. In contrast with the self-pollinated florets, the cross-pollinated florets maintained normal development in later stages and yielded fertile seeds [fig_ref] Figure S6: Phenotypes of heterozygous mutants of OsCUC1 and OsCUC3 in rice [/fig_ref]. We further investigated whether these seeds could develop normally. No defects were detected in any developmental stage in these F 1 progeny (oscuc1-KO1/+). In the F 2 population, the homozygous oscuc1-KO1 plants showed the same phenotype as the T 0 transgenic plants, while the non-KO plants and heterozygous mutants developed normally [fig_ref] Figure S6: Phenotypes of heterozygous mutants of OsCUC1 and OsCUC3 in rice [/fig_ref]. Both of the oscuc3-KO lines exhibited defects in meristem/organ boundary specification in the vegetative growth stage as well as in the reproductive growth stage, including the fusion of leaves and filaments, and defects in stamen identification [fig_ref] Figure 2: Appearance of rice oscuc3 knockout mutants [/fig_ref] [fig_ref] Table S3: Percentages of aberrant florets of oscuc1-KO and oscuc3-KO plants [/fig_ref] , which were highly similar to those of the oscuc1-KO lines. However, knocking out OsCUC3 did not lead to dwarf plant stature or twisted-rolling leaves. In addition, in contrast to the infertility of the oscuc1-KO plants, fertile seeds could be obtained from both of the oscuc3-KO lines. Similar to the oscuc1-KO1/+ plants, the heterozygous OsCUC3 mutant plants showed no defects during any developmental stages [fig_ref] Figure S6: Phenotypes of heterozygous mutants of OsCUC1 and OsCUC3 in rice [/fig_ref].
[formula] (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) (l) (m) [/formula]
The expression profiles of OsCUC1 and OsCUC3 NAC family proteins have been shown to function as transcription factors [bib_ref] GRAB proteins, novel members of the NAC domain family, isolated by their..., Xie [/bib_ref] [bib_ref] The CUP-SHAPED COTYLEDON3 gene is required for boundary and shoot meristem formation..., Vroemen [/bib_ref]. Since OsCUC1 and OsCUC3 encode NAC domain proteins, we used GFP fluorescence to determine their subcellular localization by transiently expressing OsCUC1-GFP or OsCUC3-GFP in-frame fusion proteins in rice protoplasts. The results revealed that both OsCUC1 and OsCUC3 are localized to the nucleus [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref]. In addition, to determine whether these genes have transactivation activity, OsCUC1 and OsCUC3 were fused to the GAL4 DNA binding domain and transformed into the yeast strain AH109. The yeast strains containing pBD-OsCUC1 or pBD-OsCUC3 grew in Trp-, His-and Ade-deficient SD medium while the negative control did not grow in triple-deficient SD medium [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref] , which indicated that both genes have transactivation activity. Taken together, these data indicated that OsCUC1 and OsCUC3 function as transcription factors.
We used qRT-PCR to study the expression patterns of OsCUC1 and OsCUC3 throughout plant development. OsCUC1 was highly expressed in the leaf, bract, flag leaf, and reproductive organs, especially in the stamens [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref]. In comparison to that of OsCUC1, the expression level of OsCUC3 was lower in all tested organs. However, OsCUC3 transcripts accumulated more abundantly in the vegetative organs than in the reproductive organs [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref]. To obtain the details of the expression patterns of these two genes, we carried out an in-situ hybridization assay [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref]. In the early floral developmental stage (In 5), OsCUC1 and OsCUC3 exhibited highly overlapping expression patterns, with both mRNAs accumulated mainly in the boundaries (between the leaves and between the meristems) as well as within the meristems [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. In early In 6 stage, OsCUC1 mRNA was strongly and specifically accumulated at the region where the stamen primordia will arise [fig_ref] Figure 2: Appearance of rice oscuc3 knockout mutants [/fig_ref]. Afterward, the expression of OsCUC1 became diffused and uniform in the flower organs [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref]. For OsCUC3, in early In 6 stage, its mRNA was detected in the same region where OsCUC1 mRNA accumulated [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref] ; in addition, OsCUC3 was also detected in the primordia of sterile lemmas and rudimentary glumes [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref]. In the later stages (late In 6 and In 7), in contrast to OsCUC1, OsCUC3 was mainly detected in the boundaries between the flower organs [fig_ref] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice [/fig_ref].
## Oscuc1 and oscuc3 dimerize and function in a partially redundant manner in boundary specification and sam activity maintenance
To understand the relationship between OsCUC1 and OsCUC3, we first studied the protein-protein interaction of these two proteins. A bimolecular fluorescence complementation (BiFC) assay showed that OsCUC1 and OsCUC3 could form homodimers or heterodimers [fig_ref] Figure 4: OsCUC1 and OsCUC3 dimerize and function redundantly in rice [/fig_ref]. An in vivo co-immunoprecipitation (Co-IP) assay and an in vitro pull-down assay further confirmed the dimerization between OsCUC1 and OsCUC3 [fig_ref] Figure 4: OsCUC1 and OsCUC3 dimerize and function redundantly in rice [/fig_ref]. In Arabidopsis, due to the miR164 function in negatively regulating CUC1 and CUC2, the interaction between the CUC proteins cannot be detected in protoplasts by BiFC. However, we detected that the miR164-resistant versions of CUC1 and CUC2 (mCUC1 and mCUC2) formed homodimers, while mCUC1-mCUC2, mCUC1-CUC3 and mCUC2-CUC3 formed heterodimers [fig_ref] Figure 7: OsCUC1 stabilizes CLD1 in the nucleus in rice [/fig_ref]. These results were largely consistent with previous findings [bib_ref] Temporal control of leaf complexity by miRNA-regulated licensing of protein complexes, Rubio-Somoza [/bib_ref] [bib_ref] A conserved role for CUP-SHAPED COTYLEDON genes during ovule development, Gonc ßalves [/bib_ref] , indicating the conserved protein-protein interaction behavior of CUC proteins in both rice and Arabidopsis.
Furthermore, we developed double mutant plants for these two genes. In contrast with the single mutant plants, which showed no defects during the early seedling stage [fig_ref] Figure S6: Phenotypes of heterozygous mutants of OsCUC1 and OsCUC3 in rice [/fig_ref] , the oscuc1/+ oscuc3/oscuc3 plants exhibited leaf sheath fusion and twisted-rolling leaves; these defects were greatly enhanced in the oscuc1/oscuc1 oscuc3/oscuc3 plants [fig_ref] Figure 4: OsCUC1 and OsCUC3 dimerize and function redundantly in rice [/fig_ref]. Although loss of function of OsCUC1 and OsCUC3 did not lead to defects in SAM initiation [fig_ref] Figure S8: The development of rice oscuc1 oscuc3 homozygous double mutants is arrested at... [/fig_ref]
## Research
New Phytologist plants was arrested in the early seedling stage, and they eventually died soon after that [fig_ref] Figure S8: The development of rice oscuc1 oscuc3 homozygous double mutants is arrested at... [/fig_ref]. Together, these results indicate that OsCUC1 and OsCUC3 dimerize and function in a partially redundant manner in boundary specification and SAM activity maintenance.
## Osa-mir164c is involved in organ boundary specification and leaf development by negatively regulating oscuc1
In Arabidopsis, it has been demonstrated that miR164 negatively regulates CUC1 and CUC2 [bib_ref] The early extra petals1 mutant uncovers a role for microRNA miR164c in..., Baker [/bib_ref] [bib_ref] The balance between the MIR164A and CUC2 genes controls leaf margin serration..., Nikovics [/bib_ref] [bib_ref] Redundancy and specialization among plant microRNAs: role of the MIR164 family in..., Sieber [/bib_ref] [bib_ref] Interplay of miR164, CUP-SHAPED COTYLEDON genes and LATERAL SUPPRESSOR controls axillary meristem..., Raman [/bib_ref]. To test whether OsCUC1 is regulated by microRNA, we searched for the corresponding microRNA using OsCUC1 as a putative target in the Plant Non-coding RNA Database, (PNRD http://structuralbiol ogy.cau.edu.cn/PNRD/index.php) [bib_ref] PNRD: a plant non-coding RNA database, Yi [/bib_ref]. The results indicated that OsCUC1 may be regulated by all members of the rice microRNA osa-miR164 family (osa-miR164a to osa-miR164f). These findings implied that osa-miR164 may also be involved in organ boundary specification and leaf development by negatively regulating OsCUC1. The bioinformatics analysis by PNRD indicated that OsCUC1 is considered to be the target of osa-miR164c with a high expectation value; moreover, only four genes, including OsCUC1, are predicted to be the potential targets of osa-miR164c, while the other osa-miR164 targets more genes. Thus, osa-miR164c was selected for further study. The results from RT-PCR followed by Sanger sequencing indicated that the real existence of the osa-miR164c is supported by New Phytologist transcript evidence [fig_ref] Figure S9: Transcript evidence and expression pattern for rice osa-miR64c [/fig_ref]. Moreover, the qRT-PCR results indicated that osa-miR164c was accumulated highly in leaf tissues, but almost no expression could be detected in the mature flower organs, in which OsCUC1 is relatively highly expressed [fig_ref] Figure S9: Transcript evidence and expression pattern for rice osa-miR64c [/fig_ref]. To verify the functions of osa-miR164c in regulating boundary specification and leaf morphology, we overexpressed osa-miR164c in the genetic background of Nipponbare. In comparison to the control plants carrying the empty vector, OsCUC1 was significantly downregulated in both the vegetative and reproductive organs of OE-osa-miR164c plants [fig_ref] Figure 5: osa-miR164c is involved in boundary specification and leaf development by negatively regulating... [/fig_ref]. Similar to oscuc1-KO, the OE-osa-miR164c plants also showed defects in organ boundary separation and leaf development [fig_ref] Figure 5: osa-miR164c is involved in boundary specification and leaf development by negatively regulating... [/fig_ref]. However, unlike in oscuc1-KO, the stamens developed normally in OE-osa-miR164c, indicating that a low expression level of OsCUC1 is sufficient for maintaining normal organ separation in the reproductive growth stage but not in the vegetative growth stage. Notably, two transgenic lines with higher expression of osa-miR164c (OE-osa-miR164c-4 and OE-osa-miR164c-15) showed more severe defects than the others, and no seeds could be obtained from these two lines.
## Research
## New phytologist
OsCUC1 contains an osa-miR164-target sequence in the second exon. To further test the effects of osa-miR164 on OsCUC1, we generated transgenic lines carrying an OsCUC1 in-frame fusion with the b-glucuronidase (GUS) gene driven by its endogenous promoter with or without synonymous mutations in the osa-miR164-targeted site (pOsCUC1::OsCUC1-GUS and pOsCUC1::mOsCUC1-GUS) [fig_ref] Figure 5: osa-miR164c is involved in boundary specification and leaf development by negatively regulating... [/fig_ref]. The pOsCUC1:: mOsCUC1-GUS lines showed a relatively high level of GUS staining in most of the organs [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref] , while only minor staining was observed in the pOsCUC1::OsCUC1-GUS lines [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. These results strongly suggest that osa-miR164c functions by dampening the transcription level of OsCUC1. In agreement with this idea, the transcriptional reporter of OsCUC1 (pOsCUC1::GUS) showed a strong expression pattern, as expected [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref].
osa-miR164 targeting OMTN4 and OMTN6 may not be involved in meristem/organ boundary specification or leaf development
In addition to OsCUC1, five other NAC (OMTN) genes, OMTN1 (ONAC027), OMTN2 (ONAC004/OsNAC2), OMTN3 (ONAC060), OMTN4 (ONAC011) and OMTN6 (ONAC104), are considered to be putative targets of osa-miR164 [bib_ref] Conserved miR164-targeted NAC genes negatively regulate drought resistance in rice, Fang [/bib_ref]. We have shown that osa-miR164c plays a critical role in boundary specification and leaf development by downregulating OsCUC1. Our qRT-PCR results indicated that only OsCUC1and not the other five putative osa-miR164 targetswas significantly downregulated in all the organs tested in OE-osa-miR164c plants [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. In a previous study, OMTN4-RNAi (RNA interference) and OMTN6-RNAi transgenic plants were reported to show severe abnormal phenotypes, such as twisted-rolling leaves and fused organs, which were similar to those of the oscuc1-KO plants [bib_ref] Conserved miR164-targeted NAC genes negatively regulate drought resistance in rice, Fang [/bib_ref]. To verify whether OMTN4 and OMTN6 function in controlling organ specification and leaf development, we knocked out these two genes using CRISPR/Cas9 in the Nipponbare genetic background. We did not observe any fused or twisted-rolling leaves for the homozygous knockout (KO) mutants of these two genes during the vegetative growth stage, and in the reproductive stage, defects in the florets were barely detected. Moreover, no significant defects could be observed from two other mutant lines with the Zhonghua11 background (omtn4 ZH11 and omtn6 ZH11 ) either [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref] ; [fig_ref] Table S4: Percentages of aberrant florets of omtn4-KO and omtn6-KO plants [/fig_ref]. Therefore, the defects in the two RNAi lines may be caused by RNAi off-target effects. Thus, we postulated that the targets of osa-miR164, OMTN4 and OMTN6 may not respond to meristem/organ boundary specification or leaf development as OsCUC1 does.
[formula] (a) (b) (c) (d) (e) (f) (h) (i) (j) (j1) (j2) (j3) (j4) (j5) (j6) (j7) (j8) (j9) (j10) (j11) (j12) (j13) (j14) (j15) [/formula]
OsCUC1 physically interacts with CLD1 and maintains CLD1 stability in the nucleus to control leaf morphology
Knocking out OsCUC1 in rice gives rise to defects in leaf development, which resemble the phenotype of the curled leaf and dwarf 1 (cld1) mutants [bib_ref] CLD1/SRL1 modulates leaf rolling by affecting cell wall formation, epidermis integrity and..., Li [/bib_ref]. However, the expression level of CLD1 showed no significant change in the KO plants [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. Thus, we posited that OsCUC1 might interact with CLD1 and function in maintaining the stability of the CLD1 protein.
In previous research, it has been demonstrated that CLD1, also known as Semi-Rolled Leaf 1 (SRL1), encodes a glycophosphatidylinositol (GPI)-anchored membrane protein, which localizes predominantly at the plasma membrane and modulates leaf development [bib_ref] Semi-rolled leaf1 encodes a putative glycosylphosphatidylinositol-anchored protein and modulates rice leaf rolling..., Xiang [/bib_ref] [bib_ref] CLD1/SRL1 modulates leaf rolling by affecting cell wall formation, epidermis integrity and..., Li [/bib_ref]. To verify whether CLD1 can interact with OsCUC1, we first tested whether CLD1 is located in the nucleus, as OsCUC1 is. We transiently co-expressed CLD1-GFP in-frame fusion proteins with a nuclear localization marker or membrane localization marker in rice protoplasts. Green fluorescence protein fluorescence indicated that CLD1 was located in both the membrane and the nucleus . Moreover, CLD1-GFP was detected in the nuclei-enriched fraction by immunoblotting . This result further confirmed the nuclear localization of CLD1. We then studied the protein-protein interactions between CLD1 and the NAC proteins; BiFC, yeast-two-hybrid (Y2H), in vivo Co-IP and in vitro pull-down assays revealed the interactions between CLD1 and OsCUC1 . By contrast, we did not detect any interactions between CLD1 and OsCUC3 or two other NAC proteins (the putative osa-miR164 targets OMTN4 and OMTN6) by either BiFC or Y2H assays [fig_ref] Figure 1: Appearance of rice oscuc1 knockout mutants [/fig_ref]. In the in vitro assay conditions, recombinant His-CLD1 was stable in the crude protein extract from non-KO1 OsCUC1 . By contrast, His-CLD1 proteins were rapidly degraded after the addition of crude protein extract from oscuc1-KO1, while the degradation of His-CLD1 was greatly slowed down when recombinant GST-OsCUC1 was first mixed with His-CLD1 [fig_ref] Figure 7: OsCUC1 stabilizes CLD1 in the nucleus in rice [/fig_ref]. Moreover, we carried out an assay for CLD1 stability in planta. Immunoblotting assays revealed that CLD1 was not detected in the nucleus-enriched fraction from oscuc1-KO plants, indicating that the nuclear accumulation of CLD1 is dependent on OsCUC1 functioning [fig_ref] Figure 7: OsCUC1 stabilizes CLD1 in the nucleus in rice [/fig_ref]. These results further confirmed the role of OsCUC1 in the stabilization of CLD1. Overall, the protein-protein interaction between OsCUC1 and CLD1 may stabilize CLD1 in the nucleus; loss of function of OsCUC1 may lead to acceleration of CLD1 degradation in the nucleus, resulting in a phenotype with twisted-rolling leaves, which is similar to that of the cld1 mutant.
# Discussion
In [bib_ref] The no apical meristem gene of Petunia is required for pattern formation..., Souer [/bib_ref] [bib_ref] Genes involved in organ separation in Arabidopsis: an analysis of the cup-shaped..., Aida [/bib_ref] [bib_ref] The CUP-SHAPED COTYLEDON1 gene of Arabidopsis regulates shoot apical meristem formation, Takada [/bib_ref] [bib_ref] The CUP-SHAPED COTYLEDON3 gene is required for boundary and shoot meristem formation..., Vroemen [/bib_ref] [bib_ref] The NAC-domain transcription factor GOBLET specifies leaflet boundaries in compound tomato leaves, Berger [/bib_ref] [bib_ref] NO APICAL MERISTEM (MtNAM) regulates floral organ identity and lateral organ separation..., Cheng [/bib_ref]. On the other hand, miR164 is involved in multiple developmental processes by negatively regulating its targets, which are always NAC transcription factors. In Arabidopsis, miR164 regulates boundary specification, leaf margin serration, lateral root development, pathogeninduced cell death, age-dependent cell death and the salt stress response by targeting different NAC genes [bib_ref] MicroRNA directs mRNA cleavage of the transcription factor NAC1 to downregulate auxin..., Guo [/bib_ref] [bib_ref] AtNAC2, a transcription factor downstream of ethylene and auxin signaling pathways, is..., He [/bib_ref] [bib_ref] The balance between the MIR164A and CUC2 genes controls leaf margin serration..., Nikovics [/bib_ref] [bib_ref] Redundancy and specialization among plant microRNAs: role of the MIR164 family in..., Sieber [/bib_ref] [bib_ref] Interplay of miR164, CUP-SHAPED COTYLEDON genes and LATERAL SUPPRESSOR controls axillary meristem..., Raman [/bib_ref] [bib_ref] Trifurcate feed-forward regulation of age-dependent cell death involving miR164 in Arabidopsis, Kim [/bib_ref] [bib_ref] An Arabidopsis NAC transcription factor NAC4 promotes pathogen-induced cell death under negative..., Lee [/bib_ref] ; in kiwifruit (Actinidia spp.), an interplay between ade-miR164 and AdNAC6/7 regulates fruit ripening ; in strawberry (Fragaria vesca), the miR164-CUC2 module regulates leaf and flower development [bib_ref] Conserved and novel roles of miR164-CUC2 regulatory module in specifying leaf and..., Zheng [/bib_ref] ; in M. truncatula, miR164-MtNAC1 pathway is involved in root and symbiotic nodule development ; and in wheat (Triticum aestivum), TaNAC21/22, the targets of tae-miR164, play an important role in regulating the resistance of host plants to stripe rust [bib_ref] The target gene of tae-miR164, a novel NAC transcription factor from the..., Feng [/bib_ref]. In rice, overexpression of osa-miR164b led to dwarf plant architecture and small panicles [bib_ref] Overexpression of miR164b-resistant OsNAC2 improves plant architecture and grain yield in rice, Jiang [/bib_ref]. Transgenic plants overexpressing the osa-miR164-targeted genes, including OMTN2 (ONAC004/ OsNAC2), OMTN3 (ONAC060), OMTN4 (ONAC011) and OMTN6 (ONAC104) increased the sensitivity to drought stress at the reproductive growth stage, indicating that these genes appear to be associated with the response to abiotic stresses [bib_ref] Conserved miR164-targeted NAC genes negatively regulate drought resistance in rice, Fang [/bib_ref].
## Research
## New phytologist
Arabidopsis are involved in SAM formation. Mutations in NAM, MtNAM and CUP lead to fusion in cotyledons, and no apical meristem can be formed. In Medicago, the development of the strong mutant of MtNAM is arrested in the fused cotyledon stage [bib_ref] NO APICAL MERISTEM (MtNAM) regulates floral organ identity and lateral organ separation..., Cheng [/bib_ref]. By contrast, in Petunia, the nam mutant occasionally produces escape shoots, which develop normally in the vegetative stage, but show defects in flower development [bib_ref] The no apical meristem gene of Petunia is required for pattern formation..., Souer [/bib_ref]. In Antirrhinum, the cup mutant always produces escape shoots; however, these escape shoots display defects during all developmental stages. In Arabidopsis, CUC1 and CUC2 function redundantly in SAM formation, and seedlings of the cuc1 cuc2 double mutant, which completely lacks a SAM, exhibit a cup-shaped cotyledon [bib_ref] Genes involved in organ separation in Arabidopsis: an analysis of the cup-shaped..., Aida [/bib_ref]. Further study indicated that cuc2
[formula] (a) (b) (c) (d) (e) (f) Fig. 6 [/formula]
OsCUC1 interacts with CLD1 in rice. (a) CLD1 is localized to the cell membrane as well as the nucleus. Upper row: co-localization of CLD1-GFP and an mCherry-tag nuclear localization marker. Upper middle row: co-localization of CLD1-GFP and an mCherry-tag membrane localization marker. Lower middle row: co-localization of green fluorescent protein (GFP) and an mCherry-tag membrane localization marker as a control. Lower row: co-localization of GFP and an mCherry-tag nuclear localization marker as a control. MLM, membrane localization marker (OsRac3-mCherry); NLM, nuclear localization marker (Ghd7-mCherry). Bars, 5 lm. (b) Accumulation of CLD1-GFP in the nucleus. Rice (Nipponbare) protoplast transiently expressed GFP or CLD-GFP. CLD1-GFP proteins were detected using anti-GFP antibodies in total proteins of the rice protoplast or in the nuclei-enriched fraction only. Histone H3 served as a nuclear marker and UGPase (UDP-glucose pyrophosphorylase) as a cytoplasmic marker. (c) BiFC assay. Co-expression of GFP C plus GFP N -CLD1, GFP C -CLD1 plus GFP N , GFP C -OsCUC1 plus GFP N , and GFP C plus GFP N -OsCUC1 were used as negative controls. Bars, 5 lm. (d) Y2H assay. The combination of BK-53 plus AD-T was used as a positive control, while BK-CLD1 plus AD, BK-LAM plus AD-T, and BK plus AD were used as negative controls. (e) In vivo co-immunoprecipitation (CoIP) assay. After the co-transformation of OsCUC1-GFP and CLD1-Flag in rice protoplasts, total proteins of protoplasts were immunoprecipitated using an anti-GFP antibody and were detected with anti-GFP and anti-Flag antibodies. IB, immunoblot; IP, immunoprecipitation. (f) In vitro pull-down assay. Recombinant GST-OsCUC1 and His-CLD1 proteins were used for the pull-down assay. IB, immunoblot.
Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation
New Phytologist (2021) 229: 1566-1581 www.newphytologist.com cuc3 double mutant seedlings, but not cuc1 cuc3 seedlings, have no functional SAM. Thus, only cuc2 in combination with cuc1 and/or cuc3 leads to the absence of the SAM [bib_ref] The CUP-SHAPED COTYLEDON3 gene is required for boundary and shoot meristem formation..., Vroemen [/bib_ref]. In our study, we found that the oscuc1 oscuc3 double mutant produces aberrant organs at the seedling stage, and the development of the double mutant seedling is arrested. These results indicated that, in addition to OsCUC1 and OsCUC3, other factor(s) may be involved in the regulation of SAM initiation, and loss of function of these two genes alone is not sufficient to prevent SAM formation. However, OsCUC1 and OsCUC3 are critical for maintaining normal SAM activity. Plants without functional OsCUC1 and OsCUC3 products stop developing and ultimately die at the seedling stage.
## Functional divergence between oscuc1 and oscuc3
Although OsCUC1 and OsCUC3 function together in boundary specification and SAM maintenance, their functions remain different in other developmental processes. Loss of function of OsCUC1 leads to a dwarf plant structure, pollen defection and twisted-rolling leaves, while oscuc3 mutants show no defects in these related developmental processes. Our study indicates that only OsCUC1but not OsCUC3 or the other two NAC proteins (OMTN4 and OMTN6)interacts with the leaf-rolling related protein CLD1. These results are consistent with the fact that oscuc1 mutants but not oscuc3 or omtn4/6 mutants show defects in leaf development. Thus, the different interaction behaviors of OsCUC1 and OsCUC3 with the other gene may be partly responsible for the functional divergence of these two proteins.
Despite divergence, the mechanisms of meristem/organ boundary specification between Arabidopsis and rice are conserved In Arabidopsis, due to the redundant function of CUC1 and CUC2, neither the cuc1 nor the cuc2 single mutation results in a severe phenotype, but the cuc1 cuc2 double mutation causes a complete lack of embryonic shoot meristem formation. The regenerated cuc1 cuc2 mutant exhibits abnormal flowers, in which all the sepals and most of the stamens are severely fused [bib_ref] Genes involved in organ separation in Arabidopsis: an analysis of the cup-shaped..., Aida [/bib_ref] [bib_ref] Shoot apical meristem and cotyledon formation during Arabidopsis embryogenesis: interaction among the..., Aida [/bib_ref] [bib_ref] The CUP-SHAPED COTYLEDON1 gene of Arabidopsis regulates shoot apical meristem formation, Takada [/bib_ref]. In rice, the oscuc1 single mutation causes severe defects in meristem/organ boundary specification, resulting in a reduction in stamen numbers and fusion of leaves/filaments. Consistent with this, systematic sequence analysis revealed that OsCUC1 is the only homolog of both CUC1 and CUC2 [bib_ref] Systematic approaches to using the FOX hunting system to identify useful rice..., Kondou [/bib_ref] , suggesting that the function of OsCUC1 may be executed redundantly by CUC1 and CUC2 in Arabidopsis. Despite the divergence between OsCUC1 and CUC1/2, the mechanisms of meristem/organ boundary specification are conserved between the dicot Arabidopsis and the monocot O. sativa : one or more microRNA (osa-miR164/miR164) target(s) (OsCUC1/CUC1&CUC2) function redundantly with another NAC gene family member (OsCUC3/ CUC3) in meristem/organ boundary specification. In addition, OsCUC1 is involved in leaf development by affecting the stability of CLD1 in the nucleus. In vitro protein degradation assay. Recombinant His-CLD1 protein was incubated with total crude extract from non-KO1 OsCUC1 (left) and oscuc1-KO1 (middle, without GST-OsCUC1; right, with GST-OsCUC1). HSP (Hsp90) was used as a loading control. (b) The accumulation of CLD1 in the nucleus depends on OsCUC1 functioning; accumulation of CLD1 in non-KO1 OsCUC1 and oscuc1-KO1 plants is shown. CLD1 was detected with an anti-CLD1 antibody. Histone H3 served as a nuclear marker, and UGPase (UDP-glucose pyrophosphorylase) served as a cytoplasmic marker. HSP (Hsp90) was used as a loading control. Three independent plants of each line were used for the assay.
New Phytologist (2021) 229: 1566-1581 Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation www.newphytologist.com Research New Phytologist Interplay of a microRNA and CUP-SHAPED COTYLEDON genes controls meristem/organ boundary specification in rice and Arabidopsis. The microRNA (osa-miR164/miR164) negatively regulates the CUP-SHAPED COTYLEDON genes (OsCUC1/CUC1&CUC2), which function redundantly with another NAC gene family member (OsCUC3/CUC3) in meristem/organ boundary specification in both rice and Arabidopsis. In addition, OsCUC1 controls plant height, male fertility and leaf morphology in rice. Red represents Arabidopsis, and green represents rice.
Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation New Phytologist (2021) 229: 1566-1581 www.newphytologist.com
## Supporting information
Additional Supporting Information may be found online in the Supporting Information section at the end of the article.
## Table s1
The accession numbers of the proteins listed in the phylogenetic tree.
## Table s2
The primer sequences used in this study. Please note: Wiley Blackwell are not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.
New Phytologist is an electronic (online-only) journal owned by the New Phytologist Foundation, a not-for-profit organization dedicated to the promotion of plant science, facilitating projects from symposia to free access for our Tansley reviews and Tansley insights. Regular papers, Letters, Research reviews, Rapid reports and both Modelling/Theory and Methods papers are encouraged. We are committed to rapid processing, from online submission through to publication 'as ready' via Early View -our average time to decision is <26 days. There are no page or colour charges and a PDF version will be provided for each article.
The journal is available online at Wiley Online Library. Visit www.newphytologist.com to search the articles and register for
[fig] Figure 1: Appearance of rice oscuc1 knockout mutants. (a) Schematic diagram of OsCUC1 gene structure and the CRISPR/Cas9 target sites. The black triangles indicate the two individual target sites for knockout of OsCUC1. The PAM site is underlined. The mutation site is shown in red. (b, c) The whole plant phenotypes of oscuc1-KO1 and oscuc1-KO2. (d) The height of mature non-KO1 OsCUC1 and oscuc1-KO1 plants. Values are shown as means AE SD (20 non-KO1 OsCUC1 plants and 25 oscuc1-KO1 were used for the statistical analysis). **, P < 0.01 (Student's t-test). (e, f) The twisted-rolling leaves in oscuc1-KO1 and oscuc1-KO2. (g, h) Fusion of leaf blade in oscuc1-KO1 and oscuc1-KO2. The fusion regions outlined by small white rectangles are shown in detail on the right-hand side of their respective panels. The arrows indicate the fusion of the leaf blades. (i-l) Cross-section of the leaf sheath of oscuc1-KO1, oscuc1-KO2 and their non-KO control. (m-p) Knocking out OsCUC1 leads to fusion in filaments and reduction in stamens. The arrows indicate the fusion of the filaments. (q-v) Scanning electron micrographs of early-arising non-KO1 OsCUC1 and oscuc1-KO1 floret. FM, flora meristem; le, lemma; pa, palea; sl, sterile lemma; rg, rudimentary glumes. The white asterisks represent the primordia of the stamens. Bars: (g-p) 500 lm; (q-v) 100 lm. New Phytologist (2021) 229: 1566-1581 Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation www.newphytologist.com [/fig]
[fig] Figure 2: Appearance of rice oscuc3 knockout mutants. (a) Schematic diagram of OsCUC3 gene structure and the CRISPR/Cas9 target sites. The black triangles indicate the two individual target sites for knockout of OsCUC3. The PAM site is underlined. The mutation site is shown in red. (b, c) The whole plant phenotype of oscuc3-KO1 and oscuc3-KO2. (d-g) Cross-section of the leaf sheath of oscuc3-KO1, oscuc3-KO2 and their non-KO control. (h, i) Fusion of leaf blade in oscuc3-KO1 and oscuc3-KO2. The fusion region outlined by the small white rectangle is shown in detail in the bottom right corner. The arrows indicate the fusion of the leaf blades. (j-m) Knocking out OsCUC3 leads to fusion in filaments. The arrow indicates the fusion of two filaments of oscuc3-KO2. Bars, 500 lm. Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation New Phytologist (2021) 229: 1566-1581 www.newphytologist.com [/fig]
[fig] Figure 3: The expression patterns of OsCUC1 and OsCUC3 in rice. (a) Subcellular localization of OsCUC1 and OsCUC3. NLM, nuclear localization marker (Ghd7-mCherry). Bars, 5 lm. (b) Transcriptional activity analysis of OsCUC1 and OsCUC3. The empty vector pGBKT7 was used as a negative control. (c, d) Relative expression of OsCUC1 and OsCUC3 in seedling (3 d after germination), leaf, root, UBI (unelongated basal internode), leaf sheath, bract, flag leaf, YP 0.5 cm (young panicle 0.5 cm), YP 2 cm, YP 8 cm, stamen, pistil, palea and lemma. Values are shown as means AE SD (n = 3). (e) In situ hybridization of OsCUC1 and OsCUC3. (e1-e4) Accumulation of OsCUC1 in wild-type plants. (e6-e9) Accumulation of OsCUC3 in wild-type plants. (e5, e10) Negative controls with sense probes. The black arrows indicate the accumulation regions of the mRNA. FM, flora meristem; le, lemma; pa, palea; rg, rudimentary glumes; sl, sterile lemma; st, stamens. Bars, 50 lm. Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation New Phytologist (2021) 229: 1566-1581 www.newphytologist.com [/fig]
[fig] Figure 4: OsCUC1 and OsCUC3 dimerize and function redundantly in rice. (a, b) BiFC assays showing the heterodimerization and homodimerization of OsCUC1 and OsCUC3. Co-expression of GFP C -OsCUC3 plus GFP N , GFP C plus GFP N -OsCUC3, GFP C -OsCUC1 plus GFP N , and GFP C plus GFP N -OsCUC1 were used as negative controls. Bars, 5 lm. (c) Interaction between OsCUC1 and OsCUC3, analysed by in vitro pull-down assay. Recombinant GST-OsCUC1 and MBP-OsCUC3 proteins were used for the pull-down assay. IB, immunoblot. (d) Interaction between OsCUC1 and OsCUC3, analysed by in vivo CoIP assay. After the co-transformation of OsCUC1-MYC and OsCUC3-GFP in rice protoplasts, total proteins of protoplasts were immunoprecipitated using an anti-MYC antibody and were detected with anti-GFP and anti-MYC antibodies. IB, Immunoblot; IP, immunoprecipitation. (e) The phenotypes of +/+ oscuc3/oscuc3, oscuc1/+ oscuc3/oscuc3 and oscuc1/oscuc1 oscuc3/oscuc3. (f) The fusion of leaf sheath and twisted-rolling leaf in oscuc1/+ oscuc3/oscuc3. The fusion region of the leaf sheath outlined by a small white rectangle in the main image is shown in detail in the top right corner; the arrow indicates the twisted-rolling leaf. Bar, 500 lm. (g) The enhanced defects in the oscuc1/oscuc1 oscuc3/oscuc3 double mutant. Bar, 500 lm.New Phytologist (2021) 229: 1566-1581 Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation www.newphytologist.com [/fig]
[fig] Figure 5: osa-miR164c is involved in boundary specification and leaf development by negatively regulating OsCUC1 in rice. (a) Relative expression of osa-miR164c in different individual OE-osa-miR164c lines and empty-vector control lines (EVC). Values are shown as means AE SD (n = 3). (b) Relative expression of OsCUC1 in EVC and two OE-osa-miR164c lines in different tissues. Values are shown as means AE SD (n = 3). **, P < 0.01 (Student's t-test). (c) The phenotype of OE-osa-miR164c and EVC lines. (d) The twisted-rolling leaf in OE-osa-miR164c. (e) Overexpressing osa-miR164c leads to fusion in leaf sheaths. (f, g) Detailed images of the fusion regions depicted in (e); the arrows indicate the fusion regions on the leaf sheath. (h) Overexpressing osa-miR164c leads to fusion in the leaf blade. (i) Schematic diagram depicting the construction of mOsCUC1.The red letters indicate mutant nucleotides introduced into the osa-miR164 target site in the mOsCUC1-GUS transgene, which interrupts the osa-miR164 target site without changing the amino acid residues. Watson-Crick base pairing between the mRNA and osa-miR164c is indicated by black lines. Mismatches and G:U wobbles are indicated by a star or equals sign, respectively. (j) Gus staining for pOsCUC1::mOsCUC1-GUS, pOsCUC1::OsCUC1-GUS and pOsCUC1::GUS in seedling (j1, j6, j11), leaf blade (j2, j7, j12), leaf sheath (j3, j8, j13), root (j4, j9, j14), and floret (j5, j10, j15). Bars, 200 lm.Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation New Phytologist (2021) 229: 1566-1581 www.newphytologist.com [/fig]
[fig] Figure 7: OsCUC1 stabilizes CLD1 in the nucleus in rice. (a) [/fig]
[fig] Figure S1: Original blot images from this study. [/fig]
[fig] Figure S2: Sequence analysis for OsCUC1 and OsCUC3. [/fig]
[fig] Figure S3: Loss-of-function of OsCUC1 and OsCUC3 generated by CRISPR/Cas9 in rice. New Phytologist (2021) 229: 1566-1581 Ó 2020 The Authors New Phytologist Ó 2020 New Phytologist Foundation www.newphytologist.com Research New Phytologist [/fig]
[fig] Figure S4: Boundary specification defects in the vegetative growing stage of the rice oscuc1-KO1 mutant. [/fig]
[fig] Figure S5: Pollen defects in the rice oscuc1-KO1 mutant. [/fig]
[fig] Figure S6: Phenotypes of heterozygous mutants of OsCUC1 and OsCUC3 in rice. [/fig]
[fig] Figure S7: Dimerization of the Arabidopsis CUC proteins. [/fig]
[fig] Figure S8: The development of rice oscuc1 oscuc3 homozygous double mutants is arrested at the seedling stage. [/fig]
[fig] Figure S9: Transcript evidence and expression pattern for rice osa-miR64c. [/fig]
[table] Table S3: Percentages of aberrant florets of oscuc1-KO and oscuc3-KO plants. [/table]
[table] Table S4: Percentages of aberrant florets of omtn4-KO and omtn6-KO plants. [/table]
|
Human placenta‐derived mesenchymal stem cells induce trophoblast invasion via dynamic effects on mitochondrial function
The trophoblast is a critical cell for placental development and embryo implantation in the placenta. We previously reported that placenta-derived mesenchymal stem cells (PD-MSCs) increase trophoblast invasion through several signaling pathways. However, the paracrine effects of PD-MSCs on mitochondrial function in trophoblasts are still unclear.Therefore, the objective of the study was to analyze the mitochondrial function of trophoblasts in response to cocultivation with PD-MSCs. The results showed that PD-MSCs regulate the balance between cell survival and death and protect damaged mitochondria in trophoblasts from oxidative stress. Moreover, PD-MSCs upregulate factors involved in mitochondrial autophagy in trophoblast cells. Finally, PD-MSCs improve trophoblast invasion. Taken together, the data indicate that PD-MSCs can regulate trophoblast invasion through dynamic effects on mitochondrial energy metabolism. These results support the fundamental role of mitochondrial energy mechanism in trophoblast invasion and suggest a new therapeutic strategy for infertility. K E Y W O R D S invasion, mitochondria, mitochondrial autophagy, placenta-derived mesenchymal stem cells, trophoblast
placentation and gynecological diseases, such as preeclampsia, through vascular dysfunction [bib_ref] Granulocyte colony-stimulating factor (G-CSF) upregulates beta1 integrin and increases migration of human..., Espino [/bib_ref] [bib_ref] Lefty promotes the proliferation and invasion of trophoblast cells by inhibiting nodal..., Li [/bib_ref].
Mitochondria have multiple functions, including cellular metabolism, adenosine triphosphate (ATP) synthesis, reactive oxygen species generation and calcium ion movement and signaling, that are important for cellular homeostasis. First, ROS generation is induced under hypoxic conditions and during cellular metabolism (e.g., cell survival signaling and ATP synthesis), leading to oxidative stress [bib_ref] Mitochondrial ROS regulation of proliferating cells, Diebold [/bib_ref] [bib_ref] Regulation of ATP production by mitochondrial Ca(2+), Tarasov [/bib_ref]. High ROS levels lead to critical oxidative stress and result in preeclampsia, intrauterine growth restriction (IUGR), and cell death. However, antioxidants can manage moderate increases in ROS levels and selectively remove abnormal mitochondria [bib_ref] Selenium supplementation protects trophoblast cells from mitochondrial oxidative stress, Khera [/bib_ref] [bib_ref] Mitochondrial fission, fusion, and stress, Youle [/bib_ref]. Previous reports have suggested that heme oxygenase (HO) 1 and 2 and antioxidants promote the proliferation and invasion of various trophoblast cell lines, including human umbilical vein endothelial cells and human embryonic stem cells [bib_ref] The effects of heme oxygenase by-products on the proliferation and invasion of..., Ha [/bib_ref] [bib_ref] Effects of selenium on the survival and invasion of trophoblasts, Na [/bib_ref]. Second, cellular ATP production is regulated via calcium ion movement between mitochondria and the endoplasmic reticulum (ER).
Ca 2+ uptake into the mitochondrial matrix stimulates ATP production through the activation of dehydrogenases in the Krebs cycle [bib_ref] The IP(3) receptor-mitochondria connection in apoptosis and autophagy, Decuypere [/bib_ref]. Ca 2+ enters through inositol triphosphate receptor (IP3R) at ER-mitochondria junctions. IP3R, a membrane glycoprotein complex, is a well-known activator of Ca 2+ channels and leads to proliferative arrest through effects on Ca 2+ entry. Interestingly, IP3R is activated by PI3K and FGF signaling and upregulates cytosolic Ca 2+ . Collectively, Ca 2+ uptake into the mitochondria is necessary not only for determining cell fate (survival or death) but also for regulating mitochondrial function [bib_ref] Calcium signaling in placenta, Baczyk [/bib_ref] [bib_ref] Molecular interaction and functional coupling between type 3 inositol 1,4,5-trisphosphate receptor and..., Maycotte [/bib_ref]. Recent reports suggest a correlation between HIF1-α and mitochondrial calcium uniport (MCU). In MCUdeficient cells, mitochondrial ROS levels and HIF1-α gene expression are decreased. These data suggest that the signaling roles of mitochondrial ROS and HIF1-α are downstream of the mitochondrial calcium channel [bib_ref] Overexpression of hypoxia-inducible factor and metabolic pathways: Possible targets of cancer, Singh [/bib_ref] [bib_ref] A molecular-beacon-based asymmetric PCR assay for easy visualization of amplicons in the..., Sonkar [/bib_ref]. In a similar study, [bib_ref] Mitochondrial dynamics regulates migration and invasion of breast cancer cells, Zhao [/bib_ref] showed that mitochondrial dynamics, such as fission in breast cancer cells, regulates migration and invasion. Moreover, mitochondrial fragmentation by fission enhances cellular glycolysis and lactate production via upregulating DRP expression [bib_ref] Molecular interaction and functional coupling between type 3 inositol 1,4,5-trisphosphate receptor and..., Maycotte [/bib_ref] [bib_ref] Mitochondrial dynamics regulates migration and invasion of breast cancer cells, Zhao [/bib_ref].
Mitochondrial quality control systems regulate mitochondrial autophagy (mitophagy) and mitochondrial biogenesis and are thus involved in cellular homeostasis. Mitophagy is a major survival mechanism that balances mitochondrial fission and fusion. Persistent hypoxia induces selective mitophagy through the active destruction of damaged mitochondria [bib_ref] PINK1-phosphorylated mitofusin 2 is a Parkin receptor for culling damaged mitochondria, Chen [/bib_ref]. Additionally, an imbalance between mitochondrial fission and fusion can induce trophoblast-related diseases. PTEN-induced putative kinase 1 (PINK1) protects cells from stress-induced mitochondrial damage and activates PARKIN, a cytosolic E3 ubiquitin ligase, to bind to depolarized mitochondria. PARKIN plays a role in recognizing proteins on the mitochondrial outer membrane and mediates the removal of damaged mitochondria via autophagy [bib_ref] PINK1/Parkin-mediated mitophagy in mammalian cells, Eiyama [/bib_ref]. The PINK1-PARKIN pathway is associated with the selective autophagy of damaged mitochondria. [bib_ref] PINK1-phosphorylated mitofusin 2 is a Parkin receptor for culling damaged mitochondria, Chen [/bib_ref] recently reported that PINK1 phosphorylates MFN2 and recruits PARKIN to dysfunctional mitochondria [bib_ref] The roles of PINK1, parkin, and mitochondrial fidelity in Parkinson's disease, Pickrell [/bib_ref].
Human placenta-derived mesenchymal stem cells have been in the spotlight recently for their high self-renewal capacity, immunomodulatory properties and therapeutic effects. [bib_ref] Effect of mesenchymal stem cells and extracts derived from the placenta on..., Choi [/bib_ref] reported that mesenchymal stem cells (MSCs) and extracts isolated from normal term placenta promote trophoblast invasion and immunomodulation by altering HLA-G expression [bib_ref] Effect of mesenchymal stem cells and extracts derived from the placenta on..., Choi [/bib_ref] [bib_ref] Immunomodulatory effects of placenta-derived mesenchymal stem cells on T cells by regulation..., Kim [/bib_ref]. Several reports, have suggested that MSCs improve trophoblast invasion through various mechanisms that influence cellular function, including integrin signaling [bib_ref] Placental villous mesenchymal cells trigger trophoblast invasion, Chen [/bib_ref] [bib_ref] Effects of human umbilical cord mesenchymal stem cells on human trophoblast cell..., Huang [/bib_ref]. However, there is limited evidence for the detailed mechanism by which energy metabolism and mitochondrial function affect trophoblasts. Therefore, we investigated the effect of PD-MSCs on the proliferation, death and invasion of trophoblasts.
We report the basic mechanism by which PD-MSCs regulate trophoblast invasion through effects on mitochondrial function.
## | materials and methods
## | cell culture and indirect coculture systems
Placentas were obtained from women without any medical, obstetrical, or surgical complications who delivered at term (38 ± 2 gestational weeks). It is approved by the Institutional Review Board of CHA General Hospital, Seoul, Korea (IRB 07-18). Briefly, PD-MSCs were isolated from as the inner side of the chorionic plate of the placentas obtained following Cesarean section, as described in previously reports . PD-MSCs were cultured in alpha-minimum essential medium (HyClone, GE Healthcare Life Sciences) supplemented with 10% fetal bovine serum (FBS; Gibco-BRL), 1% penicillin/streptomycin (Pen-Strep; Gibco-BRL), 25 μg/ml human fibroblast growth factor 4 (PeproTech Inc.) and 1 μg/ml heparin (Sigma-Aldrich) at 37°C in an incubator with a humidified atmosphere of 5% CO 2 . The human extravillous trophoblast cell line HTR-8/SVneo was provided by Dr. Graham (Queen's University, Kingston, Ontario, Canada). HTR-8/SVneo cells were cultured in Roswell Park Memorial Institute 1640 medium (HyClone) supplemented with 5% FBS (Gibco-BRL) and 1% Pen-Strep (Gibco-BRL) under the same conditions as PD-MSCs.
In this study, indirect coculture was performed with a six-well cell culture trans well system (8 μm pore insert; Falcon, BD Biosciences). HTR-8/SVneo cells (1 × 10 5 ) were seeded in the lower chamber with 2 ml medium, PD-MSCs (2.5 × 10 5 ) were seeded in the upper chamber with 0.7 ml serum-free medium, and the system was incubated for 12, 24, and 48 h.
## | trypan blue staining
To analyze the survival rate of HTR-8/SVneo cells according to PD-MSC cocultivation, HTR-8/SVneo cells (1 × 10 5 ) were seeded in the lower chamber with 2 ml medium, and PD-MSCs (2.5 × 10 5 ) were SEOK ET AL. | 6679 seeded in the upper chamber with 0.7 ml serum-free medium. The cells (including adherent and floating cells) were harvested at 12, 24, and 48 h and stained with trypan blue (Gibco-BRL); the positively stained cells were counted by using a hemocytometer.
2.3 | Total RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was isolated from harvested HTR-8/SVneo cells using TRIzol reagent (Ambion) according to the manufacturer's protocol. The isolated RNA concentration was quantified by a Nanodrop spectrophotometer (Thermo Fisher Scientific). Reverse transcription was performed with 1 μg total RNA, 20 pmol oligo dT, 10 mM dNTP mix, RNase OUT, and SuperScript III Reverse Transcriptase (Invitrogen). The messenger RNA (mRNA) levels of specific genes were analyzed by qRT-PCR using SYBR Green Master Mix (Rox; Roche Diagnostics). The mRNA amplification conditions were as follows: denaturation at 95°C for 5 min, followed by 40 cycles of 5 s at 95°C and 30 s at 60°C. The human glyceraldehyde 3-phosphate dehydrogenase gene was used as an internal control for normalization. All reactions were performed in duplicate or triplicate. The relative mRNA expression levels were analyzed by the comparative CT method. The human-specific primers used are presented in .
## | the western blot analysis
Protein lysates were isolated from HTR-8/SVneo cells cocultured or not with PD-MSCs by using protein lysis buffer containing a protease inhibitor cocktail (mini-tablet; Roche) and phosphatase inhibitor cocktail II (A.G. Scientific). The protein lysates were then loaded onto 6%-15% sodium dodecyl sulphate (SDS) polyacrylamide gels for the detection of specific gene products. Separated proteins were transferred onto polyvinylidene difluoride membranes, which were then incubated for 1 h at room temperature with 5% bovine serum albumin (Amresco).
After this blocking step, the membranes were incubated overnight at 4°C with antibodies against p-ERK (1:1000; Cell Signaling), BAX (1:500; Santa Cruz), Bcl-2 (1:500; Santa Cruz), HO-1 (1:1000; Novus), HO-2 (1:1000; Novus), superoxide dismutase (SOD; 1:1000; Novus), heat shock protein 60 (HSP60; 1:1000; Cell Signaling), prohibitin 1 (PHB1; 1:1000; Cell Signaling), voltage-dependent anion channel (VDAC; 1:1000; Cell Signaling), PINK1 (1:1000; Abcam), and PARKIN (1:1000; Abcam). The membranes were then incubated with secondary antibody or anti-rabbit immunoglobulin G for 1 h at room temperature in an orbital shaker. After a washing step, the bands were detected using enhanced chemiluminescence reagents (Bio-Rad Laboratories Inc.). All experiments were performed in duplicate or triplicate.
## | fluorescence-activated cell sorting (facs) analysis
## | trans well assay
The invasiveness of trophoblasts according to PD-MSC cocultivation was analyzed by using inserts (8-μm pores; Falcon) in 24-well plates. MMP-2 and MMP-9 activity was analyzed based on the intensity of the unstained bands. All experiments were performed in duplicate.
## T a b l e 1 mrna primers used in this study
## | mito sox and mito tracker immunofluorescence
HTR-8/SVneo cells were seeded on cover slips and then cocultured with PD-MSCs using inserts for 12, 24, and 48 h. Next, the cells were washed with 1X cold PBS and incubated with MitoSOX and Mito-Tracker (Invitrogen) at 37°C. The cells were then washed with 1X cold PBS and incubated with 1 μg/ml diamidino-phenylindole hydrochloride (Sigma-Aldrich) at room temperature for 1 min. The cells were mounted with mounting solution (Dako), and images were obtained by using an EVOS microscope (Thermo Fisher Scientific).
## | immunofluorescence
To analyze PINK1 and PARKIN expression in trophoblast cells according to PD-MSC cocultivation, HTR-8/SVneo cells (1 × 10 4 ) were cultured with serum-free medium on cover slips in 24-well culture plates, and PD-MSCs (5 × 10 4 ) were cultured with cell culture med-
## | xf assay
To analyze the glycolytic flux and mitochondrial stress levels in HTR-8/SVneo cells according to PD-MSC cocultivation, the extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) were determined using an XF-24 cell culture assay according to the manufacturer's protocol. Briefly, HTR-8/SVneo cells (4 × 10 3 ) were seeded in the appropriate culture plates, and PD-MSCs (1.6 × 10 4 ) were seeded in 24-well inserts. After 12, 24, and 48 h, the cells were exposed to compounds that affect glycolysis
# | statistical analysis
Student's t tests were performed for groupwise comparisons, and a p value less than .05 was considered to indicate statistical significance. All experiments were performed in duplicate or triplicate.
## | pd-msc cocultivation increased hif1-α expression and ros levels in trophoblast cells
The HIF1-α gene is a key factor in dynamic cellular metabolism and invasion that is regulated by cellular ROS levels. HIF1-α gene expression was determined by qRT-PCR and enzyme-linked immunosorbent assay (ELISA) in trophoblast cells cocultured with PD-MSCs. HIF1-α mRNA expression was significantly increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone, as shown by qRT-PCR and ELISA (*p < .05, . These findings suggest that PD-MSC cocultivation induces the expression of HIF1-α, a transcription factor, in trophoblasts.
To
## | pd-msc cocultivation affected mitochondrial function in trophoblast cells
To investigate the effect of PD-MSC cocultivation on mitochondrial function in trophoblasts, we performed JC-1, NAO and ATP production assays. The JC-1 assay results indicated that the mitochondrial membrane potential was significantly decreased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone (*p < .05, . The NAO assay revealed an increased mitochondrial mass in trophoblasts cocultured with PD-MSCs at 12 and 24 h but a decreased mass in trophoblasts cocultured with PD-MSCs at 48 h compared to trophoblasts cultured alone (*p < .05, . These findings suggest that PD-MSC cocultivation influences mitochondrial function in trophoblasts, including mitochondrial membrane potential and mass.
We also analyzed the effect of PD-MSC cocultivation on HSP60, PHB1, and VDAC gene expression in trophoblasts . The expression of HSP60, which protects against cell damage, is induced under mitochondrial stress conditions; trophoblasts cocultured with PD-MSCs had markedly lower HSP60 expression than trophoblasts cultured alone at 12, 24, and 48 h (*p < .05, . Among the proteins involved in mitochondrial stabilization and ion channel signaling, PHB1 and VDAC were analyzed by western blotting. The expression of PHB1, which is related to mitochondrial stabilization, was extremely high in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone at 12, 24, and 48 h (*p < .05, . The expression of VDAC, an ion channel in the mitochondrial outer membrane related to ATP transport and synthesis, was markedly increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone (*p < .05, . IP3R
and MCU are involved in Ca 2+ signaling between the ER and mitochondria to regulate ATP synthesis [bib_ref] Ca2+ signals regulate mitochondrial metabolism by stimulating CREB-mediated expression of the mitochondrial..., Shanmughapriya [/bib_ref].
Therefore, we analyzed the expression of genes related to Ca 2+ signaling and ATP synthesis that may play a role in trophoblast . These findings suggest that PD-MSC cocultivation activates ATP metabolism through effects on calcium channels and enhances mitochondrial stabilization, resulting in reduced mitochondrial damage in trophoblast cells.
## | pd-msc cocultivation induces mitophagy factors in trophoblast cells
To evaluate mitochondrial autophagy, which provides quality control in trophoblast cells, we analyzed the effect of PD-MSC cocultivation on the expression of genes related to mitophagy in trophoblast cells by western blotting . PINK1 protein expression was significantly increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone at 12 and 24 h (*p < .05, . In addition, PARKIN expression was markedly increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone at 12, 24, and 48 h (*p < .05, . The results for PINK1 and PARKIN expression in trophoblasts are shown in . PD-MSC cocultivation markedly increased PINK1 and PARKIN expression in trophoblasts . These findings suggest that PD-MSC cocultivation regulates mitochondrial damage in trophoblasts by upregulating PINK1 and PARKIN gene expression.
## | pd-msc cocultivation improved trophoblast cell invasion
To ascertain the effect of PD-MSC cocultivation on trophoblast invasion, we performed an invasion assay using a trans well insert system. The invasion assays showed that trophoblast cell invasion was markedly increased in the PD-MSC cocultivation group compared to the individual culture group, as shown in . We counted the number of invaded cells through hematoxylin staining. The number of invaded cells significantly increased after PD-MSC cocultivation at 12, 24, and 48 h (*p < .05, . In accordance with previous reports, we also analyzed the activity of MMP-2/9, which are involved in trophoblast invasion. MMP-2/9 activity in the supernatant was analyzed by zymography. MSCs secrete many biologically active factors, including cytokines, chemokines, growth factors, exosomes and microRNAs, that have considerable therapeutic potential in several diseases [bib_ref] Effect of the microenvironment on mesenchymal stem cell paracrine signaling: Opportunities to..., Kusuma [/bib_ref] [bib_ref] Mechanisms of mesenchymal stem/stromal cell function, Spees [/bib_ref]. Yajing and colleagues recently reported that UC-MSCs regulate human chorionic gonadotropin, placental growth factor, and soluble Endoglin levels and promote the proliferation, migration, and invasion of trophoblasts through cell-cell interactions [bib_ref] Effects of human umbilical cord mesenchymal stem cells on human trophoblast cell..., Huang [/bib_ref]. Several studies reported that placenta samples from women with preeclampsia showed mitochondrial dysfunction, including higher ROS levels and smaller mitochondria, in trophoblast cells. Therefore, mitochondrial function has an important role in trophoblast invasion to establish a successful pregnancy [bib_ref] Trophoblast mitochondrial function is impaired in preeclampsia and correlates negatively with the..., Zsengeller [/bib_ref].
## Mmp
Several studies have demonstrated the involvement of various regulatory factors, including HIF1-α and MMPs, and mitochondrial function in trophoblast invasion. Especially, HIF1-α is multiple functional transcription gene related to angiogenesis, cell survival, invasion and energy metabolism. Under hypoxia conditions. HIF1-α expression generates ROS and stimulates cell migration, invasion, proliferation and dynamic metabolism through the ERK pathway [bib_ref] Deferoxamine promotes MDA-MB-231 cell migration and invasion through increased ROS-dependent HIF-1alpha accumulation, Liu [/bib_ref]. Moreover, [bib_ref] Effects of hypoxia inducible factors-1alpha on autophagy and invasion of trophoblasts, Choi [/bib_ref] showed that upregulating HIF1-α expression induces trophoblast invasion by activating MMP-2/9. In our study, PD-MSC cocultivation promoted trophoblast invasion by activating MMP-2/9 and inducing HIF1-α expression.
Collectively, these findings demonstrate that PD-MSCs regulate several factors related to migration to enhance the invasion ability of trophoblasts. Recently reports suggested that HIFs is transcription factor, but it is presented in outside the nucleus under hypoxic conditions [bib_ref] HIF-1alpha localization with mitochondria: A new role for an old favorite, Briston [/bib_ref]. Also, HIF1-a is a regulator on mitochondrial metabolism including mitochondrial autophagy as well as ROS generation [bib_ref] HIF-1-mediated expression of pyruvate dehydrogenase kinase: A metabolic switch required for cellular..., Kim [/bib_ref] [bib_ref] Hypoxia-inducible factor 1: Regulator of mitochondrial metabolism and mediator of ischemic preconditioning, Semenza [/bib_ref] [bib_ref] Mitochondrial autophagy is an HIF-1-dependent adaptive metabolic response to hypoxia, Zhang [/bib_ref]. In our data, PD-MSC cocultivation induced mitochondrial function on trophoblast including cellular ROS levels and mitochondrial membrane potential/mass. The correlation between mitochondrial metabolism and HIF1-α expression through glycolysis and the tricarboxylic acid cycle has been reported in several publications. Recent reports stated that the HIF1-α gene is in- Moreover, previous reports suggest that in contrast to syncytiotrophoblasts, cytotrophoblasts depend on both aerobic and glycolytic energy production [bib_ref] Energy metabolism and glycolysis in human placental trophoblast cells during differentiation, Bax [/bib_ref]. Our data showed that PD-MSC cocultivation improved mitochondrial function, including glycolysis and respiration, in trophoblast cells. Collectively, these findings suggest that PD-MSCs stimulate mitochondrial metabolism, including glycolysis and respiration, during trophoblast cell invasion.
Mitochondrial autophagy (mitophagy) is a selective event that targets damaged mitochondria in cells. The high energy requirement of invading cells induces mitochondrial damage in trophoblast cells.
PINK1 interacts with PARKIN to activate mitophagy and thus protect mitochondria. Mitochondrial autophagy is initiated after the accumulation of PINK1 and phosphorylated PARKIN on the mitochondrial membrane. Therefore, PINK1 and PARKIN expression promote the upregulation of autophagy biomarkers (e.g., p62 and LC3-I/II) during mitophagy. During mitochondrial autophagy, damaged mitochondria are selectively removed to protect cells.
Heeman B et al. recently reported that PINK1 modulates the mitochondrial network, energy maintenance and calcium homeostasis [bib_ref] Depletion of PINK1 affects mitochondrial metabolism, calcium homeostasis and energy maintenance, Heeman [/bib_ref]. Our data showed that PD-MSC cocultivation reduced mitochondrial damage by downregulating HSP70 expression and enhanced mitochondrial protection by upregulating PHB1 in trophoblast cells. Moreover, PD-MSC cocultivation increased the expression of mitophagy regulators, such as PINK1 and PARKIN, in trophoblast cells. Taken together, these findings suggest that PD-MSCs attenuate mitochondrial damage in trophoblast cells.
In conclusion, PD-MSCs induce mitochondrial activity (e.g., glycolysis and mitochondrial respiration rate), ion channel signaling and/ mitophagy factors and stabilize the mitochondrial membrane in trophoblasts. Additionally, PD-MSCs enhance trophoblast invasion.
The findings of this study suggest that PD-MSCs have a positive paracrine effect on trophoblast function through effects on mitochondrial function. Although our data is normal physiological system, it is can help understanding for therapeutic effect of PD-MSCs on early implantation as well as liver disease and ovary dysfunction.
[fig] 48: h, all the cells in the upper wells were removed with a cotton swab. The invading cells that had attached to the bottom side of the filter were fixed with methanol for 20 min and stained with Mayer's hematoxylin (Dako) at room temperature. The cell invasion ability was determined by counting the number of stained cells in nine randomly selected fields on the membranes at ×100 magnification. The number of invaded trophoblast cells cocultivated with PD-MSCs was normalized to that of control cells, and the data are presented as the mean invasion. 2.7 | Zymography HTR-8/SVneo cell culture supernatants were harvested to determine MMP-2/9 activity by zymography. The conditioned medium was separated in a 12% SDS-polyacrylamide gel electrophoresis gel supplemented with 0.5% gelatin (Sigma-Aldrich). The separated proteins were incubated for 30min in renaturation buffer (Bio-Rad Laboratories), rinsed and incubated in development buffer (Bio-Rad Laboratories) for 24 h at 37°C in an orbital shaker. The gels were stained with Coomassie Brilliant Blue R-250 (Bio-Rad Laboratories) solution for 3 h at room temperature and then distained with a buffer containing 10% acetic acid and 30% methanol (Merck) until the zymogen bands could be visualized. [/fig]
[fig] |: PD-MSC cocultivation induced the rapid turnover of trophoblast cellsTo determine the effect of PD-MSC cocultivation on trophoblast proliferation and death, we stained trophoblasts using trypan blue solution and counted the positively stained cells. After 12 h of culture, there was no difference in the number of stained trophoblast cells between the coculture and individual culture groups. However, PD-MSCs than in the group cultured alone(Figure 1a). In addition, the cell cycle distribution of trophoblasts was analyzed by using FACS analysis. As shown inFigure 1b, the population of cells in the sub-G1 phase, which related to DNA fragmentation and apoptosis, not different according to PD-MSC cocultivation. However, the population of cells in G1, S and G2/M phase, which related to cell growth and DNA replicate, were slightly increased in the group cocultured with PD-MSCs than in the group cultured alone(Figure 1b). Furthermore, weanalyzed the expression of genes related to cell survival and death by western blotting. The levels of p-ERK, which is involved in cell survival and proliferation, were increased in trophoblasts cocultured with PD-MSCs compared with trophoblasts cultured alone at 24 and 48 h. The gene expression of BAX and Bcl2, which regulate apoptotic cell death, was clearly different according to PD-MSC cocultivation. Interestingly, BAX gene expression in trophoblasts cocultured with PD-MSCs was significantly increased at 12 h but decreased at 24 and 48 h. Bcl2 gene expression was increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone (*p < .05, Figure 1c-f). These results show that PD-MSC cocultivation reduces trophoblast cell death and increases cell proliferation. [/fig]
[fig] Figure 2b: determine the effect of PD-MSC cocultivation on ROS generation in trophoblast cells, ROS levels were analyzed by DCF-DA ELISA and MitoSOX/Tracker staining. ROS levels were markedly increased in trophoblasts cocultured with PD-MSCs compared with trophoblasts cultured alone at 24 and 48 h (*p < .05, Figure 2a). Similarly, superoxide accumulation was notably increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone at 24 h, as shown in The HO-1/2 and SOD genes encode catalytic enzymes with cellular protective functions that regulate oxidative stress through effects on ROS. HO-1/2 mRNA expression was increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone (Figure S2a,b). The western blot analysis showed that HO-1 protein expression was increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone at 24 and 48 h (Figure 2c,d). In addition, HO-2 protein expression was significantly increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone at 12, 24, and 48 h (*p < .05, Figure 2c,e). Similar to the mRNA expression results, SOD1 protein expression was decreased in trophoblasts cocultured with PD-MSCs (*p < .05, Figure S2c and Figure 2c,ff). These findings suggest that PD-MSC cocultivation induces HIF1-α expression and regulates ROS levels in trophoblast cells. F I G U R E 1 PD-MSC cocultivation did not affect the trophoblast cell survival rate through rapid cell cycle progression. HTR-8/SVneo cells were cocultured with or without PD-MSCs for 12, 24, and 48 h and subjected to trypan blue staining, FACS analysis and western blot analysis. (a) Fold change (dead/live rate) in trypan blue-stained cells according to PD-MSC cocultivation. (b) Quantification of trophoblast cells in Sub-G1, G1, S, and G2/M phase according to PD-MSC cocultivation. (c) Levels of p-ERK, Bax and Bcl2, which are markers of cell survival and death, in trophoblast cells according to PD-MSC cocultivation. (d-f) Fold change in the expression of specific genes normalized to the internal control (i.e., GAPDH). Values represent the mean ± SD of three independent experiments; *p < .05, t test. White bar: control; black bar: PD-MSC cocultivation. *: Without versus with PD-MSC cocultivation at each time point. Control: trophoblasts cultured alone. FACS, fluorescence-activated cell sorting; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; PD-MSC, placenta-derived mesenchymal stem cell F I G U R E 2 PD-MSC cocultivation induced the expression of genes related to oxidative stress in trophoblasts. HTR-8/SVneo cells were cocultured with or without PD-MSCs for 12, 24, and 48 h and subjected to DCF-DA assays, MitoSOX/Tracker staining and western blot analysis. (a) Fold change in ROS levels in trophoblast cells according to PD-MSC cocultivation. (b) The MitoSOX/Tracker ratio in trophoblast cells according to PD-MSC cocultivation was determined as described in the Section 2. Representative confocal images are shown (magnification: ×200). (c) Expression of HO-1, HO-2, and SOD1, which are markers of oxidative stress, in trophoblast cells according to PD-MSC cocultivation. (d-f) Fold change in the expression of specific genes normalized to the internal control (i.e., GAPDH). Values represent the mean ± SD of three independent experiments; *p < .05, t test. White bar: control; black bar: PD-MSC cocultivation. *: Without versus with PD-MSC cocultivation at each time point. Control: trophoblasts cultured alone. GAPDH, glyceraldehyde 3-phosphate dehydrogenase; PD-MSC, placenta-derived mesenchymal stem cell; SOD, superoxide dismutase F I G U R E 3 3 PD-MSC cocultivation increased glycolysis in trophoblasts. HTR-8/SVneo cells were cocultured with or without PD-MSCs for 12, 24, and 48 h, and XF analysis was performed to measure glycolysis. (a) Schematic diagram of the glycolysis experiment using XF analysis. ECAR levels were measured by XF assays under basal conditions and after the sequential addition of glucose (10 mM), oligomycin (1 µM) and 2-DG (50 mM), as indicated. (b) The phenotype of trophoblast cells according to PD-MSC cocultivation. (c) Baseline ECAR, (d) glycolysis ability, (e) glycolytic capacity and (f) glycolytic reserve in trophoblasts according to PD-MSC cocultivation. Values represent the mean ± SD of three independent experiments; *p < .05, t test. White bar: control; black bar: PD-MSC cocultivation. *: Without versus with PD-MSC cocultivation at each time point. Control: trophoblasts cultured alone. ECAR, extracellular acidification rate; PD-MSC, placenta-derived mesenchymal stem cell 3.3 | PD-MSC cocultivation enhanced glycolysis in trophoblast cells To measure glycolysis in trophoblast cells according to PD-MSC cocultivation, the glycolysis XF analyzer was used to analyze live trophoblast cells. Glycolysis was measured based on the ECAR, which indicates a change in pH of the culture media of trophoblast cells. After PD-MSCs and trophoblast cells were cocultivated, the trophoblast cells were treated with glucose, oligomycin or 2-DG to analyze each step-in glycolysis individually. XF analysis was performed as shown in Figure 3a. The energetic phenotype was determined in trophoblasts cocultured with PD-MSCs compared to that in trophoblasts cultured alone at 24 and 48 h (Figure 3b). Comparative quantitative analysis of the results showed that the basal ECAR was significantly increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone at 12, 24, and 48 h (*p < .05, Figure 3c). Among trophoblasts treated with glucose, those cocultivated with PD-MSCs showed an increase in glycolysis ability compared to those cultured alone (Figure 3d). Among the trophoblast cells treated with oligomycin, those cocultivated with PD-MSCs showed an increase in glycolysis capacity at 48 h compared to those cultured alone (*p < .05, Figure 3e). After 2-DG treatment, the trophoblast cells cocultivated with PD-MSCs had an increased glycolytic reserve at 24 h (*p < .05, Figure 3f). These findings suggest that PD-MSC cocultivation activates glycolysis in trophoblast cells. the effect of PD-MSC cocultivation on the mitochondrial OCR of trophoblast cells, the mitochondrial OCR was analyzed in live trophoblast cells. Mitochondrial stress levels can be deduced by the OCR, which is detected based on changes in the pH of trophoblast cell culture media. Trophoblast cells cocultivated with PD-MSCs were treated with oligomycin, FCCP or rotenone/antimycin A to analyze each step of mitochondrial oxygen consumption individually. XF analysis was performed as shown in Figure 4a. The results revealed the energetic and glycolytic phenotypes of trophoblasts cocultured with PD-MSCs compared to those of trophoblasts cultured alone (Figure 4b). The comparative quantitative analysis revealed that the basal OCR at 24 and 48 h was significantly increased in trophoblasts cocultured with PD-MSCs compared to trophoblasts cultured alone (*p < .05, Figure 4c). After oligomycin treatment, trophoblast cells cocultivated with PD-MSCs showed significantly increased ATP production at 12 h (Figure 4d). After FCCP treatment, the maximal respiration of trophoblast cells was markedly increased in the PD-MSC cocultivation group compared to the individual culture group at 12 h (*p < .05, Figure 4e). In contrast, the spare capacity of trophoblast cells was not significantly different under the two culture conditions (Figure 4f). These findings suggest F I G U R E 4 4 PD-MSC cocultivation regulated mitochondrial respiration in trophoblasts. HTR-8/SVneo cells were cocultured with or without PD-MSCs for 12, 24, and 48 h, and XF analysis was performed to measure mitochondrial respiration. (a) Schematic diagram of the mitochondrial stress experiment (i.e., mitochondrial respiration) using XF analysis. The OCR was measured by XF assay under basal conditions and after the sequential addition of oligomycin (1 µM), FCCP (0.5 µM) and rotenone A/A (0.5 µM), as indicated. (b) The phenotype of trophoblast cells according to PD-MSC cocultivation. (c) Baseline OCR, (d) ATP production, (e) maximal respiration and (f) spare capacity were analyzed in trophoblasts according to PD-MSC cocultivation. Values represent the mean ± SD of three independent experiments; *p < .05, t test. White bar: control; black bar: PD-MSC cocultivation. *: Without versus with PD-MSC cocultivation at each time point. Control: trophoblasts cultured alone. ATP, adenosine triphosphate; OCR, oxygen consumption rate; PD-MSC, placenta-derived mesenchymal stem cell that PD-MSC cocultivation induces mitochondrial oxygen consumption and ATP production in trophoblast cells. [/fig]
[fig] F: I G U R E 5 PD-MSC cocultivation changed gene expression and mitochondrial function with effect on stress levels and ion channels. HTR-8/ SVneo cells were cocultured with or without PD-MSCs for 12, 24, and 48 h and subjected to western blotting and qRT-PCR analyses. (a) Expression of HSP60, PHB1, and VDAC, which are markers of mitochondrial stress, in trophoblast cells according to PD-MSC cocultivation. (b-d) old change in the expression of specific genes normalized to the internal control (i.e., GAPDH). (e, f) IP3R and MCU mRNA expression in trophoblast cells according to PD-MSC cocultivation. Values represent the mean ± SD of three independent experiments; *p < .05, t test. White bar: control; black bar: PD-MSC cocultivation. *: Without versus with PD-MSC cocultivation at each time point. Control: trophoblasts cultured alone. GAPDH, glyceraldehyde 3-phosphate dehydrogenase; MCU, mitochondrial calcium uniport; mRNA, messenger RNA; PD-MSC, placenta-derived mesenchymal stem cell; qRT-PCR, quantitative real-time polymerase [/fig]
|
Clinical and Pathological Characteristics of Patients With Nonproteinuric Diabetic Nephropathy
Introduction: As the most common complication of diabetes mellitus (DM), diabetic nephropathy (DN) was initially considered to begin with proteinuria preceding the progression of renal insufficiency. This clinical paradigm has been questioned in the late decades, as many DM patients without proteinuria have progressive renal insufficiency. However, the characteristics of nonproteinuric DN were not fully clear yet.Patients and Methods: A total of 390 patients with renal biopsy-proven DN in our center were retrospectively recruited in the current study. Clinical and histopathological data of the patients were analyzed. We used propensity score-matching methods to address the imbalance of age, sex, and diabetes duration for comparative analyses.Results: Among all the renal biopsy-proven DN patients with renal biopsy proven DN, 18 patients were classified as nonproteinuric DN. Compared with 36 propensity scorematched proteinuric DN patients, diabetic retinopathy (DR) was less frequent in nonproteinuric DN patients (38.9% vs. 66.4%, p<0.05). During the follow-up of 24.0 (12.0-42.0) months, the probability of developing the end-stage renal disease (ESRD) was significantly lower in nonproteinuric DN patients than in proteinuric ones in both the propensity score-matched cohort and overall cohort (log-rank test, p<0.001 and p<0.001, respectively).Conclusions: Compared with proteinuric DN patients, DR was less frequent in nonproteinuric DN patients. Nonproteinuric DN patients had better renal outcomes than proteinuric DN patients.
# Introduction
Diabetic nephropathy (DN) is the most common complication of diabetes mellitus (DM) and the leading cause of end-stage renal disease (ESRD) in China. DN was initially considered to begin with proteinuria preceding the progression of renal insufficiency [estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m 2 ]. The natural history was divided into normoalbuminuria (urinary albumin-to-creatinine ratio [UACR] <30 mg/g), microalbuminuria (UACR 30-300 mg/g), and macroalbuminuria (UACR >300 mg/ g), which was mainly based on the typical progression course of type 1 DM (4).
However, this concept of the clinical paradigm has changed over the last decades, and it has been noted that DM patients without proteinuria could also have progressive renal insufficiency. Therefore, the latest diagnostic criteria for diabetic kidney disease (DKD) include low eGFR or the persistent presence of elevated urinary albumin excretion (albuminuria). Nonproteinuric DKD was defined as an eGFR <60 mL/min/1.73 m 2 with a UACR <300 mg/g (6-10). As a diagnosis term, DKD covered both clinical diagnosis and histological diagnosis (DN).
The characteristics of nonproteinuric DN patients are not yet thoroughly investigated. Previous studies showed that the renal histopathological findings of DN are heterogeneous regardless of the level of GFR or UACR. According to the previous results, we speculated that nonproteinuric DN patients might have typical histopathological features of DN and a lower risk of CKD progression. Therefore, in the current study, using the cohort of our center and propensity score-matching methods, we investigated clinicopathological characteristics and outcomes in patients with the nonproteinuric phenotype of DN in comparison with patients with the classical proteinuric DN.
# Patients and methods
## Patients
A total of 390 DM patients with renal biopsy-proven DN who were diagnosed from January 1, 2015, to December 31, 2020, were analyzed retrospectively. DM was defined according to the criteria proposed by the American Diabetes Association in 2017. The investigation was conducted according to the Declaration of Helsinki and was approved by the Ethics Committee of Peking University First Hospital (2017-1280). Written informed consent was obtained from each participant.
Among the 390 patients with renal biopsy-proven DN, 298 were male and 92 were female, with an age of 53.11 ± 12.59 years at renal biopsy. The median level of UACR was 2718.56 (1195.57-4897.83) mg/g. Of the 390 patients, 167 patients who had coexisting non-diabetes-related renal disease, including 54 patients with membranous nephropathy, 45 patients with IgA nephropathy, 15 patients with immune complex-mediated glomerulonephritis, 10 patients with ANCA-associated glomerulonephritis, 7 patients with C3 glomerulonephritis, 6 patients with IgG4-related kidney disease and 30 patients with other renal diseases, were excluded. The comparison between patients with and without coexisting non-diabetes-related renal disease was provided in Supplementary Table 1, . 55/390 patients with eGFR>60 ml/min/1.73m 2 were excluded. Ultimately, 168 patients were eligible for further analysis for different proteinuria groups. Among them, 18/168 patients were classified as nonproteinuric DN (UACR <300 mg/g) and 150/168 patients were classified as proteinuric DN .
## Clinical characteristics
The clinical data of these patients at the time of renal biopsy and during follow-up were systematically recorded, including age, sex, diabetic retinopathy (DR), use of renin-angiotensin-aldosterone system (RAAS) inhibitors, hemoglobin, serum creatinine (Scr), eGFR, serum albumin, fasting blood glucose (FBG), HbA1c, triglyceride (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), and plasma complements. Proteinuria was expressed as the UACR. Nonproteinuric DN was defined as an eGFR <60 mL/min/1.73 m 2 with a UACR <300 mg/g at the time of renal biopsy according to the previously described criteria. Cardiovascular disease (CVD) history was self-reported and included a history of congestive heart failure, coronary heart disease, heart attack, angina, stroke, or periphery atherosclerosis. The eGFR was calculated using the CKD-EPI equation. HbA1c levels were measured using a high-performance liquid chromatographic assay.
## Renal histopathology
Renal specimens were evaluated using direct immunofluorescence (for immunoglobulins and complement components), light microscopy, and electron microscopy. Periodic acid-Schiff (PAS), silver methenamine, hematoxylin and eosin (HE), and Masson's trichrome staining were used for light microscopy. Biopsies were scored independently by two pathologists. A standard classification system was used based on histological scores for glomerular lesions, tubulointerstitial lesions, vascular lesions and non-diabetic renal lesions.
Diabetic glomerulopathy is classified as class I through IV according to the Renal Pathology Society in 2010. Interstitial fibrosis and tubular atrophy (IFTA) were scored semiquantitatively based on the proportion of the tubulointerstitial compartment affected (0, none; 1, <25%; 2, 25-50%; 3, >50%). Interstitial inflammation was scored semi-quantitatively (0, absent; 1, infiltration only in areas related to IFTA; 2, infiltration in areas without IFTA). Vascular lesions were scored according to the presence of arteriolar hyalinosis and large-vessel arteriosclerosis (grades 0-1). For direct immunofluorescence, the intensities of staining of immunoglobulins, complements, fibrin-associated antigen (FRA), and albumin (Alb) were semi-quantitatively graded on a scale of 0-4+. Outcomes ESRD was defined as the initiation of hemodialysis/peritoneal dialysis, renal transplantation, or death due to uremia. The patients were followed up until the end of 2020 or ESRD, whichever came first. New-onset CVD events included congestive heart failure, coronary heart disease, heart attack, angina, stroke, or periphery atherosclerosis until 2020.
# Statistical analysis
Normally distributed data were presented as mean ± standard deviation, while non-normally distributed data were presented as median values with an inter-quartile range (IQR). Categorical variables were expressed as percentages or ratios. Chi-square, one-way analysis of variance (ANOVA), and t-tests were performed as appropriate. Differences in semi-quantitative and quantitative parameters that were not normally distributed were assessed using Kruskal-Wallis or Mann-Whitney U-tests, as appropriate. Differences were considered significant if the pvalue was <0.05. In the current study, the sample size of patients with nonproteinuric DN (n=18) was relatively small compared with the proteinuric DN patients (n=150). We conducted propensity score matching analysis to address the imbalance of background factors such as age, sex, and diabetes duration that affect outcomes. We matched the nonproteinuric DN group with the proteinuric DN group using propensity scores with a one-totwo nearest-neighbor caliper width of 0.01, which is the maximum allowable difference in propensity scores. Analyses were performed using the SPSS statistical software package (version 11.0; Chicago, IL, USA) and R studio 4.0.2.
# Results
## General data of the patients at renal biopsy
General data at the renal biopsy of the whole cohort of 390 DN patients were listed in
## Comparison of clinical manifestations
Clinical features of patients stratified by proteinuria before and after propensity score matching are shown in. For example, advanced DN pathology manifestations (class III and class IV) were observed in only 4/18(22.2%) of nonproteinuric DN patients, whereas they were found in 27/36(75.0%) of matched proteinuric ones. No significant difference in tubulointerstitial damage was found between the two matched groups. The proportion of patients with arteriolar hyalinosis was significantly lower in the nonproteinuric DN group than in matched proteinuric group (66.7% vs. 88.9%, p<0.05). All nonproteinuric and proteinuric DN patients showed arteriosclerosis in the kidneys.
Regarding direct immunofluorescence, there were significantly lower proportions of IgM and C1q depositions in nonproteinuric DN patients than in matched proteinuric ones (11.1% vs. 77.8%, p<0.001 and 0.0% vs. 58.3%, p<0.05, respectively). A significantly higher proportion of C3 deposition was found in patients with proteinuria in the overall cohort (44.4% vs. 72.0%, p<0.05).
## Outcomes
During a median follow-up duration of 24.0 (12.0-42.0) months, none of the nonproteinuric DN patients progressed to ESRD, whereas 21/36 (65.6%) of the matched proteinuric DN patients progressed to ESRD. Among the patients with proteinuria from the overall cohort, 92/150 (61.3%) progressed to ESRD. Kaplan-Meier analysis showed that the probability of developing ESRD was significantly lower in nonproteinuric DN patients than in proteinuric ones in both the propensity score-matched cohort and overall cohort (log-rank test, p<0.001 and p<0.001, respectively). Only 1/18 patients with nonproteinuric DN and 22/150 patients with proteinuria DN had new-onset CVD in the current study (P>0.05), which might be due to the relatively short follow-up.
# Discussion
DN is the leading cause of ESRD and is associated with increased cardiovascular morbidity and all-cause mortality. Traditionally, persistent microalbuminuria has been considered the first clinical sign of DN, inevitably progressing to macroalbuminuria and subsequent renal dysfunction. However, over recent decades, there has been increasing recognition that GFR reduction may precede the development of proteinuria in several patients with diabetes . These patients were therefore defined as nonproteinuric DKD/ DN. The prevalence of proteinuric DKD declined, while the prevalence of nonproteinuric DKD increased, attributable to a higher rate of RAAS inhibitors prescription. Although the paradigm has been renewed, the characteristics of nonproteinuric DN have not been thoroughly investigated.
In patients with DKD, the prevalence of nonproteinuria varies between 20% and 40%. In the current study, a total of 18/ 223 (8.1%) DN patients were classified as nonproteinuric DN, which was lower than that in previous reports. Of the patients with reduced eGFR (<60 mL/min/1.73 m²) from the National A B Values are expressed as a mean ± standard deviation, percentage or median with upper and lower quartile or percentage. Chi-square tests were performed in percentages or ratios variables.
Health and Nutrition Examination Survey (NAHNES III) in 2003, 81% had nonproteinuric DKD, and only 19% had proteinuria. In the UK Prospective Diabetes Study (UKPDS-74), during 15 years of follow-up in 4,006 patients with type 2 diabetes, 1,132 (28.3%) developed renal impairment. Of the latter, 575 (50.8%) patients were classified as nonproteinuric DKD. We have noted that all patients in the current study underwent renal biopsy, which was not highly recommended in nonproteinuric DKD patients unless they were suspected of having either superimposed non-diabetic kidney disease or de novo non-diabetic kidney disease. The relatively lower prevalence of nonproteinuric DN patients in the current study might be associated with the lower rate of renal biopsy in this subgroup of patients. In summary, the prevalence of nonproteinuric DKD is not low. The traditional nonproteinuric DKD should also be paid attention and concern on, mainly due to lower eGFR and renal insufficiency. Compared with proteinuric DN patients, a significantly lower proportion of DR in nonproteinuric DN patients was found in both the overall and the matched cohorts. The prevalence of DR in patients with nonproteinuric DKD varies across studies. A study from RIACE with 2,959 DKD patients found that 2,028 (68.5%) patients did not have DR, and 538 patients (18.2%) showed both proteinuria and retinopathy. The varying prevalence of DR suggests that the development of nonproteinuric DKD may be independent of diabetic microangiopathic lesions.
Only a limited number of studies have investigated the renal histopathological features of nonproteinuric DN. Results from previous biopsy-based studies were inconsistent, which may be due to the small sample size and the timing of renal biopsy. Studies of the renal histopathology in patients with type 2 DM showed that nonproteinuric patients had less frequent typical glomerular injuries. The findings were not consistent for tubulointerstitial and arterial injuries. Yamanouchi et al. reported that patients with nonproteinuric DN have both milder glomerular injuries and tubulointerstitial injuries. In the current study, consistent with previous reports, most of the nonproteinuric DN patients showed typical but milder glomerular injuries, including mesangial expansion and nodular sclerosis (Kimmelstiel-Wilson lesions), while tubulointerstitial injuries were heterogeneous. More importantly, these results suggest that typical glomerular injuries may precede overt proteinuria in DN. For immunofluorescence, there was a significantly lower proportion of IgM and C1q deposition in nonproteinuric DN patients compared with matched proteinuric DN patients. A higher proportion of C3 deposition was found in patients with proteinuria in the overall cohort. Previous studies have shown that complement deposition in renal histopathology is associated with severe kidney damage in DN patients. Persistent proteinuria may induce local complement activation and aggravate renal injury. The pathogenic role of complement overactivation warrants further investigation.
In the current study, the renal outcome was more favorable in nonproteinuric DN patients than those with proteinuria. None of the nonproteinuric DN patients progressed to ESRD. These results were consistent with previous studies. Proteinuria remains a crucial independent predictor of eGFR decline in DM patients, especially those with low eGFR. However, even if the risk for ESRD was low, nonproteinuric patients showed an equal or even higher risk of CVD morbidity and mortality than those with proteinuria. The results suggest that nonproteinuric DN may represent a distinct phenotype, with macroangiopathic and tubulointerstitial lesions instead of microangiopathic lesions involved in the underlying pathology. Close attention and care for CVD morbidity and mortality in these patients are needed.
This study has some limitations. First, the sample size was small, and the follow-up duration was short for assessing the probability of developing ESRD. The current study was a singlecenter study that recruited only 18 nonproteinuric DN patients. Therefore, the true prevalence of nonproteinuric DKD cannot be accurately assessed. Second, there was an inevitable bias in patients receiving renal biopsy. Third, we only referred to Chinese DN patients in the current study. Studies involving multi-ethnic and multi-center are needed.
# Conclusion
In conclusion, compared with proteinuric DN patients, DR was less frequent in nonproteinuric DN patients. Nonproteinuric DN patients had better renal outcomes than proteinuric patients. Multicenter studies with larger sample sizes are needed to further understand nonproteinuric DN.
# Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
# Ethics statement
The studies involving human participants were reviewed and approved by Ethics Committee of Peking University First Hospital (2017-1280). The patients/participants provided their written informed consent to participate in this study.
# Author contributions
D-YC and MC designed the study. D-YC, M-RL, X-JY, and S-XW contributed data. D-YC and MC drafted the analysis plan. D-YC performed the statistical analysis. D-YC wrote the manuscript. MC and M-HZ revised the manuscript and supervised the study. MC is the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. All authors contributed to the article and approved the submitted version.
# Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fendo.2021. 761386/full#supplementary-material
|
Colorimetric detection of Shewanella oneidensis based on immunomagnetic capture and bacterial intrinsic peroxidase activity
[bib_ref] Respiration of metal (hydr) oxides by Shewanella and Geobacter: a key role..., Shi [/bib_ref] [bib_ref] Global transcriptome analysis of Shewanella oneidensis MR-1 exposed to different terminal electron..., Beliaev [/bib_ref] [bib_ref] Genome sequence of the dissimilatory metal ion-reducing bacterium Shewanella oneidensis, Heidelberg [/bib_ref] [bib_ref] c-Type cytochrome-dependent formation of U (IV) nanoparticles by Shewanella oneidensis, Marshall [/bib_ref] [bib_ref] Manganese sulfide formation via concomitant microbial manganese oxide and thiosulfate reduction, Lee [/bib_ref] [bib_ref] Isolation of a high-affinity functional protein complex between OmcA and MtrC: two..., Shi [/bib_ref] [bib_ref] Global transcriptome analysis of Shewanella oneidensis MR-1 exposed to different terminal electron..., Beliaev [/bib_ref] [bib_ref] Biodecolorization of Naphthol Green B dye by Shewanella oneidensis MR-1 under anaerobic..., Xiao [/bib_ref] [bib_ref] Extracellular respiration of dimethyl sulfoxide by Shewanella oneidensis strain MR-1, Gralnick [/bib_ref] [bib_ref] Immunomagnetic capture rt-PCR for detection of norovirus from foods implicated in a..., Kobayashi [/bib_ref] [bib_ref] Sensitive detection of biotoxoids and bacterial spores using an immunomagnetic electrocheminescence sensor, Gatto-Menking [/bib_ref] [bib_ref] Detection of Clostridium perfringens type A enterotoxin in faecal and food samples..., Cudjoe [/bib_ref] [bib_ref] Enzymatic Assay for Cu (II) with Horseradish Peroxidase and Its Application in..., Xianyu [/bib_ref] [bib_ref] Irregular-shaped platinum nanoparticles as peroxidase mimics for highly efficient colorimetric immunoassay, Gao [/bib_ref] [bib_ref] Enhanced colorimetric detection on porous microarrays using in situ substrate production, Goff [/bib_ref] [bib_ref] Horseradish peroxidase: a modern view of a classic enzyme, Veitch [/bib_ref] [bib_ref] Probing structure-function relations in heme-containing oxygenases and peroxidases, Dawson [/bib_ref] [bib_ref] Detection of c-type cytochromes using enhanced chemiluminescence, Vargas [/bib_ref] [bib_ref] Haem staining in gels, a useful tool in the study of bacterial..., Goodhew [/bib_ref] [bib_ref] Localization of cytochromes to the outer membrane of anaerobically grown Shewanella putrefaciens..., Myers [/bib_ref]
# Results
The detection mechanism of colorimetric immunomagnetic assay. [fig_ref] Figure 1 |: Schematic for the antibody/MBs preparation and colorimetric immunomagnetic assay [/fig_ref] depicts the principle of the colorimetric immunomagnetic assay for detecting S. oneidensis cells. S. oneidensis cells were first captured by anti-S. oneidensis antibody-conjugated MBs (antibody/MBs) through antigen-antibody reaction, yielding bacteria/ MBs complexes. Then, the bacteria/MBs complexes were attracted by a magnet, separated from the sample matrix, and finally transferred to a microplate for further colorimetric assay.
[formula] H 2 O 2 zTMB red ÀÀ? HRP 2H 2 OzTMB oxð1ÞH 2 O 2 zTMB red ÀÀÀÀÀÀÀÀÀÀ? c{type cytochrome 2H 2 OzTMB oxð2Þ [/formula]
In a conventional HRP/TMB system (Equation (1)), HRP (a heme-containing protein) catalyzes the oxidation and reduction reaction between TMB and hydrogen peroxide, yielding a characteristic strong absorption at 650 nm. In this study, c-type cytochromes such as OmcA and MtrC, located at the outer membrane of S. oneidensis cells, are similar in structure to HRP and play a similar role (Equation [bib_ref] Ecology and biotechnology of the genus Shewanella, Hau [/bib_ref] in the interaction between TMB and hydrogen peroxide. Consequently, a blue color developed in the S. oneidensis/ MBs/TMB/H 2 O 2 system, which provides a colorimetric detection method for S. oneidensis cells. The optical density of color-developing reaction product was measured with a microplate reader. Based on the obtained data, a regression model was developed to describe the relationship between S. oneidensis cell number and optical density.
Characterization of the antibody/MBs-S. oneidensis interaction. As shown in [fig_ref] Figure 2 |: Optical and SEM images of S [/fig_ref] , the isolated S. oneidensis cells were observed with an optical microscope (inset). Attachment of S. oneidensis to antibody/MBs was further confirmed with a scanning electron microscope (SEM). These results suggest that S. oneidensis cells bind to antibody/MBs due to antigen-antibody recognition. To evaluate the In addition, the antibody/MBs-captured S. oneidensis cells were investigated to be capable of a conventional color reaction where peroxidase activity was responsible for catalyzing the oxidization of TMB. As shown in [fig_ref] Figure 3 |: Peroxidase-like activities of S [/fig_ref] , similar to S. oneidensis alone [fig_ref] Figure 3 |: Peroxidase-like activities of S [/fig_ref] and cytochrome c control [fig_ref] Figure 3 |: Peroxidase-like activities of S [/fig_ref] , S. oneidensis/MBs complex [fig_ref] Figure 3 |: Peroxidase-like activities of S [/fig_ref] displayed a color developing ability when immersed in the substrate solution containing H 2 O 2 and TMB. The color product was further characterized by UV-vis spectroscopy. As shown in [fig_ref] Figure 3 |: Peroxidase-like activities of S [/fig_ref] , for Curve a, b and c, a strong absorption peak appeared at 650 nm, which is the characteristic absorption peak of the product from peroxidasecatalyzed H 2 O 2 /TMB reaction. Yet, there was neither blue color developed nor characteristic optical absorption for the MBs control [fig_ref] Figure 3 |: Peroxidase-like activities of S [/fig_ref] , Curve d) or the H 2 O 2 /TMB solution control [fig_ref] Figure 3 |: Peroxidase-like activities of S [/fig_ref] , Curve e). Although previous work has showed that there are peroxidase-like activities in magnetic particles [bib_ref] Fe 3 O 4 magnetic nanoparticles as peroxidase mimetics and their applications..., Wei [/bib_ref] , the MBs used in this study presented almost no peroxidase activity. This may be attributed to the special structure of polystyrene MBs which are produced by coating the silica core first with a layer of magnetic particles and then an additional surface modification layer. These results demonstrated the feasibility of the proposed colorimetric immunomagnetic assay for rapid detection of S. oneidensis cells.
Optimization of immunoassay conditions. To obtain the best detection sensitivity, parameters including antibody/MBs concentration, H 2 O 2 concentration, immune-reaction time and color developing time were optimized using 5.0 3 10 6 CFU/mL S. oneidensis cells as analyte. [fig_ref] Figure 4 |: Optimization of the working conditions [/fig_ref] shows the influence of antibody/MBs concentration on detection signal. The OD value first increased and then dropped with increasing antibody/MBs amount, reaching its maximum when 40 mL of the antibody/MBs was used. Hence, 40 mL of antibody/MBs was selected for further studies. The proposed colorimetric assay relies on the bacteria intrinsic peroxidase activity whose sensitivity is affected by the concentration of H 2 O 2 . Herein, the effect of H 2 O 2 concentration on the detection signal was evaluated with H 2 O 2 concentration ranging from 0.1 mM to 5 mM. Similarly, the OD value, associated with the amount of target bacteria, first increased and then dropped with increasing H 2 O 2 concentration, reaching its maximum when H 2 O 2 concentration was 2 mM [fig_ref] Figure 4 |: Optimization of the working conditions [/fig_ref]. As a result, 2 mM H 2 O 2 was used for subsequent trials. The immunoreaction time can impact the detection sensitivity. As shown in [fig_ref] Figure 4 |: Optimization of the working conditions [/fig_ref] , the OD value increased gradually with increasing reaction time from 10 min to 50 min and leveled off after that, indicating that the antigen-antibody reaction reached its equilibrium after 50 min. Thus, in following experiments, reactions were allowed to continue for 50 min. For colorimetric immunoassay, the color developing time can strongly influence detection sensitivity. [fig_ref] Figure 4 |: Optimization of the working conditions [/fig_ref] shows that the peak value increased with color developing time, and the maximum absorbance occurred at 30 min. After that, color began to fade due to the precipitation of TMB out of the solution. Consequently, 30 min was selected to be the optimal color developing time.
Specificity and reproducibility. To investigate the specificity of the proposed colorimetric immunomagnetic assay, the proposed detecting procedure was performed on both S. oneidensis and control organisms. [fig_ref] Figure 5 |: Specificity of the colorimetric immunomagnetic assay [/fig_ref] shows the absorbance for samples containing 5.0 3 10 6 CFU/mL S. oneidensis, Shewanella putrefaciens (S. putrefaciens), Shewanella decolorationis (S. decolorationis), Pseudomonas agglomerans (P. agglomerans), Aeromonas hydrophila (A. hydrophila), Staphylococcus aureus (S. aureus), E. coli or B. subtilis. There were minor changes in absorbance of the control samples containing P. agglomerans, A. hydrophila, S. aureus, E. coli and B. subtilis as compared with that of the blank control (pure water). For samples containing phylogenetically close Shewanella species such as S. putrefaciens and S. decolorationis, there were only weak detection signals, which can be attributed to the partial cross-reaction between antibody/MBs and bacteria that shared the conservative epitopes with Shewanella species. However, for samples containing S. oneidensis, the response signal increased drastically.
In order to evaluate the effects of sample matrix components on the proposed assay method, we further challenged the system with river samples spiked with different microorganisms. Analyte solutions were prepared by adding S. oneidensis (5.0 3 10 6 CFU/mL), E. coli (5.0 3 10 6 CFU/mL), B. subtilis (5.0 3 10 6 CFU/mL), or a mixture of the three bacteria (each at 5.0 3 10 6 CFU/mL) into river water samples collected from a river near our institute. As indicated in [fig_ref] Figure 6 |: The effects of sample matrix components on the colorimetric immunomagnetic assay [/fig_ref] , higher absorbance was observed for samples with S. oneidensis than for those without S. oneidensis. The present of other sample matrix components seems to exert no significant interfering effects on the response signal to target S. oneidensis. The results clearly suggested the high selectivity of the proposed immunoassay to S. oneidensis.
In addition, the assay method was tested for its reproducibility. Five repetitive measurements were conducted on S. oneidensis of 5.0 3 10 6 CFU/mL. The method was demonstrated to be highly reproducible with a relative standard deviation of 5.41%.
Analytical performance. Under the optimized working conditions, the sensitivity and detection range of the colorimetric immunomagnetic assay were determined with varying amounts of S. oneidensis cells. As shown in [fig_ref] Figure 7 |: The regression curve for the detection of S [/fig_ref] , the absorbance increased linearly as the amount of S. oneidensis cells increased from 5.0 3 10 3 to 5.0 3 10 6 CFU/mL. The relationship can be described as Y 5 1.133 27 X 1 0.027 with a correlation coefficient of 0.9988 (R 2 ), where Y is the absorbance and X is the amount of S. oneidensis cells (CFU/mL). The detection limit was estimated to be 5.0 3 10 3 CFU/mL according to the signal-to-noise ratio of three. The results suggested that the proposed assay method was sensitive for detecting S. oneidensis cell with a wide detection range.
Real sample analysis. In order to evaluate the analytical reliability and potential applications, the proposed colorimetric immunomagnetic assay was used to detect S. oneidensis level in real water samples. A series of samples were prepared by spiking river water samples collected from a local river with different amounts of S. oneidensis cells. Sequentially, the prepared samples were analyzed using the proposed colorimetric assay and the results are summarized in Supplementary Table S1 online. The recovery obtained is in the range between 83.32% and 102.40% with the relative standard deviation between 1.63% and 11.98%. The results indicate that the proposed assay method is feasible for S. oneidensis detection in environmental samples.
# Discussion
S. oneidensis inhabits various environments, such as lake [bib_ref] Bacterial manganese reduction and growth with manganese oxide as the sole electron..., Myers [/bib_ref] , ocean [bib_ref] Towards environmental systems biology of Shewanella, Fredrickson [/bib_ref] water and marine sediments [bib_ref] Low-temperature growth of Shewanella oneidensis MR-1, Abboud [/bib_ref]. To estimate its numbers, microbiological ecology methods are currently employed based on DNA extracting and PCR testing. Unfortunately, a problem with DNA extraction is that the subsequent PCR testing is easily inhibited by the impurities that are present in the extracted DNA templates [bib_ref] Purification and characterization of a common soil component which inhibits the polymerase..., Watson [/bib_ref]. In spite of high specificity to target cells in environmental samples [bib_ref] Abundance and phylogenetic affiliation of iron reducers in activated sludge as assessed..., Nielsen [/bib_ref] , reverse transcription fluorescent in situ hybridization, which utilizes sequence-specific nucleic acid labeled with fluorophores as probes, involves a very time-consuming and labor-intensive sample treatment procedure, including fixation, hybridization and washing. Fortunately, the recently developed IMC technique is not only simple and time-saving, but efficient in removing inhibitory substances and heterogeneous bacteria in samples [bib_ref] Immunomagnetically captured thermophilic sulfate-reducing bacteria from North Sea oil field waters, Christensen [/bib_ref] [bib_ref] Selective recovery of Streptosporangium fragile from soil by indirect immunomagnetic capture, Mullins [/bib_ref]. In addition, IMC is non-destructive for the target cells. This is a very important characteristic that allows subsequent detection of such bacteria as S. oneidensis with bacterial intrinsic enzyme activity.
This IMC followed by bacterial intrinsic peroxidase activity-catalyzed redox-reduction reaction can be useful to detect S. oneidensis cells in the environment. Our results showed a detection limit of 5.0 3 10 3 CFU/mL toward S. oneidensis. Compared to recently developed methods such as surface-enhanced Raman spectroscopy and microarray hybridization (see online), our method showed competitive or better performance. Compared with a previously reported electrochemical immunoassay 32 , the advantages of this method includes speed and minimum sample manipulation. In addition, a previous study revealed that even lowbiomass water samples from a bioremediation site contained cells at a density of 3 3 10 6 cells/mL 33 , which indicates the method proposed here is good enough to detect and monitor bioremediation processes. Therefore, the present work provides an alternative approach for S. oneidensis detection.
Our method relies on IMC for concentrating and isolating target organism from sample matrix. Since selective separation is achieved by using specific antibodies-attached magnetic spheres, the efficacy of this method can be affected by the properties of antibodies and magnetic spheres. In the study, commercially available polystyrene magnetic microspheres were employed to prepare immunomagnetic beads. These superparamagnetic magnetic particles with homogeneous sizes ensured the reproducibility of the detection. Polyclonal antibodies raised against S. oneidensis cell, in the study, are directed against multiple epitopes rather than a single epitope, which results in reliable immunological recognition to the target cells. With these magnetic microspheres and polyclonal antibodies, the prepared antibodies/MBs would enable a reliable and reproducible S. oneidensis detection. Virtually, the antibodies/MBs do not react with the bacteria 34 usually found in river waters and the environment such as P. agglomerans, A. hydrophila, S. aureus, E. coli and B. subtilis, and only partially cross-react with phylogenetically close Shewanella species such as S. putrefaciens and S. decolorationis. In addition, the magnet strength also affects IMC efficacy. In fact, a homemade magnetic separator equipped with strong permanent magnets (8000 Gauss) was employed in the experiment, and it was efficient enough to completely separate the magnetic particles from the solution within 5 min. Although previous reports have shown that sample matrix would affect IMC performance 35 , our study demonstrated that sample matrix (river water spiked with E. coli or B. subtilis) exerts only minor interfering effects on detection.
The color developing procedure was designed for colorimetric detection of the bead-captured S. oneidensis cells. This bacterial outer membrane peroxidase activity-based signal amplification eliminates the need for enzyme or fluorophores-conjugated antibody or PCR amplification to identify the captured cells, and enables easy detection of the target cells in a micro-well plate either qualitatively with naked eyes or quantitatively by using an OD reader. Although we have only demonstrated S. oneidensis cell detection, this strategy can be applied to other metal-reducing microorganisms, such as Geobacter sulfurreducens, Shewanella decolorationis and Shewanella putrefaciens, by substituting the antibody immobilized on the magnetic bead with antibodies that are specific to the target bacteria. Coupled with an automated immunomagnetic separation system, this detection strategy can be an alternative method for high throughput analysis of metal-reducing bacteria in environmental samples.
In summary, a novel signal amplification strategy for rapid detection of S. oneidensis was successfully designed. This is the first study that coupled bacterial intrinsic peroxidase activity with IMC. The primary advantages of this method include sensitivity, speed, and minimum sample manipulation, which are realized via the highefficiency of IMC and signal amplification of bacteria extracellular peroxidase. This method offers a technique for determining functional microbes in bioremediation.
# Methods
Microbial strains and culture. S. oneidensis (ATCC700550) was grown from a glycerol stock in Luria-Bertani broth as described in a previous report 32 . The S. oneidensis cell culture was centrifuged at 5000 g for 20 min and washed three times with 0.01 M phosphate buffer solution (PBS, pH 7.4). The collected cell precipitate was resuspended in PBS and stored at 4uC for further use. Before each measurement, cell solution was serially diluted with PBS and determined via agar plate count for cell number. E. coli and B. subtilis were used as control.
Antibody production and antibody/MBs preparation. S. oneidensis cells were processed to yield whole-cell antigens. Briefly, cell solution was centrifuged at 5000 g for 20 min and cells were resuspended to a concentration of approximately 10 9 cells/ mL. The cells were treated with formaldehyde for 30 min at room temperature, washed three times with PBS, resuspended in PBS to a concentration of 10 8 cells/mL and stored at 4uC. Using the prepared whole-cell antigen, a female New Zealand white rabbit weighing 2 kg was immunized intramuscularly for a total of four times at intervals of ten days for producing polyclonal antibody against S. oneidensis. The obtained antisera were purified with ammonium sulfate precipitation and characterized by enzyme-linked immunosorbent assay according to a previous protocol [bib_ref] Preparation and characterization of egg yolk immunoglobulin Y specific to influenza B..., Wen [/bib_ref]. The purified anti-S. oneidensis antibodies were stored at 220uC.
To prepare antibody/MBs, anti-S. oneidensis antibodies were attached to carboxylate-functionalized MBs (MBs-COOH) (Affimag PSC, 1-2 mm in diameter, Tianjin BaseLine ChromTech Research Centre, Tianjin, China) using an N-ethyl-N9-(3-(dimethylamino) propyl) carbodiimide and N-hydroxysuccinimide (EDC/NHS) amidization protocol [fig_ref] Figure 1 |: Schematic for the antibody/MBs preparation and colorimetric immunomagnetic assay [/fig_ref] according to a previous report [bib_ref] SERS-based sandwich immunoassay using antibody coated magnetic nanoparticles for Escherichia coli enumeration, Guven [/bib_ref]. Briefly, MBs-COOH was first washed three times with 2-(N-morpholino) ethanesulfonic acid (MES) buffer (0.05 M, pH 5.5), then mixed with 1 mL of MES buffer (containing 20 mg of EDC and 20 mg of NHS) and activated for 30 min. The MES buffer was later removed by attracting the beads to the wall of the vial with a magnet. The MBs were washed with MES buffer for three times to remove excess EDC and NHS. Subsequently, the resulting functionalized-beads were dispersed in 0.3 mL of PBS and sonicated for 5 min to obtain a homogeneous dispersion. Then 0.3 mL of anti-S. oneidensis antibody at 1.4 mg/mL was added to the dispersion. After mixed thoroughly, the mixture was stirred for 3 h at room temperature. The reaction mixture was placed on a magnet for 5 min, and the pellet was collected after removal of supernatant, and then washed three times with PBS. The resulting antibody/MBs complex was redispersed in 1.0 mL of PBS containing 1% BSA, and stored at 4uC.
Immunomagnetic capture of S. oneidensis. Serial dilutions of bacterial sample were made for IMC. To begin with, 500 mL of diluted sample was added to a 1.5 mL centrifuge tube, and then 40 mL of antibody/MBs was added and mixed with the bacterial sample in the 1.5-mL tube. The tube was rotated for 1 h, allowing the target bacteria to attach to the antibody/MBs. Bacteria/beads complexes were then magnetically separated by placing the tube on a magnet. After removal of supernatant, the collected complexes were washed for three times, resuspended in approximately 100 mL of pure water and transferred to a microplate.
Colorimetric quantification of S. oneidensis. Colorimetric assays were performed on a 96-well plate as illustrated in [fig_ref] Figure 1 |: Schematic for the antibody/MBs preparation and colorimetric immunomagnetic assay [/fig_ref]. In short, 100 mL of bacteria/MBs complex was mixed with equal volume of substrate solution (containing 100 mM citric acid, 200 mM Na 2 HPO 4 , 0.32 mM TMB and 2 mM H 2 O 2 , pH 4.3). The plate was incubated for 30 min at room temperature for color developing. After color development, the reaction was stopped with 2 M sulfuric acid (50 mL/well), and then the solution in the wells was transferred to a new plate with the aid of a magnetic separator. Optical density (OD) was measured at 450 nm using a microplate reader (iMark, Bio-rad).
[fig] Figure 1 |: Schematic for the antibody/MBs preparation and colorimetric immunomagnetic assay. (a) Preparation procedure of antibody/MBs. (b) The procedure of the colorimetric immunomagnetic assay. [/fig]
[fig] Figure 2 |: Optical and SEM images of S. oneidensis/MBs. Optical image (inset) and SEM images show S. oneidensis cells attached to antibody/MBs. Cells appear as faint rods with beads attached. www.nature.com/scientificreports SCIENTIFIC REPORTS | 4 : 5191 | DOI: 10.1038/srep05191 capture efficiency of antibody/MBs, S. oneidensis cells at a concentration of 5.0 3 10 6 CFU/mL were captured with antibody/MBs and then plated on plain LB agar. A capture efficiency of 75.2 6 4.6% for S. oneidensis cells was found. [/fig]
[fig] Figure 3 |: Peroxidase-like activities of S. oneidensis. UV-Vis of color development in (a) S. oneidensis, (b) Cytochrome c, (c) S. oneidensis/MBs, (d) MBs and (e) Pure water. The color developing was performed in substrate solution containing H 2 O 2 and TMB for 30 min. Inset: photographs of the corresponding solutions. www.nature.com/scientificreports SCIENTIFIC REPORTS | 4 : 5191 | DOI: 10.1038/srep05191 [/fig]
[fig] Figure 4 |: Optimization of the working conditions. (a) Antibody/MBs concentration, (b) H 2 O 2 concentration, (c) Immune-capturing time and (d) Color developing time. S. oneidensis cell at 5.0 3 10 6 CFU/mL. [/fig]
[fig] Figure 5 |: Specificity of the colorimetric immunomagnetic assay. Blank control (no bacterial cells), E. coli (5.0 3 10 6 CFU/mL), B. subtilis (5.0 3 10 6 CFU/mL), S. aureus (5.0 3 10 6 CFU/mL), P. agglomerans (5.0 3 10 6 CFU/mL), A. hydrophila (5.0 3 10 6 CFU/mL), S. decolorationis (5.0 3 10 6 CFU/mL), S. putrefaciens (5.0 3 10 6 CFU/mL), S. oneidensis (5.0 3 10 6 CFU/mL). Immunoreaction time, 50 min; antibody/MBs, 40 mL; color developing time, 30 min. Error bars represent S.D., n 5 3. [/fig]
[fig] Figure 6 |: The effects of sample matrix components on the colorimetric immunomagnetic assay. Immunoreaction time, 50 min; antibody/MBs, 40 mL; color developing time, 30 min. Error bars represent S.D., n 5 3. [/fig]
[fig] Figure 7 |: The regression curve for the detection of S. oneidensis cell.Immunoreaction time, 50 min; antibody/MBs, 40 mL; color developing time, 30 min. Error bars represent S.D., n 5 3. [/fig]
|
A Review on the Expression Pattern of Non-coding RNAs in Patients With Schizophrenia: With a Special Focus on Peripheral Blood as a Source of Expression Analysis
## Lncrnas and schizophrenia
LncRNAs are a huge and dissimilar group of ncRNA which have more than 200 nucleotides. LncRNAs signify the bulk of the noncoding transcriptome and utilize numerous mechanisms to exert their regulatory functions on gene expression among them being suppression or recruitment of transcription factors, modulation of chromatin structure and regulation of the stability of transcripts. High throughput microarray investigations have demonstrated significant changes in lncRNA signature in the peripheral blood or post-mortem brain tissues of patients with schizophrenia compared with control samples.
## Lncrna profile in central tissues
Expression of the lncRNA Gomafu has been regulated by neuronal activation. This lncRNA has been shown to directly interact with two splicing factors namely QKI and SRSF1. Therefore, aberrant expression of Gomafu changes the splicing patterns of DISC1 and ERBB4 genes to a pattern which is similar to what is reported in schizophrenia. Expression of Gomafu is substantially decreased in post-mortem cortical gray matter of patients with schizophrenia. Level of Gomafu activity has been correlated with neuronal structural plasticity which is strong in the course of development and subsequently is decreased in the adulthood. Therefore, abnormal activity of Gomafu in the brain tissues of subjects with schizophrenia might reflect abnormal neurodevelopment. Moreover, abnormal levels of this lncRNA might also affect human behavior, since it has been associated with susceptibility to substance abuse.
Hu et al. have assessed RNA profile of post-mortem brain samples in patients with schizophrenia and those with bipolar disorder and control subjects. They reported differential expression of 20 long intergenic non-coding RNAs (lincRNAs) in orbitofrontal cortex of bipolar patients and aberrant expression of 34 and 1 lincRNAs in anterior cingulate cortex and dorsolateral prefrontal cortex of patients with schizophrenia, respectively. Thus, they reported brain area-specific profiles for lincRNAs. Differentially expressed lincRNAs were enriched in pathways namely immune system development and oligodendrocyte differentiation. Moreover, they reported altered DNA methylation as a possible mechanism for dysregulation of lincRNAs.
## Lncrna profile in peripheral blood
Chen et al. have examined expression profile of lncRNAs in the peripheral blood mononuclear cells (PBMCs) of patients with schizophrenia compared with healthy subjects. They reported differential expression of 125 lncRNAs between these two subgroups. Notably, expression levels of ENST00000394742, TCONS_l2_00025502, ENST00000563823, ENST00000521622, and TCONS_l2_00021339 were suggestively down-regulated in patients. LncRNA profiling has also revealed up-regulation of three lncRNAs in schizophrenia, down-regulation of six lncRNAs in major depressive disorder, and up-regulation of three lncRNAs in generalized anxiety disorder (GAD). Notably, lncRNAs observed to be up-regulated in schizophrenia were significantly decreased in patients with GAD. Furthermore, down-regulated lncRNAs in major depressive disorder were up-regulated in patients with schizophrenia. Finally, there were significant differences in the expression levels of lncRNAs between patients with schizophrenia and GAD. Therefore, a number of lncRNAs are putative biomarkers for differentiation of schizophrenia from major depressive disorder and generalized anxiety disorder. Sudhalkar et al. have examined expression levels of MEG3, PINT, and GAS5 in the PBMCs of patients with psychosis compared with healthy controls. They reported diagnostic differences with MEG3, PINT, and GAS5, and symptom acuity effect with MEG3 and GAS5. Moreover, there was significant difference in the expression of MEG3 between drug naïve patients and patients received risperidone. IFNG-AS1 expression has been shown to be down-regulated in patients with schizophrenia compared with healthy subjects in correlation with IFNG expression indicating a putative role for inflammation in this disorder. We have recently demonstrated downregulation of FAS-AS1, PVT1, and TUG1 in patients with schizophrenia compared with controls. Yet, expressions of GAS5, NEAT1, and OIP5-AS1 were similar between patients and controls.
FAS-AS1 50 SCZ patients and 50 healthy controls were enrolled.
## Blood
The association between FAS-AS1 expression and schizophrenia was remarkable in a subgroup of men.
PVT1 PVT1 and TUG1 were appropriate biomarkers in male patients.
## Tug1
Frontiers in Psychiatry | www.frontiersin.org schizophrenia compared with including 62 over-expressed and 63 under-expressed lncRNAs in patients. Antipsychotic treatment has resulted in reduction in NONHSAT089447 and NONHSAT041499 levels, parallel with decrease in the post-treatment Positive And Negative Syndrome Scale (PANSS) scores. Moreover, reduction in NONHSAT041499 levels have been associated with improvement of several clinical manifestations and better response to therapies (17). Chen et al. have reported the effects of olanzapine treatment in suppression of expression of the NONHSAT089447 lncRNA. Small interfering RNA-mediated NONHSAT089447 silencing has reduced expression of dopamine receptors DRD3 and DRD5. In addition, Western blot studies verified the role of this lncRNA in regulation of DRD signaling. We have demonstrated up-regulation of HOXA-AS2, Linc-ROR, MEG3, SPRY4-IT1, and UCA1 in patients with schizophrenia compared with healthy subjects. Yet, when assessing their expressions in sex-based subclasses, the differences in their expressions were significant just among females. Moreover, we reported correlations between expressions of Linc-ROR and SPRY4-IT1 and age of patients. Ni et al. have profiled peripheral blood transcriptome of monozygotic twins discordant for schizophrenia. Using this approach, authors have demonstrated up-regulation of AC006129.1 lncRNA in patients. This lncRNA regulates inflammatory reactions through promoting expression of SOCS3 and CASP1. Further experiments showed that AC006129.1 interacts with the promoter region of the transcriptional repressor Capicua (CIC) to enhance the interactions of DNA methyltransferases with its promoter and decrease CIC expression, thus reversing CIC-associated SOCS3 and CASP1 suppression. Activation of SOCS3 increases the anti-inflammatory reactions by obstructing JAK/STAT pathway.exhibits the list of up-regulated lncRNAs in schizophrenia.
## Association between lncrna(s) snp(s) and risk of schizophrenia
Few studies have appraised association between lncRNAs SNPs and risk of schizophrenia. Rao et al. have performed a twophase association study on 8 tag SNPs that encompass the entire MIAT region in two independent cohorts from Han Chinese population. They demonstrated significant association between paranoid schizophrenia and the rs1894720. Moreover, there was a weak association between rs4274 and this condition. No specific haplotype was detected that modulate risk of paranoid schizophrenia in the assessed population.
## Diagnostic value of lncrnas in schizophrenia
Expression levels of lncRNAs could be used as diagnostic markers for schizophrenia. We have assessed this possibility in a cohort of Iranian patients with schizophrenia. Based on the obtained results in a limited number of patients, GAS5 and OIP5-AS1 have been proposed as appropriate biomarkers in female subjects.summarizes the results of these studies.
## Mirnas and schizophrenia
These small transcripts have intricate temporospatial signature in the brain tissue and can modulate expressions of myriad of genes by working as the specificity elements for genesilencing apparatus in the cells. Based on the observed association between miRNA dysregulation and substantial alterations in the network organization in the course of neurodevelopment, miRNAs are considered as important regulators of several neurological processes. Numerous studies have indicated aberrant expression of miRNAs in the brain and peripheral blood of patients with schizophrenia. Sun et al. have demonstrated over-expression of miR-132, miR-195, miR-30e, and miR-7 in plasma samples of patients with schizophrenia, and up-regulation of miR-212, miR-34a, and miR-30e in their PBMCs {#164}. They suggested that miRNA signature is more distinctive in plasma samples compared with PBMCs. A high throughput miRNA profiling revealed up-regulation of eight miRNAs in plasma samples obtained from schizophrenia patients, among them were miR-130b and miR-193a-3p which were up-regulated in schizophrenia but not in non-schizophrenia psychotic disorders. These results indicated these miRNAs as state-independent biomarkers for schizophrenia.
## Mirna levels in peripheral blood
Lai et al. have subsequently assessed the impact of hospitalization on the expression levels of these miRNAs. Notably, expression of none of these miRNAs did not change after 2 months hospitalization of patients even when clinical symptoms were remarkably ameliorated. Thus, these miRNAs have been suggested as trait biomarkers instead of statedependent biomarkers. Assessment of expression profile of CT value of NONHSAT041499 was significantly higher in patients after the treatment, representing the substantial down-regulation of this lncRNA expression by the treatment. The symptomatology score and total score were meaningfully reduced following treatment.
## Nonhsat098126
These transcripts have been suggested as markers for the diagnosis and prognostic evaluations.
hsa-miR-34a and hsa-miR-548d in post-mortem brain samples showed no difference between patients and controls.
## Mirna profile in central tissues/cell lines
Expression profile of miRNAs has also been assessed in the olfactory epithelium as one of the limited available neural tissues that have neurons and neural stem cells. Mor et al. have detected over-expression of miR-382-5p in cultured olfactory cells obtained from schizophrenia patients compared with controls. Up-regulation of this miRNA was also verified in microdissected olfactory epithelium neuronal tissues of these patients. However, miR-382 was not expressed in lymphoblastoid cell lines originated from either patients with schizophrenia or control subjects. This miRNA has been shown to target FGFR1 and SPRY4, two genes whose expressions were decreased in the olfactory cells obtained from patients with schizophrenia. Other experiments in HEK293 and SH-SY5Y cell lines have validated CALN1 as a target of miR-137. A number of other miRNAs can also regulate development of dendrites, thus being implicated in the functionality of synapses and neuronal interactions. For instance, miR-214 has been shown to increase dendrite dimension, complexity and morphogenesis. Its involvement in the pathogenesis of schizophrenia has been suggested through the observed interaction between this miRNA and quaking (Qki), a candidate gene in schizophrenia. The miR-214-Qki axis is a fundamental axis in the modulation of dendritic development in the neurons. Experiments in a mouse neuroblastoma cell line (Neuro2A) have shown the regulatory role of miR-137 on the expression of several important proteins in the PI3K-Akt-mTOR axis which have functions downstream of neuregulin/ErbB and BDNF. Therefore, this miRNA controls neuronal reactions to these factors and dendritic development, thus contributing in the risk of schizophrenia. shows the mechanism of involvement of miR-214 in the pathogenesis of schizophrenia.show the list of down-/up-regulated miRNAs in the schizophrenia patients, respectively.
## Mirnas variations in schizophrenia
Some studies have verified associations between SNPs within miRNA-coding genes and risk of schizophrenia. For instance, in a genome-wide association study, Ripke et al. have shown significant association between the rs1625579 of the miR-137 and risk of schizophrenia. Notably, others have confirmed associations between miR-137 target genes CUB, CSMD1, C10orf26, CACNA1C, TCF4, and ZNF804A, and risk of schizophrenia. Notably, the mentioned SNP in miR-137 has been shown to influence cognitive activity. The T/T genotype of this SNP is correlated with working memory defects in patients with schizophrenia as reflected by their reduced scores in the brief assessment of cognition in schizophrenia instrument. The risk genotype has also been associated with dorsolateral prefrontal cortex hyperactivation (DLPFC) in patients with schizophrenia. In addition, the risk allele of this SNP has been associated with down-regulation of miR-137 in schizophrenia and is potentially involved in the modulation of expression of the schizophrenia risk locus TCF4, more emphasizing on the participation of miR-137 and its downstream molecules in this disorder. These studies further support the role of this miRNA in the pathogenesis of schizophrenia.
## Diagnostic value of mirnas in schizophrenia
Diagnostic value of miRNAs has been assessed in different biological sources of patients with schizophrenia using statistical methods, such as plotting the receiver operating characteristic curve and calculation of the area under the curve (AUC) values. Such method has shown the diagnostic values of 0.767 and 0.756 for plasma and PBMC expression patterns of miR-30e, respectively. Plasma levels of this miRNA had sensitivity and specificity of 90.90 and 60.00%, respectively. These values for its PBMC levels were 81.80 and 68.00%, FIGURE 1 | miR-214 is overexpressed in patients with schizophrenia. This miRNA binds with the 3 ′ UTR of Qki transcript to reduce its expression. Qki is an RNA-binding protein which modulates stability of several mRNAs among them are AIP1, Wnt2a, and Rab5. These transcripts are involved in the dendritic development. Qki can also bind with FEZ1 mRNA to enhance its stability. FEZ1 binds with DISC1 and increases neurite outgrowth in response to NGF. Qki can bind with intronic regions and participate in the biogenesis of circRNAs. Abnormal expression of Qki might be associated with the observed dysregulation of cricRNAs in schizophrenia.
respectively. Moreover, logistic regression analysis showed higher sensitivity of plasma levels of this miRNA in differentiating schizophrenia patients from healthy subjects compared with its PBMC levels. The diagnostic value of a panel of over-expressed miRNAs including miR-30e, miR-181b, miR-34a, miR-346, and miR-7 has been reported to be 0.713. Notably, suitable pharmacotherapy resulted in down-regulation of miR-132, miR-181b, miR-432, and miR-30e expressions. Besides, there was a remarkable correlation between amelioration of clinical symptoms and alterations in the expression levels of miR-132, miR-181b, miR-212, and miR-30e. A certain molecular axis including the transcription factor the early growth response protein 1 (EGR1), miR-30a-5p, and its target gene NEUROD1 has been shown to differentiate schizophrenia patients from healthy subjects with diagnostic accuracy of 0.962 which was far higher than the diagnostic power of miR-30a-5p alone. This axis has also been proved useful for monitoring of patients in acute psychotic phase.shows the results of studies which assessed diagnostic role of miRNAs in the schizophrenia.
# Discussion
Schizophrenia is complex disorder caused by interaction between several genomic loci. The speculation that this disorder is caused by one or a few common principal gene effects has been experimentally examined in genome-wide linkage studies yet results generally showed no genome-wide significance. In addition to genomic variants that contribute in the pathogenesis of this disorder, some other mechanisms might modify the risk of development of schizophrenia. Expressions of ncRNAs are modulated by neuronal activation, suggesting a role for these transcripts in the pathophysiology of neuropsychiatric disorders. In the current study, we reviewed the literature about the role of ncRNAs in the pathophysiology of schizophrenia. Based on the above-mentioned evidence, several lncRNAs and miRNAs have been dysregulated in blood or brain samples of patients with schizophrenia. Not discounting the role of ncRNAs as biomarkers for schizophrenia, peripheral expression profile of these transcripts does not necessarily reflect their expression in the brain tissues. Aberrant expression of a number of Prefrontal cortex these transcripts has been associated with PANSS score. Moreover, clinical diagnoses of psychosis and symptom severity have been shown to alter expression of a number of lncRNAsindicating a substantial role for these transcripts in the pathogenesis of schizophrenia. Although there is a considerable level of overlap between psychiatric disorders in the terms of contributing genetic factors, expression levels of some ncRNAs could be used to differentiate a number of these conditions. Moreover, expression profile of certain ncRNAs can distinguish patients with schizophrenia from healthy subjects. At the present, schizophrenia is principally diagnosed based on the clinical symptoms and signs instead of on the pathophysiological biomarkers. Identification of such biomarkers would facilitate detection of malingering, thus has practical significance in the forensic medicine in the assessment of cases pretending psychological disorders for a particular gain. Moreover, molecular biomarkers have implications in the establishment of targeted therapies. Notably, expression profile of ncRNAs in the circulation has the potential to improve the diagnostic and prognostic assessment of patients with schizophrenia. A former study has demonstrated up-regulation of NONHSAT089447, NONHSAT021545, and NONHSAT041499 lncRNAs in patients with schizophrenia, while down-regulation of these transcripts in patients with GAD. Notably, authors have reported six lncRNAs with opposite expression patterns in schizophrenia and major depressive disorder. Moreover, they have identified three GADrelated lncRNAs whose expressions were significantly differences between patients with schizophrenia and GAD patients. However, most of these studies do not test if expression profile of ncRNAs can distinguish people with schizophrenia from people with other disorders, such as major depression, bipolar disorder, or autism. Based on the similarities in many of clinical symptoms between subjects with schizophrenia and other disorders, including bipolar disorder, autism, and major depression, identification of specific markers for each disease has practical significance. Therefore, this field should be explored in future investigations. Particularly, miRNA signature has been correlated with clinical course, patients' response to pharmacologic interventions and prognosis of patients with schizophrenia. Perhaps, the most intensively assessed miRNA in this regard is miR-30 family. Consistent with this observation, expression of EGR1 which regulates expressions of this miRNA family has been decreased in PBMCs of patients with schizophrenia. On the other hand, expression of NEUROD1 as a target gene of miR-30a-5p has been increased in these patients. Based on the results of this study, identification of transcription factor/ miRNA/ target gene axes would be a practical method for recognition of molecular pathways in the pathogenesis of schizophrenia and development of diagnostic/ prognostic panels for this disorder.
[formula] STRADB & IRGQ & PWWP2A & CADM2 & SH3TC2 & DCUN1D3 & CDC73 & BRWD3 & COL5A3 & COL5A3 & CDK19 & RIMKLB & CALD1 & CDC73 & SESN3 & [/formula]
Several schizophrenia-associated ncRNAs have been shown to modulate immune responses and inflammatory pathways. Recent studies have highlighted the presence of intricate interplay between the immune system, systemic inflammatory responses, and the central nervous system, which can result alterations in mood, cognitive functions, and behavior. All of these aspects contribute in the pathogenesis of schizophrenia. Immune responses can influence activity of neurotransmitters as well as neurodegenerative and neurodevelopmental processes all of which are related with this disorder. Assessment of lncRNA signature in Amygdala samples from schizophrenia patients has further endorsed dysregulation of immune-associated lncRNAs in these samples. Thus, the functional axes between ncRNAs and immune-related genes provide an explanation for involvement of these transcripts in the pathophysiology of schizophrenia.
Besides, high throughput studies in schizophrenia subjects have shown a tendency toward global up-regulation of miRNAs normally abundant in infants, whereas down-regulation of those normally abundant in prepuberty. Therefore, dysregulation of miRNAs dynamic changes might be involved in the pathogenesis of schizophrenia. A number of ncRNAs might explain the observed dissimilarity in the brain activation modes between patients with schizophrenia and controls. For instance, Gomafu has been shown to be intensely regulated in reaction to neuronal activation. This lncRNA is also implicated in schizophreniarelated alternative splicing. Therefore, aberrant expression of this lncRNA might reflect the abnormal brain activity in these patients.
Finally, it is worth mentioning that miRNAs have distinctive roles as trait-dependent markers or state-dependent markers. This speculation is supported by amelioration of dysregulated expression of a number of miRNAs after pharmacotherapy, while no change in the expression of other miRNAs following suitable treatments. Therefore, it is necessary to define the role of each miRNA as trait-or state-dependent marker to design distinctive diagnostic/ prognostic miRNA panels for schizophrenia.
Taken together, lncRNAs and miRNAs are potential transcripts that can explain the difference in expression profile of protein coding genes in brain and blood tissues of patients with schizophrenia and healthy subjects. Moreover, as a number of above-mentioned ncRNAs are functionally related with dopamine neurotransmission, these ncRNAs might alter response of patients to some types of antipsychotic drugs. In addition, several ncRNAs can be used as disease markers in schizophrenia.
Several questions should be addressed about the role of ncRNAs in the development of schizophrenia. Studies reviewed in this article have mostly assessed expression of ncRNAs in adult patients. Although differentially expressed ncRNAs among patients and controls have been enriched in neurodevelopment, neurotransmission, and synaptic plasticity, the functional impact of these ncRNAs in the development of neurons should be assessed through knock-out/-in studies in animal models and cell lines.
Finally, most of studies reviewed here have not appraised the effects of antipsychotic drugs on expression of ncRNAs. The data regarding the therapeutic regimens and types of antipsychotic drugs have not been presented in the main articles. This is possibly because patients have been under treatment with different antipsychotic drugs. We mention this point as a limitation of these studies.
# 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.
|
Snack food consumption among Bangladeshi children, supplementary data from a large RCT
Childhood obesity has been associated with consumption of energy-dense foods such as caloric beverages and fast foods. Many low-and middle-income countries like Bangladesh are now experiencing a rising problem of noncommunicable diseases along with the long-standing problem of stunting and undernutrition. WASH Benefits Bangladesh was a large community-based cluster randomized controlled trial conducted in rural Bangladesh. Study clusters were randomized into seven arms: single nutrition (N); water (W); sanitation (S); hygiene (H); combined water, sanitation, and hygiene (WSH); WSH and nutrition (N + WSH); and a double sized control (C).Nutrition intervention messages included four promotional components: maternal nutrition, breastfeeding, complementary feeding, and lipid-based nutrient supplements. The World Health Organization infant food frequency questionnaire (24-hr recall and 7-day recall) was administered at Year 1 and Year 2 of intervention.The likelihood of any snack food consumption was significantly lower (odds ratio 0.37: 95% confidence interval [0.28, 0.49]) in the nutrition intervention arms compared to the control arm in Year 2 follow-up. In addition, in the water intervention arm, fewer children (about 50% less) consumed soft drinks, but not the other sugar-sweetened beverages, compared with control in Year 2. There were no other differences between groups. Simple messages about balanced diet and feeding family foods were effective in lowering commercially produced snack food consumption of the young children in low-income rural communities of Bangladesh. Provision of safe water apparently encouraged mothers to reduce offering unhealthy beverages to the young children.K E Y W O R D Sdietary patterns, infant and child nutrition, infant feeding behaviour, infant feeding decisions, randomized controlled trial, water
# | introduction
A substantial worldwide increase in childhood obesity has posed a major global health challenge. The Global Burden of Disease Study 2013 identified that overweight and obesity among children have increased markedly in high-, middle-, and low-income countries [bib_ref] Global, regional, and national prevalence of overweight and obesity in children and..., Ng [/bib_ref]. Although the level of undernutrition continues to be high, some low-and middle-income countries have experienced similar or greater increases in childhood obesity compared with highincome countries [bib_ref] Child and adolescent obesity: Part of a bigger picture, Lobstein [/bib_ref]. All dimensions of health, physical, emotional, and social, are profoundly influenced by childhood obesity [bib_ref] Childhood obesity: Causes and consequences, Sahoo [/bib_ref]. The World Health Assembly set a global nutrition target to halt the rise in childhood overweight by .
Childhood obesity has been associated with consumption of energy-dense foods like caloric beverages and fast foods [bib_ref] Poverty and obesity: the role of energy density and energy costs, Drewnowski [/bib_ref]. A study with more than 4,000 USchildren aged 5 and 11 years showed that there was an increased likelihood of developing overweight/obesity with increased frequency of fast food and sugar-sweetened drink consumption [bib_ref] Double trouble: Commingled effects of fast food and sugar-sweetened beverage consumption and..., Berry [/bib_ref]. A longitudinal study on children of low socioeconomic status from Brazil also found that early ultraprocessed food consumption as a proportion of daily energy was associated with altered lipoprotein profile; ultraprocessed foods are not made from whole foods but from their derivatives and food additives [bib_ref] The food system. Ultra-processing: the big issue for nutrition, disease, health, well-being, Monteiro [/bib_ref].
Snacking habits such as consumption of sweet snacks, candy, and chips were associated with other adverse health outcomes including dental caries in young children [bib_ref] Snacking habits and caries in young children, Johansson [/bib_ref]. Energy-dense, nutrient-poor foods have become more accessible to people at a lower cost due to developments in agriculture and food technologies [bib_ref] Poverty and obesity: the role of energy density and energy costs, Drewnowski [/bib_ref].
South Asian countries, including Bangladesh, have shown remarkable economic advancement in recent years. Improvement in household economic status, local production of packaged food, improved transportation, and extensive food marketing have bought changes in dietary patterns even to rural populations. Up to 75% of 2-year-old Asian children from countries like Philippines and Nepal were found to consume sugary snack foods [bib_ref] Babies, soft drinks and snacks: A concern in low-and middle-income countries?, Huffman [/bib_ref]. A study in rural Indonesia showed that snack food that included fast food, soft drinks, traditional snack foods, candies, and desserts, and modern snack foods consumption, in respect to total energy and nutrient intake, was associated with lower height-for-age z-score among schoolchildren [bib_ref] Snack foods consumption contributes to poor nutrition of rural children in West..., Sekiyama [/bib_ref]. Countries like Bangladesh and India are now detecting increasing rates of noncommunicable diseases along with the long-standing problem of stunting and undernutrition. With a national stunting prevalence of 36%, Bangladesh is simultaneously experiencing an increase in the prevalence of overweight (3.6 to 7.9% from 1998 to 2015) and obesity (2 to 9% from 2004 to 2015) over time [bib_ref] Overweight and obesity among children and adolescents in Bangladesh: A systematic review..., Biswas [/bib_ref].
Young children's eating behaviour is presumed to be greatly influenced by their parents [bib_ref] Parental feeding and child eating: An investigation of reciprocal effects, Steinsbekk [/bib_ref].
A parenting program in Bangladesh to address early childhood health, growth, and development found the intervention was effective in changing mother's practices related to dietary diversity [bib_ref] Effectiveness of a parenting program in Bangladesh to address early childhood health,..., Aboud [/bib_ref]. A recent systematic review on the factors determining eating behaviour among preschool children in low-and middle-income countries found that better nutritional knowledge of caregivers was associated with an increased healthy eating practices of the children [bib_ref] Family and community factors shaping the eating behaviour of preschool-aged children in..., Sirasa [/bib_ref]. Therefore, an intervention focused on improved caregiver knowledge of appropriate complementary feeding practice may result in improved dietary quality and less unhealthy snack food consumption in their children.
The primary objective of this study was to measure commercially available snack food and packaged food consumption patterns among children less than 3 years of age participating since birth in a randomized controlled trial of nutrition, water, sanitation, and hygiene interventions. We previously reported a significant increase in dietary diversity among the children in the nutrition intervention group compared with control . Thus, we were interested to explore if there were any relationships between snack food consumption and complementary food diversity among the children who took part in the intervention trial. We hypothesized that a nutrition inter- These sites were selected according to the quality of ground water
## Key messages
- The promotion of optimal infant and young child feeding practices was associated with a passive reduction of commercially produced snack food consumption.
- Provision of safe water appeared to encourage mothers not to offer unhealthy beverages to young children.
- Snack food consumption was not associated with displacement of nutritious complementary foods; however, more research is needed using quantitative dietary intake assessment methods. and presence or absence of ongoing or upcoming water, sanitation, and nutrition interventions. At the initial stage of the study, research assistants conducted a community census to identify pregnant women. Women at their first or second trimester of pregnancy were eligible to enrol in the study. Eight nearest pregnant women, identified using a global positioning system, formed a study cluster. These study clusters were randomized into seven arms: single nutrition (N); water (W); sanitation (S); hygiene (H); combined water, sanitation, and hygiene (WSH); combined WSH and nutrition (N + WSH); and a double sized control (C). Each intervention arm contained 90 clusters, and the control arm contained 180 clusters. A total of 5,551 women and their offspring were enrolled in the study. The randomized group assignment could not be blinded due to the nature of the interventions. Detailed study design and rationale have been published previously [bib_ref] Cluster-randomised controlled trials of individual and combined water, sanitation, hygiene and nutritional..., Arnold [/bib_ref] [bib_ref] Complementary feeding practices among rural Bangladeshi mothers: Results from WASH Benefits study, Jannat [/bib_ref]. [bib_ref] Combining intensive counseling by frontline workers with a nationwide mass media campaign..., Menon [/bib_ref] and delivered small quantity lipid-based nutrient supplements to children from 6 to 24 months of age. To deliver the intervention to the households, especially to the mothers, women from the local community were recruited as community health promoters (CHPs). CHPs were under regular supervision of the research assistants. Supervision of the research staffs were arranged in a ranked order so that the implementation remained even and precise. Supervisors were trained in direct intervention delivery; subsequently, they trained the CHPs. First round of training was more intensive including basic project description, principals of behaviour change, communication skills, intervention specific training, and role play. This was followed by regular monthly meeting between supervisors and CHPs. Refresher training was also arranged about a year after initiation of intervention .
## | intervention delivery
Each cluster of 6-8 months was monitored by one CHP. On average, a cluster was about 1 km in diameter that could easily be covered by a CHP. We maintained a 1 km of buffer zone in between the study clusters to prevent possible spillover effects. to the households. Lipid-based nutrient supplements were distributed to the mothers on a monthly basis. Two 10-g sachets were recommended for daily consumption starting from 6 months until 2 years of age.
## | data collection
During the baseline assessment, before the index children were born, research assistants visited participants in their home to collect household socio-economic and demographic data. The research assistants conducted two follow-up visits at about 1 year and 2 years of commencement of intervention delivery. The infant food frequency questionnaire (24-hr recall and 7-days recall) was completed at both follow-up visits. The infant and young child feeding module was adapted from indicators recommended by the World Health Organization (WHO, 2010a) with an added module to assess consumption of commercially produced snack and packaged foods. Commercially produced snacks were given special consideration due to their high salt, sugar, and trans-fat content, and also taking into account the presence of food substances that are not commonly used in culinary preparations such as hydrogenated oil and food additives like colouring agents, flavourings, and nonsugar sweeteners [bib_ref] Industrialized foods in early infancy: a growing need of nutritional research, Araújo [/bib_ref]. We grouped different commonly available snack foods and packaged foods in the rural market according to their taste and pattern. The questionnaire module focused on store-bought sweet and savoury snacks rather than those that were home prepared. Store-bought foods may be ultraprocessed, contain high amounts of sugar, high sodium, or are deep fried with frying oil used and reheated several times, which increases the trans-fat content of the food [bib_ref] Effect of heating/reheating of fats/oils, as used by Asian Indians, on trans..., Bhardwaj [/bib_ref]. We grouped them into six categories: If any item from the mentioned food categories was consumed, mothers were asked for how many days the particular food item was consumed in past 7 days.
# | data analysis
The outcomes presented in this paper were added during the course of the study and were not considered primary or secondary outcomes of the trial. Those results have been published elsewhere [bib_ref] Effect of water quality, sanitation, hand washing, and nutritional interventions on child..., Tofail [/bib_ref]. Nevertheless, the statistical analyses were planned and pre-specified prior to beginning the analysis. The WHO has defined snack foods for young children as foods that are easy to prepare, mostly self-fed, and eaten in between main meals . In this paper, we defined snack food as unhealthy when they contained excess sugar (these included as soft drinks, artificial fruit juice, sweetened milk, and sweet snacks), high salt ( groups among seven recommended food groups all days within past 7 days. We used a multivariate logistic regression model to measure if there was any association between snack food consumption and minimum dietary diversity in the past 7 days. We added the covariates: child age and sex, parent's education, father's occupation, study districts, household socio-economic status, and food insecurity in the model. This model was used to assess the association among children from the nutrition or nonnutrition arms and controls separately. We examined that the association between snack food consumption and dietary diversity in the nutrition arm and other arms as dietary diversity was similar among children from control and nonnutrition arms . Food insecurity was measured using WHO Household Food Insecurity Access Scale. Because participating children were randomized at cluster level, it was expected that responses were correlated within cluster. In order to obtain cluster-robust standard errors, all analyses were adjusted for clustering using the sandwich estimator. Baseline characteristics, child age, and sex were balanced across study arms . About half of the mothers had an educational level of more than grade five. Electricity connection was available in around 60% of the households. The household size was around five persons (2 SD) per family. Food insecurity was reported by only around a quarter (30%) of the households.
# | ethical considerations
The prevalence of snack food consumption in past 7 days in different arms was 0-7% at the Year 1 follow-up. This prevalence increased to 3-76% in Year 2 follow-up . The consumption of any snack food at least once in the past 7 days was significantly lower (N: 73%; N + WSH: 69% vs. C: 87%) in the nutrition intervention arms compared with the control arm at Year 2 follow-up .
Some of the snack foods consumption was associated with increased dietary diversity . Specifically, among the control and nonnutrition arms, canned fruit juices consumption was associated with a higher odds of dietary diversity at Year 2 . In the nutrition arms at Year 2 follow-up, consumption of soft drinks, sweet snacks, and pickles was associated with an increased consumption of diverse complementary foods . In addition, when measuring association between snack food consumption and dietary diversity, higher educational attainment of fathers was consistently associated with an increased food diversity.
# | discussion
In this randomized controlled trial of nutrition, water, sanitation, and hygiene, children participating in the nutrition intervention were less likely to consume sweets or packaged snack food items. We previously reported that the nutrition intervention, which focused on the timely introduction of diverse complementary foods, resulted in a higher dietary diversity score at both time points and high adherence to lipid-based nutrient supplementation recommendations . The present data suggest that parents prioritized feeding the promoted nutritious foods or lipid-based nutrient supplements over commercially available sweet, salty, or packaged snack foods to their children. Parents have more control over children's food environment and F I G U R E 1 Summary of participant enrolment, randomization, retention, and analysis for snack food consumption experiences at an early age [bib_ref] Parental influence on children's early eating environments and obesity risk: Implications for..., Anzman [/bib_ref]. Similar to the WASH Benefits intervention, studies that delivered interventions targeting the parents were found to be effective in improving children's dietary behaviour [bib_ref] Effectiveness of universal parental support interventions addressing children's dietary habits, physical activity..., Kader [/bib_ref]. The The significant reduction in soft drinks in the water intervention arm is also notable. It is possible that when households felt that their own water was safer, they were less likely to feed their children with packaged drinks. Studies in the United States have also found that perceived water safety was associated with lower intake of sugarsweetened beverages [bib_ref] Perceptions of Tap Water and School Water Fountains and Association with Intake..., Onufrak [/bib_ref] Abbreviations: CI, confidence interval; OR, odds ratio; SD, standard deviation; WSH, water, sanitation and hygiene; WSHN, WSH and nutrition. a None of the indicators were significantly different compared to the control arm. b Food insecurity was measured using World Health Organization Household Food Insecurity Access Scale.
## T a b l e 2
Effect of the nutrition intervention on snack and packaged food consumption (at least once in last 7 days) [bib_ref] Frequent consumption of sugar-sweetened beverages and sweets starts at early age, Laitala [/bib_ref]. Consumption of energy dense foods and sugar-sweetened drinks may result in an increased body mass index in children and adults [bib_ref] Sugar-sweetened carbonated beverage consumption correlates with BMI, waist circumference, and poor dietary..., Collison [/bib_ref] [bib_ref] Dietary energy density and body weight in adults and children: A systematic..., Pérez-Escamilla [/bib_ref]. To regulate this practice of increased consumption, positive food parenting has an important role [bib_ref] How parental dietary behavior and food parenting practices affect children's dietary behavior...., Larsen [/bib_ref]. It is thus important to educate caregivers to support building healthy food habits among children.
We were interested to know how snack food consumption was associated with dietary diversity. We presumed that children who consumed more snack foods would have less diverse complementary The WASH Benefits Bangladesh trial was a large randomized control trial, which was designed to measure small differences in outcome variables, maintained high levels of implementation fidelity, and achieved high follow-up rates throughout the 2-year trial . Promoted behaviours showed a high uptake , including an increase in dietary diversity in the nutrition intervention arm . There were some limitations to this study that should be noted, however. Both complementary feeding and snack food consumption information were reported by the mothers. All food consumption was measured by frequency of intake not by quantity; thus, precise estimation of calories consumed was not possible. Therefore, we could not calculate snack food consumption as a percent of energy or directly estimate the possibility of displacement of healthy foods. Moreover, there might be inconsistencies in food categories with respect to their nutrient density. In the future, a nutrient profiling method would allow for a more standardized method for categorization of food types [bib_ref] A proposed nutrient density score that includes food groups and nutrients to..., Drewnowski [/bib_ref]. Because the intervention was not blinded to either the participants or the data collectors, the chance of information bias cannot be ruled out. However, we used standardized dietary indicators to assess food diversity (WHO, 2010a). We adapted a new module for assessing commercially produced snack food consumption that covered most of the common snack items available in the rural community of Bangladesh. Rigorous training was conducted for the data collectors. The survey team worked independent of the intervention team to minimize surveyor's bias. Loss to follow-up was reasonable (~15%) over a 2-year period.
Simple messages about balanced diet and feeding family food were effective in lowering commercially produced snack food consumption of the young children in low-income rural communities of Bangladesh. Provision of safe water appeared to encourage mothers to reduce offering unhealthy soft drinks to the children. Nevertheless, a sharp increase in snack food consumption in the second
year of child age suggests that greater attention is needed to prevent the establishment of unhealthy eating behaviours in children.
# Acknowledgments
## Conflicts of interest
The authors declare that they have no conflicts of interest.
# Author contributions
KJ drafted the manuscript under the guidance of CPS and input from all listed co-authors. SPL drafted the research protocol; he coordinated input from the study team throughout the project. Peter J. Winch https://orcid.org/0000-0001-8569-5507
Christine P. Stewart https://orcid.org/0000-0003-4575-8571
|
Exosomes in cancer: small particle, big player
Exosomes have emerged as a novel mode of intercellular communication. Exosomes can shuttle bioactive molecules including proteins, DNA, mRNA, as well as non-coding RNAs from one cell to another, leading to the exchange of genetic information and reprogramming of the recipient cells. Increasing evidence suggests that tumor cells release excessive amount of exosomes, which may influence tumor initiation, growth, progression, metastasis, and drug resistance. In addition, exosomes transfer message from tumor cells to immune cells and stromal cells, contributing to the escape from immune surveillance and the formation of tumor niche. In this review, we highlight the recent advances in the biology of exosomes as cancer communicasomes. We review the multifaceted roles of exosomes, the small secreted particles, in communicating with other cells within tumor microenvironment. Given that exosomes are cell type specific, stable, and accessible from body fluids, exosomes may provide promising biomarkers for cancer diagnosis and represent new targets for cancer therapy.
# Introduction
Exosomes are small, lipid bilayer membrane vesicles of endocytic origin. Exosomes can be defined by several common characteristics, including size (50-100 nm in diameter), density (1.13-1.19 g/ml), morphology ("cup" or "dish" shaped in transmission electron microscopy), and certain enriched protein markers (tetraspanins, TSG101, Hsp70). Initially discovered as the garbage bags for removal of unwanted material from cells, the role of exosomes in immune response is gradually recognized as they function in antigen presentation. More recently, the researchers reveal that exosomes contain proteins and nucleic acids that are functional when transferred into recipient cells. Exosomes have been shown to act as shuttles between cells by transmitting signals (referred to as communicasomes). In this review, we highlight the recent advances in the roles of exosomes in cancer with an emphasis on the potential of exosomes as diagnosis biomarker and therapy target.
## Biogenesis, release, and uptake of exosomes
Exosome formation is a fine-tuned process which includes four stages: initiation, endocytosis, multivesicular bodies (MVBs) formation, and exosome secretion [bib_ref] Exosomes: composition, biogenesis and function, Thery [/bib_ref]. Multivesicular bodies (MVBs) are endocytic structures formed by the budding of an endosomal membrane into the lumen of the compartment. After vesicular accumulation, the MVBs are either sorted for cargo degradation in the lysosome or released into the extracellular space as exosomes by fusing with the plasma membrane . The mechanisms underlying the sorting of cargo into the intraluminal vesicles are not yet fully elucidated. Both endosomal sorting complex required for transport (ESCRT)-dependent and independent signals have been suggested to determine the sorting of exosomes [bib_ref] Ceramide triggers budding of exosome vesicles into multivesicular endosomes, Trajkovic [/bib_ref]. The formation of exosomes has been shown to be controlled by the syndecan heparan sulfate proteoglycans and their cytoplasmic adaptor syntenin [bib_ref] Syndecan-syntenin-ALIX regulates the biogenesis of exosomes, Baietti [/bib_ref].
The Rab guanosine triphosphatases (GTPases) have been found to critically regulate exosome secretion. Ostrowski et al. have identified that Rab27a/b affects the size and localization of MVBs [bib_ref] Rab27a and Rab27b control different steps of the exosome secretion pathway, Ostrowski [/bib_ref]. Hsu et al. suggest that Rab3 regulates MVBs docking to tethering at the plasma membrane [bib_ref] Regulation of exosome secretion by Rab35 and its GTPaseactivating proteins TBC1D10A-C, Hsu [/bib_ref]. The accumulation of intracellular Ca 2+ results in increased exosome secretion [bib_ref] Exosome release is regulated by a calcium-dependent mechanism in K562 cells, Savina [/bib_ref]. In addition, intracellular and intercellular pH has been shown to affect exosome release. When the microenvironmental pH is low, exosome secretion and uptake by recipient cells increases [bib_ref] Microenvironmental pH is a key factor for exosome traffic in tumor cells, Parolini [/bib_ref]. There is evidence that oncogenes and tumor suppressors regulate exosome secretion in cancer [bib_ref] Oncogenic events regulate tissue factor expression in colorectal cancer cells: implications for..., Yu [/bib_ref]. demonstrate that p53-regulated protein tumor suppressor-activated pathway 6 (TSAP6) induces exosome secretion under stressed conditions [bib_ref] The regulation of exosome secretion: a novel function of the p53 protein, Yu [/bib_ref] [bib_ref] Exosome secretion, including the DNA damage-induced p53-dependent secretory pathway, is severely compromised..., Lespagnol [/bib_ref]. Heparanase is an enzyme with elevated level in cancer. Overexpression of heparanase promotes exosome secretion [bib_ref] Heparanase regulates secretion, composition, and function of tumor cell-derived exosomes, Thompson [/bib_ref]. Intriguingly, exosomes from normal mammary epithelial cells inhibit exosome secretion by breast cancer cells, implicating a feedback control to maintain dynamic equilibrium [bib_ref] Regulation of exosome release from mammary epithelial and breast cancer cells -a..., Riches [/bib_ref].
Exosomes transfer information to the target cells through three main ways: (1) receptor-ligand interaction;
(2) direct fusion with plasma membrane; (3) endocytosis by phagocytosis . Although the specific receptors that mediate the uptake of exosomes have not been found, there are several proteins that may act as potential receptors for exosome uptake, such as Tim1/4 for B cells [bib_ref] Identification of Tim4 as a phosphatidylserine receptor, Miyanishi [/bib_ref] and ICAM-1 for APCs [bib_ref] ICAM-1 on exosomes from mature dendritic cells is critical for efficient naive..., Segura [/bib_ref]. The uptake of exosomes by direct plasma membrane fusion mode has not been well studied. Melanoma cells could take up exosomes by fusion and low pH facilitates this process [bib_ref] Visualizing of the cellular uptake and intracellular trafficking of exosomes by live-cell..., Tian [/bib_ref]. Phagocytosis is an efficient way of exosome uptake. Phagocytic cells have a greater uptake of exosomes than non-phagocytic cells [bib_ref] Cellular internalization of exosomes occurs through phagocytosis, Feng [/bib_ref]. The uptake of exosomes by recipient cells is energy dependent [bib_ref] Interaction and uptake of exosomes by ovarian cancer cells, Escrevente [/bib_ref]. Heparan sulfate proteoglycans (HSPGs) function as internalizing receptors of cancer cell-derived exosomes. Enzymatic depletion of cell-surface HSPG or pharmacological inhibition of endogenous proteoglycan biosynthesis significantly attenuates exosome uptake [bib_ref] Cancer cell exosomes depend on cell-surface heparan sulfate proteoglycans for their internalization..., Christianson [/bib_ref].
## Structure and contents of exosomes
Exosomes consist of a lipid bilayer membrane surrounding a small cytosol . The structured lipids not only mold the exosomes but are also involved in exosome function. In addition to lipids, nucleic acids and proteins have also been detected in exosomes. Thakur et al. demonstrate that double-stranded DNA is present in exosomes from cancer cells and reflects the mutational status of the originated cells [bib_ref] Double-stranded DNA in exosomes: a novel biomarker in cancer detection, Thakur [/bib_ref]. Valadi et al. demonstrate that exosomes contain mRNA and miRNA [bib_ref] Exosomemediated transfer of mRNAs and microRNAs is a novel mechanism of genetic..., Valadi [/bib_ref]. Exosome-carried RNA can shuttle between cells and thus is called "exosomal shuttle RNA" (esRNA). The protein composition of tumor cell-derived exosomes has been well characterized for a number of cancers by using different proteomic methods. The most common proteins, mRNA, and miRNAs found in exosomes have been deposited in ExoCarta (www.exocarta.org). To date, 4563 proteins, 1639 mRNAs, and 764 miRNAs have been identified in exosomes from different species and tissues by independent examinations. The exosomal contents vary between different physiological and pathological conditions and original cell types. Moreover, the composition of Biogenesis, release, structure, and uptake of exosomes. Exosomes are produced from the multivesicular bodies (MVBs) (also known as late endosomes). The membrane of the MVBs bulges inward to form exosomes. During this process, proteins (e.g., receptor, cytoplasmic proteins, tetraspanin), nucleic acids (e.g., DNA, mRNA, miRNA), and lipids (e.g., cholesterol, ceramide) are packed into exosomes in a cell type-dependent manner. MVBs fuse with the cellular membrane to release exosomes into the extracellular space. Several mechanisms have been suggested to mediate the uptake of exosomes, including a exosome fusion with the cellular membrane of the recipient cell, leading to the release of the exosomal cargo into the cytoplasm, b juxtracrine signaling through receptor-ligand interactions, c and endocytosis by phagocytosis exosomes can be distinct from the originated cells due to the selective sorting of the cargo into exosomes.
## Isolation, detection, and analysis of exosomes
Exosomes have been isolated and characterized from distinct cells under normal and stressed conditions. At present, the most commonly used methods for exosome isolation include ultracentrifugation, combined with sucrose gradient, and the immune-bead isolation (e.g., magnetic activated cell sorting; MACS). There are many commercial kits available for the extraction of exosomes. Transmission electron microscopy (TEM), Western blot, and FACS are frequently used to characterize the isolated exosomes based on their biochemical properties (e.g., morphology, size, exosomal markers). There is a lack of the accurate method to determine the concentration of exosomes. The researchers have to rely on inaccurate measurements of protein concentration or nanoparticle tracking analysis. Quantitative RT-PCR, nucleic acid sequencing, Western blot, or ELISA are used for exosome RNA and protein identification. The International Society for Extracellular Vesicles (ISEV) has recently released minimal experimental requirements for definition of extracellular vesicles and their functions [bib_ref] Minimal experimental requirements for definition of extracellular vesicles and their functions: a..., Lotvall [/bib_ref].
## Roles of exosomes in cancer
Accumulating evidence indicates that exosomes play important roles in cancer. Exosomes transfer oncogenic proteins and nucleic acids to modulate the activity of recipient cells and play decisive roles in tumorigenesis, growth, progression, metastasis, and drug resistance [fig_ref] Figure 2: Roles of exosomes in cancer [/fig_ref]. Exosomes can act on various recipient cells. The uptake of exosomes may induce a persistent and efficient modulation of recipient cells. In this section, we will discuss about the roles of exosomes in cancer and the molecular mechanisms [fig_ref] Table 1: Overview on the function of exosomes in cancer [/fig_ref].
## Tumorigenesis
Normal cells are transformed into cancer cells in the process of tumorigenesis. Exosomes from malignant cells have shown the potential to induce normal cell transformation. For instance, prostate cancer cell-derived exosomes could induce neoplastic transformation of adipose-derived stem cells (ASCs) [bib_ref] Neoplastic reprogramming of patient-derived adipose stem cells by prostate cancer cell-associated exosomes, Elmageed [/bib_ref] , which is associated with trafficking of oncogenic proteins (Ras superfamily of GTPases), mRNA (K-ras and H-ras), as well as miRNAs (miR-125b, miR-130b, and miR-155) by exosomes. In addition, Melo et al. suggest that breast cancer cell-derived exosomes contain precursor microRNAs (pre-miRNAs) associated with RNA-induced silencing complex (RISC)-loading complex proteins, which could induce a rapid and efficient silencing of mRNAs in nontumorigenic epithelial cells, resulting in transcriptome reprogramming and oncogenic transformation [bib_ref] Cancer exosomes perform cell-independent microRNA biogenesis and promote tumorigenesis, Melo [/bib_ref]. They further demonstrate that the exosomes from serum specimen from breast cancer patients but not those from healthy donors induce tumor formation in mice when coinjected with the nontumorigenic epithelial cells, suggesting a potential mechanism for exosome in tumorigenesis. Cancer is composed of heterogeneous cell populations. Side population (SP) cells are a sub-population of cells that exhibit stem cell-like characteristics and can be isolated in cancer by adapting the Hoechst33342 staining method. Koch et al. demonstrate that in diffuse large B-cell lymphoma, side population cells could export Wnt3a via exosomes to neighboring cells, thus modulating SP-non-SP transitions and maintaining population equilibrium [bib_ref] Populational equilibrium through exosome-mediated Wnt signaling in tumor progression of diffuse large..., Koch [/bib_ref]. Altogether, these findings indicate that exosomes may contribute to tumor development and uncontrolled tumor progression by acting as a mediator in the transformation of normal cells to malignant cells and a modulator for the balance between cancer stem cells (CSCs) and non-CSCs.
## Tumor growth
The promoting effects of exosomes from distinct sources on tumor cell proliferation have been widely reported. Cancer cells uptake exosomes that contain survivin, an anti-apoptotic protein, to protect them from genotoxic stress-induced cell death [bib_ref] cell-permeable survivin inhibits apoptosis while promoting proliferative and metastatic potential, Khan [/bib_ref]. Exosomes from serum of glioblastoma patients contain EGFRvIII mRNA, which stimulate the proliferation of human glioma cells through a self-promoting way [bib_ref] Intercellular transfer of the oncogenic receptor EGFRvIII by microvesicles derived from tumour..., Al-Nedawi [/bib_ref]. Colon cancer cellderived exosomes are enriched in ΔNp73 mRNA. The proliferation potential of target cells is greatly enhanced Linc-ROR HCC cells HCC cells Reduces chemotherapy sensitivity [bib_ref] Extracellular vesicle-mediated transfer of long non-coding RNA ROR modulates chemosensitivity in human..., Takahashi [/bib_ref] by incubation with ΔNp73-containing exosomes [bib_ref] Tumor-derived exosomes are enriched in DeltaNp73, which promotes oncogenic potential in acceptor..., Soldevilla [/bib_ref]. The interaction between tumor stromal cells and tumor cells also efficiently promote tumor growth. Exosomes from chronic myelogenous leukemia (CML) cells stimulate bone marrow stromal cells to produce IL-8, which in turn promote the growth of leukemia cells [bib_ref] Exosome-mediated crosstalk between chronic myelogenous leukemia cells and human bone marrow stromal..., Corrado [/bib_ref]. Bone marrow mesenchymal stromal cells (BM-MSCs) from multiple myeloma (MM) patients release exosomes that express increased levels of oncogenic proteins, cytokines, and adhesion molecules to facilitate the growth of MM cells [bib_ref] BM mesenchymal stromal cell-derived exosomes facilitate multiple myeloma progression, Roccaro [/bib_ref]. Thus, exosomes from tumor cells and microenvironment could act coordinately to promote tumor growth.
## Tumor angiogenesis
The formation of new blood vessels is required for tumor growth and progression. Proteomic analysis has revealed that abundant angiogenic factors are present in malignant mesothelioma-derived exosomes [bib_ref] Hypoxic tumor cell modulates its microenvironment to enhance angiogenic and metastatic potential..., Park [/bib_ref]. Exosome uptake induces upregulation of angiogenesisrelated genes and results in enhanced endothelial cell proliferation, migration, and sprouting [bib_ref] Cell surface tetraspanin Tspan8 contributes to molecular pathways of exosome-induced endothelial cell..., Nazarenko [/bib_ref]. Exosomes derived from hypoxic glioblastoma cells are more potent to induce angiogenesis [bib_ref] Exosomes reflect the hypoxic status of glioma cells and mediate hypoxia-dependent activation..., Kucharzewska [/bib_ref]. Exosomes from metastatic breast cancer cells contain miR-105. Exosome-mediated transfer of miR-105 degrades ZO-1 protein, disturbs tight junctions, and induces vascular permeability in distant organs [bib_ref] Cancer-secreted miR-105 destroys vascular endothelial barriers to promote metastasis, Zhou [/bib_ref]. Exosomal miR-92a from K562 leukemia cells targets integrin α5 to enhance endothelial cell migration and tube formation [bib_ref] Leukemia cell to endothelial cell communication via exosomal miRNAs, Umezu [/bib_ref]. MiR-210 is significantly enriched in exosomes from hypoxic K562 cells, which promotes the angiogenic activity of endothelial cells [bib_ref] Exosomes derived from hypoxic leukemia cells enhance tube formation in endothelial cells, Tadokoro [/bib_ref]. Multiple myeloma cells grown under hypoxic condition produce more exosomes containing miR-135b, which directly suppresses FIH-1, an inhibitor of HIF-1, to enhance endothelial tube formation in endothelial cells [bib_ref] Exosomal miR-135b shed from hypoxic multiple myeloma cells enhances angiogenesis by targeting..., Umezu [/bib_ref]. Exosomes are critically involved in tumor angiogenesis by directly delivering angiogenic proteins into endothelial cells or modulating the angiogenic function of endothelial cells by exosomal miRNAs.
## Tumor metastasis
Exosomes contribute to tumor metastasis by enhancing tumor cell migration and invasion, establishing premetastatic niche, and remodeling the extracellular matrix. EBV-positive nasopharyngeal carcinoma (NPC) cell-derived exosomes contain HIF-1α, which increases migration and invasiveness of EBV-negative NPC cells [bib_ref] Exosomal HIF1alpha supports invasive potential of nasopharyngeal carcinoma-associated LMP1-positive exosomes, Aga [/bib_ref]. Metastatic cancer cells secrete increased level of miRNA with tumor-suppressor function, which may suggest another mechanism for the role of exosomes in metastasis [bib_ref] Cellular disposal of miR23b by RAB27-dependent exosome release is linked to acquisition..., Ostenfeld [/bib_ref]. The formation of pre-metastatic niche is a prerequisite for tumor metastasis. Exosomes from highly metastatic melanoma enhance the metastatic ability of primary tumors by converting bone marrow progenitor cells to a pro-vasculogenic and pre-metastatic phenotype via the MET receptor [bib_ref] Melanoma exosomes educate bone marrow progenitor cells toward a pro-metastatic phenotype through..., Peinado [/bib_ref]. Gastrointestinal stromal tumor cells release exosomes containing protein tyrosine kinase to convert progenitor smooth muscle cells to a premetastatic phenotype [bib_ref] Oncogenic KIT-containing exosomes increase gastrointestinal stromal tumor cell invasion, Atay [/bib_ref]. Suetsugu et al. show that highly metastatic breast cancer cells can transfer their own exosomes to other cancer cells and normal lung tissue cells in vitro and in vivo by using fluorescent protein imaging method [bib_ref] Imaging exosome transfer from breast cancer cells to stroma at metastatic sites..., Suetsugu [/bib_ref] , which provides direct evidence for the involvement of exosomes from highly metastatic cancer cells in educating stromal cells. Luga and colleagues have shown that exosomes produced by stromal cells are taken up by breast cancer cells and are then loaded with Wnt11, which is associated with stimulation of the invasiveness and metastasis of the breast cancer cells [bib_ref] Exosomes mediate stromal mobilization of autocrine Wnt-PCP signaling in breast cancer cell..., Luga [/bib_ref]. Exosomes from activated CD8+ T cells promote cancer cell invasion and lung metastasis via the Fas/FasL pathway [bib_ref] Activated T cell exosomes promote tumor invasion via Fas signaling pathway, Cai [/bib_ref] , which adds another layer of mechanism for the role of tumor-infiltrating lymphocytes in cancer metastasis. Exosome-mediated transfer of oncogenic microRNAs into cancer cells is associated with enhanced metastatic potential. IL-4-activated macrophage-derived exosomes transfer miR-223 to co-cultivated breast cancer cells, leading to increase of cell invasion [bib_ref] Microvesicles secreted by macrophages shuttle invasion-potentiating microRNAs into breast cancer cells, Yang [/bib_ref]. Exosomemediated delivery of miR-221/222 from MSCs to gastric cancer cells greatly enhances gastric cancer cell migration [bib_ref] Deregulated microRNAs in gastric cancer tissue-derived mesenchymal stem cells: novel biomarkers and..., Wang [/bib_ref]. Fabbri et al. suggest that miRNAs in tumor-secreted exosomes can directly bind toll-like receptor (TLR) in immune cells to promote tumor metastasis [bib_ref] MicroRNAs bind to toll-like receptors to induce prometastatic inflammatory response, Fabbri [/bib_ref]. Recently, Costa-Silva and colleagues demonstrate that MIF-containing exosomes from pancreatic ductal adenocarcinoma (PDAC) cells induce TGF-β production in liver Kupffer cells, which in turn upregulates fibronectin (FN) expression by hepatic stellate cells and enhances recruitment of bone marrowderived cells, finally leading to the formation of liver pre-metastatic niche [bib_ref] Pancreatic cancer exosomes initiate pre-metastatic niche formation in the liver, Costa-Silva [/bib_ref] , suggesting a complicated network that involves cancer cells, stromal cells, and immune cells in exosome-initiated pre-metastatic niche formation. Intriguingly, Zomer et al. use the Cre-LoxP system to visualize extracellular vesicle (EV) exchange between tumor cells in living mice [bib_ref] In Vivo imaging reveals extracellular vesicle-mediated phenocopying of metastatic behavior, Zomer [/bib_ref]. They show that the less malignant tumor cells that take up EVs released by malignant tumor cells display enhanced migratory behavior and metastatic capacity, indicating that the metastatic behavior can be phenocopied through extracellular vesicle exchange. Taken together, these findings reveal that the intercellular communication mediated by exosomes may be an important mechanism for tumor metastasis.
## Tumor drug resistance
Exosomes contribute to the development of therapy resistance in tumor cells through a variety of mechanisms.
Tumor-derived exosomes can transfer multi-drug resistance (MDR)-associated proteins and miRNAs to target cells [bib_ref] Docetaxel-resistance in prostate cancer: evaluating associated phenotypic changes and potential for resistance..., Corcoran [/bib_ref] [bib_ref] Exosomal miR-221/222 enhances tamoxifen resistance in recipient ER-positive breast cancer cells, Wei [/bib_ref]. In addition, exosomes participate in the process of tumor resistance by mediating drug efflux. The drugs and their metabolites can be encapsulated and exported by exosomes [bib_ref] Expulsion of small molecules in vesicles shed by cancer cells: association with..., Shedden [/bib_ref] [bib_ref] Abnormal lysosomal trafficking and enhanced exosomal export of cisplatin in drug-resistant human..., Safaei [/bib_ref]. Melanosomal sequestration of cytotoxic drugs contributes to the intractability of malignant melanomas [bib_ref] Melanosomal sequestration of cytotoxic drugs contributes to the intractability of malignant melanomas, Chen [/bib_ref]. Moreover, exosomes may counteract the effect of antibody drugs by modulating their binding to tumor cells. Lymphoma exosomes carry CD20, which bind therapeutic anti-CD20 antibodies and protect target cells from antibody attack [bib_ref] Exosomal evasion of humoral immunotherapy in aggressive B-cell lymphoma modulated by ATP-binding..., Aung [/bib_ref]. Exosomes from HER2-overexpressing breast cancer cells express active HER2 and can bind to the HER2 antibody trastuzumab to inhibit its activity [bib_ref] Potential role of HER2-overexpressing exosomes in countering trastuzumab-based therapy, Ciravolo [/bib_ref]. Exosomes secreted by stromal cells also contribute to tumor drug resistance. BM-MSC-derived exosomes induce multiple myeloma cells resistant to bortezomib through the activation of several survival relevant pathways [bib_ref] Tumour exosomes inhibit binding of tumour-reactive antibodies to tumour cells and reduce..., Battke [/bib_ref]. Therefore, exosomes released by cancer cells and stromal cells may have a potential to modulate sensitivity of cancer cells to distinct therapies.
## Tumor immune escape
Initially reported as tumor-associated antigens and tumor immune response stimulators, the recent studies have shown that tumor-derived exosomes might rather perform immunosuppressive functions. Tumor exosomes block the differentiation of murine myeloid precursor cells into dendritic cells (DC) [bib_ref] Tumor exosomes inhibit differentiation of bone marrow dendritic cells, Yu [/bib_ref]. Tumor exosome-carried TGF-β1 skews IL-2 responsiveness in favor of regulatory T cells and away from cytotoxic cells [bib_ref] Human tumor-derived exosomes selectively impair lymphocyte responses to interleukin-2, Clayton [/bib_ref]. Human nasopharyngeal carcinoma-derived exosomes recruit, expand, and regulate the function of regulatory T cells through CCL20 [bib_ref] Effect of nasopharyngeal carcinoma-derived exosomes on human regulatory T cells, Mrizak [/bib_ref]. NPC cell-derived exosomes impair T cell function, which is associated with upregulated miRNAs in the exosomes [bib_ref] Tumor-derived exosomes promote tumor progression and T-cell dysfunction through the regulation of..., Ye [/bib_ref]. Tumor cell-derived exosomes switch the differentiation of myeloid cells to myeloid-derived suppressor cells (MDSCs) and induce accelerated lung metastasis in a MyD88-dependent manner [bib_ref] Induction of myeloid-derived suppressor cells by tumor exosomes, Xiang [/bib_ref] [bib_ref] Contribution of MyD88 to the tumor exosome-mediated induction of myeloid derived suppressor..., Liu [/bib_ref]. Hsp72 on tumor-derived exosomes promotes the immunosuppressive activity of MDSCs via autocrine activation of IL-6/STAT3 pathway [bib_ref] Membrane-associated Hsp72 from tumor-derived exosomes mediates STAT3-dependent immunosuppressive function of mouse and..., Chalmin [/bib_ref]. Breast cancer cell-derived exosomes simulate the activation of NF-κB and enhance the secretion of pro-inflammatory cytokines in macrophages [bib_ref] Macrophage immunomodulation by breast cancer-derived exosomes requires toll-like receptor 2-mediated activation of..., Chow [/bib_ref]. Exosomes from human prostate cancer cells express ligands for NKG2D on their surface and downregulate NKG2D expression on natural killer (NK) and CD8+ T cells, leading to the impairment of their cytotoxic function [bib_ref] Prostate tumor-derived exosomes down-regulate NKG2D expression on natural killer cells and CD8+..., Lundholm [/bib_ref]. Collectively, these data suggest that tumor-derived exosomes interfere on multiple levels with the immune system to drive tumor immune evasion.
## Tumor-stroma interaction
Tumor stroma is believed to be critically involved in tumor development and progression. Webber et al. suggest that prostate cancer cells could trigger differentiation of fibroblasts into myofibroblasts through exosomal TGF-β [bib_ref] Cancer exosomes trigger fibroblast to myofibroblast differentiation, Webber [/bib_ref]. In addition, prostate cancer exosomes triggered TGFβ1dependent fibroblast differentiation resemble stromal cells isolated from cancerous prostate tissue [bib_ref] Differentiation of tumour-promoting stromal myofibroblasts by cancer exosomes, Webber [/bib_ref] , which accelerates tumor growth by supporting angiogenesis. MSCs function as precursors for tumor myofibroblast. The research from our lab suggests that tumor cellderived exosomes could induce differentiation of human MSCs to carcinoma-associated fibroblasts (CAFs) [bib_ref] Gastric cancer exosomes trigger differentiation of umbilical cord derived mesenchymal stem cells..., Gu [/bib_ref]. Adipose tissue-derived MSCs treated with breast cancerderived exosomes also display the characteristics of myofibroblasts [bib_ref] Exosomes from breast cancer cells can convert adipose tissue-derived mesenchymal stem cells..., Cho [/bib_ref]. [bib_ref] Exosomes mediate stromal mobilization of autocrine Wnt-PCP signaling in breast cancer cell..., Luga [/bib_ref]. Therefore, exosomes may mediate a reciprocal interplay between tumor cells and stromal cells to synergistically promote tumor progression.
## Tumor thrombosis
Tissue factor (TF) overexpression is closely associated with tumor progression. TF can get incorporated into tumor-derived exosomes. The hypercoagulable state in cancer patients may be partially influenced by the release of TF-bearing exosomes from tumor cells. Garnier et al. demonstrate that exosomes link the procoagulant status with metastatic phenotype in cancer. Induction of EMT changes in epithelial cancer cells results in the release of exosomes containing elevated level of tissue factor. Importantly, TF-rich exosomes can be transferred to endothelial cells and cause their exaggerated procoagulant conversion [bib_ref] Cancer cells induced to express mesenchymal phenotype release exosome-like extracellular vesicles carrying..., Garnier [/bib_ref] , suggesting that EMT influences tumor-vascular interaction through altered TF-containing exosomes. However, the exact roles of exosomes in tumor thrombosis and consequent impact on tumor growth, progression, and metastasis remain to be further explored.
## Exosomes as cancer biomarkers and targets
The findings that exosomes play critical roles in almost all aspects of cancer provide opportunities for the development of exosomes as ideal diagnostic biomarkers and therapeutic targets. Exosome-shuttled proteins and nucleic acids have been suggested as novel diagnostic and prognostic indicators for a variety of cancers. Moreover, utilizing tumor-derived exosomes as vaccines and exosomes Urine [bib_ref] Can urinary exosomes act as treatment response markers in prostate cancer?, Mitchell [/bib_ref] from distinct sources as carriers for drugs and small molecules have been proved to be effective in pre-clinical studies and clinical trials.
## Exosomes as cancer diagnostic biomarkers
Exosomes are readily accessible in nearly all body fluids including blood, urine, saliva, and ascites. Exosomes contain bioactive molecules that reflect the pathological state of the originated cells, thus providing an enriched source of biomarkers [fig_ref] Table 2: Exosomes from distinct biofluids of cancer patients as biomarkers [/fig_ref]. The level of exosomes is elevated in the plasma of some cancer patients as compared to healthy controls. There is a positive correlation between the abundance of tumor exosomes and tumor stage in ovarian cancer patients [bib_ref] MicroRNA signatures of tumor-derived exosomes as diagnostic biomarkers of ovarian cancer, Taylor [/bib_ref]. Tumor is characterized by a specific miRNA profile. The majority of circulating microRNAs is concentrated in exosomes [bib_ref] The majority of microRNAs detectable in serum and saliva is concentrated in..., Gallo [/bib_ref]. Exosomal miRNAs have been suggested as diagnostic and prognostic indicators for lung cancer, esophageal squamous cell carcinoma, prostate cancer, breast cancer, glioblastoma, ovarian cancer, and other cancer types [bib_ref] Circulating miRNAs in cancer: from detection to therapy, Wang [/bib_ref] [bib_ref] Glioblastoma microvesicles transport RNA and proteins that promote tumour growth and provide..., Skog [/bib_ref] [bib_ref] Exosomal microRNA: a diagnostic marker for lung cancer, Rabinowits [/bib_ref] [bib_ref] Role of pancreatic cancer-derived exosomes in salivary biomarker development, Lau [/bib_ref] [bib_ref] Prostate cancer-derived urine exosomes: a novel approach to biomarkers for prostate cancer, Nilsson [/bib_ref] [bib_ref] Intracellular and extracellular microRNAs in breast cancer, Corcoran [/bib_ref] [bib_ref] Clinical impact of serum exosomal microRNA-21 as a clinical biomarker in human..., Tanaka [/bib_ref]. Exosomal miRNAs are positively correlated with the stage and degree of cancer progression. In addition to miRNAs, long non-coding RNAs (LncRNAs) are also detected in exosomes [bib_ref] Extracellular vesicle-mediated transfer of a novel long noncoding RNA TUC339: a mechanism..., Kogure [/bib_ref] [bib_ref] Extracellular vesicle-mediated transfer of long non-coding RNA ROR modulates chemosensitivity in human..., Takahashi [/bib_ref]. LncRNA from serum of gastric cancer patients is defined as a novel exosomal biomarker [bib_ref] Plasma long noncoding RNA protected by exosomes as a potential stable biomarker..., Li [/bib_ref] [bib_ref] Combined detection of serum exosomal miR-21 and HOTAIR as diagnostic and prognostic..., Wang [/bib_ref].
## Exosomes as cancer therapy targets exosome-based immunotherapy
Dendritic cell-derived exosomes (dexosomes) have been developed as immunotherapeutic anticancer agents [bib_ref] Dendritic cell-derived exosomes as immunotherapies in the fight against cancer, Pitt [/bib_ref]. Tumor peptide-pulsed DC-derived exosomes suppress growth of established murine tumors in a T celldependent manner [bib_ref] Eradication of established murine tumors using a novel cell-free vaccine: dendritic cell-derived..., Zitvogel [/bib_ref]. Exosomes secreted by living tumor cells contain and transfer tumor antigens to dendritic cells and induce potent CD8+ T cell-dependent antitumor effects on mouse tumors [bib_ref] Tumor-derived exosomes are a source of shared tumor rejection antigens for CTL..., Wolfers [/bib_ref]. Dexosomes have entered clinical trials for colorectal cancer, metastatic melanoma, and non-small cell lung cancer and have achieved modest therapeutic effects [bib_ref] The application of exosomes as a nanoscale cancer vaccine, Tan [/bib_ref].
## Exosome removal for cancer therapy
The removal of exosomes from advanced cancer patients is a novel strategy to treat cancer [bib_ref] Exosome removal as a therapeutic adjuvant in cancer, Marleau [/bib_ref]. Exosome depletion by dimethyl amiloride (DMA) in mice restores the antitumor efficacy of cyclophosphamide (CTX) through the inhibition of MDSC functions. Amiloride, a drug used to treat high blood pressure, inhibits exosome formation and blunts MDSC suppressor functions in colorectal cancer patients [bib_ref] Membrane-associated Hsp72 from tumor-derived exosomes mediates STAT3-dependent immunosuppressive function of mouse and..., Chalmin [/bib_ref]. The biotechnology company Aethlon Medical has developed an adjunct therapeutic method HER2osome, which is able to reduce tumor-secreted HER2 positive exosomes in the circulation and thus inhibit HER2-positive breast cancer progression. However, further work is needed to evaluate the clinical safety of such a treatment strategy based on exosome removal.
## Exosomes as anti-cancer drug delivery vehicles
The use of exosomes as nucleic acid or drug delivery vehicles has gained considerable interest due to their excellent biodistribution and biocompatibility [bib_ref] Exosomes as nucleic acid nanocarriers, Van Den Boorn [/bib_ref]. Exosome-mediated delivery of therapeutic short interfering RNA (siRNA) to the target cells has been tested. The exosome-delivered siRNA is effective at causing posttranscriptional gene silencing and inducing cell death in recipient cancer cells [bib_ref] Exosomes are natural carriers of exogenous siRNA to human cells in vitro, Shtam [/bib_ref] [bib_ref] Microvesicle-mediated delivery of transforming growth factor beta1 siRNA for the suppression of..., Zhang [/bib_ref] [bib_ref] Delivery of siRNA to the mouse brain by systemic injection of targeted..., Alvarez-Erviti [/bib_ref]. To improve drug delivery efficacy to tumors, the researchers have modified exosomes with targeting ligands such as iRGD-Lamp2b. The modified exosomes show highly efficient targeting to αV integrin-positive breast cancer cells, and intravenous injection of these exosomes obviously inhibits tumor growth [bib_ref] A doxorubicin delivery platform using engineered natural membrane vesicle exosomes for targeted..., Tian [/bib_ref]. In addition, exosomes have been utilized as effective vehicle for drug delivery [bib_ref] Biodistribution and delivery efficiency of unmodified tumor-derived exosomes, Smyth [/bib_ref]. Exosomes from MSCs have been tested as the vehicle to package and deliver active drugs such as paclitaxel [bib_ref] Paclitaxel is incorporated by mesenchymal stromal cells and released in exosomes that..., Pascucci [/bib_ref].
# Conclusion
The rapid expansion of the number of published studies on exosomes clearly shows that research on exosomes and their functions is now a very exciting field. Exosomes are small particles with big roles and are emerging as major players in intercellular communication. Exosomes have been suggested as active transporters for proteins, DNA, mRNA, and non-coding RNAs. The roles of exosomes in cancer have been gradually realized. Although some reports have suggested anti-tumor roles of exosomes due to their potential to elicit immune response, most of the reports have revealed the various pro-tumor effects of exosomes, which is further supported by the observations that the level of circulating exosomes is increased in cancer patients and correlated with tumor progression. In this review, we discussed several aspects of exosome biology in cancer. Cancer cells communicate with the surrounding and distant cells via exosomes, which constitutes a LncRNA-p21 Prostate cancer qRT-PCR Higher level of exosomal lncRNA-p21 in patients with prostate cancer than those with benign hyperplasia Plasma [bib_ref] Exosomal lncRNA-p21 levels may help to distinguish prostate cancer from benign disease, Isin [/bib_ref] bi-directional interaction network to synergistically promote cancer development, progression, metastasis, and drug resistance. However, the exact mechanisms mediating the complex roles of exosomes in cancer have not yet fully elucidated. Exosomes would be ideal biomarkers for cancer diagnosis and targeted therapy because they closely represent the state of their parental cells and are relatively stable in the circulation and could be easily collected from body fluids. The potential of exosomal contents for diagnostic and prognostic biomarkers have been investigated in various cancers. It is required to develop faster and more convenient methods for validating the proposed exosomal cargos as biomarkers in specimens from human cancer patients. The use of nanotechnology to load exosomes with small molecules or drugs for cancer therapy has also been exploited. Improvements in developing new strategies to obtain a large amount of exosomes from appropriate donor cells, efficiently introducing the therapeutic agents into exosomes, and optimizing the targeted delivery of exosomes to particular tissues will facilitate the use of exosomes as natural carrier in clinical therapy. Future studies of exosomes will not only shed lights on their roles in the pathogenesis of cancer but will open new avenues for cancer diagnosis and therapeutics.
## Competing interests
The authors declare that they have no competing interest
Authors' contributions XZ and WRX were responsible for the conception and design of the manuscript. All authors participated in the drafting of the manuscript and approved its final version. All authors read and approved the final manuscript.
[fig] Figure 2: Roles of exosomes in cancer. Exosomes are critically involved in tumor initiation, growth, progression, metastasis, and drug resistance by transferring oncogenic proteins and nucleic acids. Tumor-derived exosomes can activate endothelial cells to support tumor angiogenesis and thrombosis. Tumor-derived exosomes can convert fibroblasts and MSCs into myofibroblasts to facilitate tumor angiogenesis and metastasis. Tumor-derived exosomes contribute to create an immunosuppressive microenvironment by inducing apoptosis and impairing the function of effector T cells and NK cells, inhibiting DC differentiation, expanding MDSCs, as well as promoting Treg cell activity. Tumor-derived exosomes can mobilize neutrophils and skew M2 polarization of macrophages to promote tumor progression. Moreover, tumor-derived exosomes can help tumor cells develop drug resistance by transferring multidrug-resistant proteins and miRNAs, exporting tumoricidal drugs, and neutralizing antibody-based drugs. In turn, exosomes from activated T cells, macrophages, and stromal cells can promote tumor metastasis and drug resistance [/fig]
[table] Table 1: Overview on the function of exosomes in cancer [/table]
[table] Table 2: Exosomes from distinct biofluids of cancer patients as biomarkers [/table]
|
Well-differentiated liposarcoma disguised as a recurrent lipoma of the forearm flexor compartment: A case report
# Introduction
Lipoma is one of the most common benign soft tissue tumors. It commonly occurs subcutaneously, however its appearance on any other plane has been reported. Large-sized lipoma or 'giant lipoma' is considered if the size is greater than 5 cm in diameter. We presented a case of a 63-year-old male with giant lipoma on the flexor compartment of his right forearm which was treated with complete removal of the lipomatous mass with marginal excision. Post-operative histopathological diagnosis was well-differentiated liposarcoma lipoma-like. No recurrence had been observed on the 1-year follow up. Our work has been reported in line with the Surgical Case Report (SCARE) Guidelines.
## Case
A 63-year old male was referred to orthopedic outpatient clinic at our institute with a large lump on the right forearm region. The lump initially appeared 28 years ago and had re-appeared for three times after three times removal. The latest removal was 7 years before admitted to our center and the latest recurrence was in the * Corresponding author at: Department of Orthopaedic and Traumatology, Faculty next 2 years. The patient was only complaining about mild discomfort and difficulty in wearing clothes. The non-tender lump was increasing in size slowly from the size of quail eggs with welldefined border around the wrist until it occupied the whole flexor compartment of the right forearm. The right forearm circumference was 20 cm compared to 7 cm on the contralateral. There was neither neurological deficit nor signs of compartment syndrome distal to the lump. The laboratory markers showed no abnormality except for high blood glucose since the patient has diabetes mellitus since the past 11 years. He also had a hyperlipidemia with increase of both total and LDL cholesterol level.
Patient underwent radiograph assessment with plain x ray which suggest lipomatous massand Magnetic Resonance Imaging (MRI) which concluded a prominent lipomatous mass on intermuscular region with possible benign lesion or low-grade malignancy. On the Clinico-Pathological Conference (CPC), this more likely to be benign lipomatous lesion and was decided to undergo surgical exploration, marginal excision and biopsy.
Surgical exploration was therefore performed under general anesthesia and with the application of a hemostatic tourniquet. A lazy-S design was made approximately one cm lateral to the antecubital fossa and extended distally until distal wrist crease. The incision started on the ulnar-volar side on the distal forearm where the initial mass appeared and where it most superficially located under the skin. After skin incision and entering the subcutaneous plane on the wrist region, there was a lipoma-like lesion which continued sub-muscularly underneath the FDS muscle. Blunt dissection through muscle fiber confirmed a deep-seated lipoma lying over the Flexor Pollicis Longus (FPL) muscle. Running through the lipoma was a shiny cord-like structure, which was identified as the median nerve. The mass was carefully dissected off the median n. until releasing the transverse carpal ligament to ensure that the nerve was left intact. The mass was found to compress the surroundings; however, it was relatively easy to remove due to the pseudo-capsule and the intermuscular location.
The 15 × 8 × 5.5 cms mass was sent for histopathological analysis. The mass was well-vascularized as observed that there were big vessels that gave blood supply to the mass and these vessels infiltrated the mass especially at the superior pole, where it gave a yellow-red appearance on that region. The mass consistency was generally rubbery with a firmer consistency at the superior pole..
Histopathological analysis suggested a lipomatous tumor, low grade, most suitable with well-differentiated liposarcoma (WDL)lipoma like subtype.
During routine follow-up in the outpatient clinic at one and three months, the wound had healed well and there were no neurological deficit, functional impairment or recurrence of the lump observed. At one-year follow-up the scar healed nicely without recurrence of the mass.
# Discussion
The etiology and pathogenesis of lipomas remain unknown, although genetic endocrine and traumatic factors have been suggested. Our patient had diabetes mellitus in the past 11 years which was treated by oral anti-diabetic drugs. He also had a longstanding hyperlipidemia with the increase of total cholesterol and LDL cholesterol level.
Typical lipoma occurs more often in young adults, women (thigh), and in overweight populations. They vary in size; when the size is greater than 5 cm in diameter, the term 'giant lipoma' is considered, as it may raise the awareness to other malignant form of lipomatous mass such as liposarcoma. Our patient is a 63-yearold normal-weight man which had first presentation of the lipoma approximately 28 years ago; this is an uncommon characteristic for the typical lipoma. Furthermore, its presentation in volar part of the forearm is also unusual from those of giant lipomas which occur most frequently in cervical region, thorax, and lower limbs (inguinal region and thigh).
Although the definitive diagnosis of giant lipoma can be made only by histopathological examination, the differential diagnosis between giant lipoma and liposarcoma is of great importance. Once suspected, other investigations must be conducted to provide more information about the tumor such as ultrasonography (USG), computed tomography (CT) scan and magnetic resonance imag- ing (MRI) as well as technetium-99 diethylenetriaminepentaacetic acid scanning.
In our case, the plain radiograph showed a radiolucent soft tissue mass along the shaft of radius ulna on the volar part of the right forearm. The MRI showed homogeneous hyperintense fat signal on T1-weighted MR images whereas T2-weighted images showed hypointense fat-suppression signal. The mass appeared inter-muscularly between FCR and FDC with well-defined margin and no infiltration to surrounding structures, raising the susceptibility of a lipomatous tumor with benign characteristics.
Surgical is the mainstay treatment for giant lipoma due to the size, recurrence risk and potential malignant transformation. Complete surgical resection, with preferably wide-margin, is recommended to prevent dedifferentiation and recurrence. Adjuvant radiation therapy may benefit in large high-grade liposarcoma and in any other liposarcoma when the wide resection cannot be obtained. The value of chemotherapy is still debatable as adjuvant to surgical therapy.
In our case due to the complicated structure and relatively narrow space at the forearm, we could not perform wide resection with much negative margins. Intraoperatively, we found that the mass was deep-seated, well-vascularized, well-circumscribed, lobulated and had a rubbery consistency. Those characteristics were suitable for the lipomatous mass but not specifically present liposarcoma. The cut surface was yellow with more hyperemic areas on the superior pole of the mass. A pseudo-capsule which was formed by the continuous pressure on surrounding tissue made the dissection and enucleation of the mass uncomplicated. Since the mass was intermuscular, we were able to remove the tumor without affecting the adjacent muscular tissue while preserving the neurovascular bundle, in this case, the median nerve particularly.
Post-operative histopathological exam revealed this giant lipomatous mass as the well differentiated liposarcomas (WDL) lipoma-like subtype. Account for 40-45 % of liposarcomas, WDL usually diagnosed at the 5th decades of life with slight male predominance. WDL are locally aggressive mesenchymal tumors composed of mature adipocytes and stromal cells with at least focal cytologic atypia. WDL are considered as low grade lipogenic tumors with morphologic subtypes dependent on the adipocytic component and background cellularization characteristic. Several histologic subtypes have been described: lipoma-like, sclerosing, inflammatory, mixed, lipo-leiomyosarcoma, WDL with low grade osteosarcoma-like and spindle cell subtype. Lipoma-like subtype, being the most common form of WDL, is grossly indistinguishable from lipoma. It frequently contains lipoblast and scattered atypical cells may be diffuse or extremely rare.
In this report, surgical excision resulted in complete relief of symptoms and significant esthetic also became evident in the affected forearm. There was no recurrence in the 1-year follow up. However, we still closely observe this patient for any recurrence since the possible risk of recurrence in the WDL as well as inter-muscularly located lipomatous mass.
# Conclusion
The slow growing nature and radiological characteristic of lipoma brought confusion with the benign lesion. However, the giant size and other unusual characteristics (gender, age, mass location and recurrences) raised the awareness of malignant lipomatous mass. Appropriate diagnosis and complete surgical resection of this tumor provided symptoms relief and esthetic improvement on the affected limb. Long-term evaluation must be conducted in order to detect recurrence. We plan for the wide excision followed by radiotherapy for possible recurrence mass in the future.
## Declaration of competing interest
The authors declare no conflicts of interest.
# Funding
The authors report no external source of funding during the writing of this article.
# Ethical approval
Ethical approval was not required in the treatment of the patient in this report.
## Consent
Written consent has been received from the subject.
# Author contribution
Dina Aprilya contributes to the study concept or design, data collection and writing the paper.
Wildan Latief contributes in the study concept or design, data collection, analysis and interpretation, oversight and leadership responsibility for the research activity planning and execution, including mentorship external to the core team.
Wahyu Widodo contributes in the study concept or design, data collection, analysis and interpretation, oversight and leadership responsibility for the research activity planning and execution, including mentorship external to the core team.
## Registration of research studies
Not required.
## Guarantor
Wahyu Widodo is the sole guarantor of this submitted article.
## Provenance and peer review
Editorially reviewed, not externally peer-reviewed.
|
Erdheim-Chester Disease with Emperipolesis: A Unique Case Involving the Heart
Histiocytosis is an uncommon disease characterized by excessive accumulation of histiocytes. Here, we report a rare case of non-Langerhans-cell histiocytosis in a 51year-old woman who presented with severe symptoms of pericardial effusion. Radiologic investigation also detected multiple bone (lower limbs, vertebrae, ribs, and ilium) lesions. Resected pericardium showed abundant mono-or multi-nucleated non-foamy histiocytes (CD68 + /CD163 + /S-100 + /CD1 /langerin ) in a fibroinflammatory background. The histiocytes demonstrated emperipolesis of lymphocytes, a hallmark feature of Rosai-Dorfman disease (RDD). However, molecular analysis revealed a BRAF V600E mutation of the proliferating histiocytes, highlighting the neoplastic features frequently observed in another non-Langerhans-cell histiocytosis known as Erdheim-Chester Disease (ECD). We consider this case to be a unique presentation of ECD harboring some RDD-like cells with emperipolesis, but not a case of RDD with a BRAF mutation concerning its clinical manifestation (involvement of the heart and bones) and neoplastic features.
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
# Introduction
Histiocytosis refers to a group of rare diseases characterized by the abnormal accumulation of macrophages, dendritic cells or monocyte-derived cells [bib_ref] Revised classification of histiocytoses and neoplasms of the macrophage-dendritic cell lineages, Emile [/bib_ref]. Recently, a revised classification schema divided histiocytosis into five groups: (1) Langerhans-related, [bib_ref] Erdheim-Chester disease harboring the BRAF V600E mutation, Blombery [/bib_ref] cutaneous and muco-cutaneous, (3) malignant histiocytoses, (4) Rosai-Dorfman disease (RDD), and (5) haemophagocytic lymphohistiocytosis and macrophage activation syndrome [bib_ref] Revised classification of histiocytoses and neoplasms of the macrophage-dendritic cell lineages, Emile [/bib_ref]. Erdheim-Chester disease (ECD) and RDD are common histiocytic disorders affecting adult bone and/or heart. ECD, a member of the "Langhans-related" group according to the latest revised classification, is a histiocytosis characterized by foamy histiocytes (CD68 + /CD163 + /S-100 /+ /CD1 /langerin ) accompanied by lymphocytes, Touton giant cells, and dense fibrosis [bib_ref] Erdheim-Chester disease harboring the BRAF V600E mutation, Blombery [/bib_ref]. The radiographic features of symmetric diaphyseal and metaphyseal osteosclerosis in the lower limbs are mostly present in patients with ECD. Other classic extraskeletal organs involved include the cardiovascular, pulmonary, neurological, and endocrine systems [bib_ref] Detection of an NRAS mutation in Erdheim-Chester disease, Diamond [/bib_ref] [bib_ref] Consensus guidelines for the diagnosis and clinical management of Erdheim-Chester disease, Diamond [/bib_ref]. RDD, which is also known as sinus histiocytosis with massive lymphadenopathy, is diagnosed on the basis of abundant pale cytoplasm in the histiocytes (CD68 + /CD163 + /S-100 + /CD1a -/langerin -) and the presence of lymphocytes and plasma cells infiltrating the background. Emperipolesis of histiocytes is a hallmark feature of RDD that has not yet been observed in ECD [bib_ref] Emperipolesis: a review, Rastogi [/bib_ref]. In contrast to ECD, which is characterized by its neoplastic features and often harbors a BRAF mutation, RDD is mostly defined as a reactive disorder of histiocytosis [bib_ref] Erdheim-Chester disease, Haroche [/bib_ref].
Here, we present an extraordinary case of non-Langerhans-cell histiocytosis (CD68 + /CD163 + /S-100 + /CD1a -/langerin -) involving the heart and skeleton in which emperipolesis and BRAF mutation was confirmed. We consider this case the first report of an atypical ECD with emperipolesis rather than RDD with a BRAF mutation. This study was approved by the Ethics Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. Written informed consent was obtained from the patient.
## Case report
A 51-year-old woman was referred to our hospital because of shortness of breath, chest pain, recurrent edema of the lower extremities, and night sweats without fever or vomiting. Ultrasound and a chest enhanced computed tomography scan showed liquid density in the pericardial cavityand B) and soft tissue density in the pericardial wall. Positron emission tomography showed radiotracer uptake in the wall of the pericardium and multiple bones, including the vertebrae, ilium, ribs, and lower limbs. Analysis using 99m Tc bone scintigraphy also demonstrated significant symmetrical radiotracer uptake in the distal ends of the femurs and the proximal and distal tibia, in addition to the vertebrae, ribs, and ilium. To relieve the symptoms caused by the pericardial effusion, pericardial fenestration and pericardiectomy were performed. A fragment of the hard and grey-white pericardial tissue measuring 7 cm2.5 cm0.4 cm was excised and submitted for pathological examination. A biopsy sample from the bone lesion could not be obtained due to patient refusal.
The surgical samples were fixed in 4% buffered formalin, embedded in paraffin, cut into 4 µm sections and stained with hematoxylin and eosin (H&E) for histological evaluation. Immunohistochemical staining was performed on paraffin-embedded sections using a three-step Avidin-Biotin Complex staining method (VECTASTAIN ABC kit, Vector Laboratories, Burlingame, CA). The primary antibodies included langerin (1:100, clone 12D6, Novocastra, Newcastle upon Tyne, UK), CD1 (1:100, clone O10, Dako, Glosrup, Denmark), S-100 (1:1,000, clone Z0311, Dako), CD68 (1:50, clone KP-1, Dako), CD163 (1:100, clone Ber-MAC3, Dako), smooth muscle actin (SMA; 1:400, clone M0879, Dako), desmin (1:200, clone D33, Dako), and pan-cytokeratin (pCK; 1:300, Dako). Diaminobenzidine was used as chromogen. Nuclei were stained with Mayer's hematoxylin. Appropriate positive controls were included. For molecular analysis, genomic DNA was extracted from formalin-fixed, paraffinembedded tumor specimens using the TIANamp Genomic DNA Kit (Tiangen Biotech Co., Ltd., Beijing, China) according to the manufacturer's instructions. The BRAF mutation was detected with the AmoyDx Human BRAF V600E Mutation Detection kit using a BioRad CFX96 Real-Time PCR Machine (BioRad Laboratories, Inc., Berkeley, CA). Briefly, 4.7 µL DNA was added to 35.3 µL polymerase chain reaction (PCR) master mix containing PCR primers, fluorescent probes, PCR buffer, and Taq DNA polymerase. The PCR conditions were as follows: 5-minute denaturation at 95°C, followed by 15 cycles of 95°C for 25 seconds, 64°C for 20 seconds, 72°C for 20 seconds, and then 31 cycles of 95°C for 25 seconds, 60°C for 35 seconds, and 72°C for 20 seconds. The fluorescent signal was collected from the FAM and HEX channels. The results were analyzed according to the instructions in the user manual.
Microscopic examination of the H&E-stained pericardial lesion revealed an exuberant histiocytic and chronic fibroinflammatory process. As shown inand B, the inflammatory cells were primarily composed of lymphocytes, plasma cells, sparse neutrophils, and eosinophils. Intermixed with the infla mmatory cells and fibrotic collagen fibers were scattered mononucleated or multinucleated histiocytes characterized by round or oval vesicular or hyperchromatic nuclei containing distinct nucleoli and abundant amphophilic cytoplasm. Under higher magnification, some of the histiocytes exhibited emperipolesis. Immunohistochemically, these histiocytes showed positive staining of CD68, CD163, and S-100and negative staining of Langerin, CD1, and other relevant markers (SMA, desmin, and pCK). The positive S-100 staining highlighted the emperipolesis of the histiocytes. Reverse transcription polymerase chain reaction was performed, and a BRAF V600E mutation was detected .
Based on the overall findings, a diagnosis of atypical ECD with emperipolesis was considered. After diagnosis, the patient underwent interferon injections subcutaneously for 7 months. During the follow-up period, the patient indicated that her quality of life had improved because she could now get out of bed and was functioning well in her daily life, but no radiologic investigation was performed. VOLUME 49 NUMBER 2 APRIL 2017 555
# Discussion
Histiocytosis refers to a heterogeneous group of diseases characterized by proliferation of cells thought to be derived from dendritic cells, macrophages, or monocyte derived cells. Differentiation between the subtypes relies on the morphology, immunohistochemistry, molecular analysis, and the clinical setting. The major differential diagnosis in this case included Langerhans cell histiocytosis (LCH), RDD, and ECD.
In the present case, LCH was easily excluded because (1) the lesional histiocytes did not show intranuclear pseudoinclusions, prominent nuclear indentations, or grooves; (2) the histiocytes were CD1 and langerin negative, which is inconsistent with the diagnostic immunophenotype for LCH (S-100 + /CD1 + /langerin + ); and (3) LCH in adults mostly involves bones (52%), lungs (40%), or skin (7%) [bib_ref] Langerhans cell histiocytosis: a review of the current recommendations of the Histiocyte..., Satter [/bib_ref]. Cardiovascular involvement has only been reported in one case based on radiographic findings and pathologic examination [bib_ref] Langerhans' cell histiocytosis presenting with a para-aortic lesion and heart failure, Chen [/bib_ref].
Based on the histopathological features, the morphological characteristics observed in this case were consistent with those of RDD. The lesion was characterized by cytoplasm rich, proliferative mononucleated or multinucleated histiocytes in a fibroinflammatory background. Furthermore, the immunophenotype of the cells was identical to that found in RDD (CD68 + /S-100 + /CD1 /langerin ). Moreover, the his-
## A b
C D F E tiocytes demonstrated emperipolesis, a conventional hallmark feature of RDD, which is facilitated by positive staining of S-100 protein. However, these findings are inconsistent with the typical histopathological features of ECD, which is generally characterized by xanthomatous infiltration of lipidladen foamy histiocytes (CD68 + /S-100 /+ /CD1 /langerin ) within a variably fibrous stroma with admixed chronic inflammatory cells and Touton giant cells [bib_ref] Consensus guidelines for the diagnosis and clinical management of Erdheim-Chester disease, Diamond [/bib_ref]. It has been reported that, in the extraskeletal ECD, histiocytes have a less or non-foamy appearance, but display a granular appearance in a fibrous stroma without the presence of Touton giant cells [bib_ref] Consensus guidelines for the diagnosis and clinical management of Erdheim-Chester disease, Diamond [/bib_ref]. This atypical morphological feature of extraskeletal ECD may make differential diagnosis of ECD and RDD challenging if emperipolesis is not taken into consideration. Emperipolesis is an uncommon physiological or pathological process in which the engulfed hematopoietic or inflammatory cells exist in another living cell, are viable, and can exit at any time. To date, emperipolesis has been considered one of the diagnostic features of RDD, as well as some hematolymphoid disorders (lymphoma, leukemia, myeloproliferative disorders, etc.) and non-hematological malignant tumors (giant cell tumors of the lung, neuroblastoma, etc.) [bib_ref] Emperipolesis: a review, Rastogi [/bib_ref]. However, emperipolesis has never been observed in ECD and LCH. Apparently, the histopathological features of the pericardial lesion are more closely related to RDD.
With regard to its clinical manifestation, besides lymphadenopathy, various extranodal involvement of RDD has also been reported, such as skin, soft tissue, respiratory tract, bone, eye, and retro-orbital tissue, as an isolated lesion or a concomitant lesion of lymphadenopathy [bib_ref] Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease): review of the entity, Foucar [/bib_ref]. However, involvement of the cardiovascular system is extremely rare. To date, there have only been eight adult cases of RDD with heart lesions reported. All of the patients with RDD were diagnosed between the ages of 40 and 69 years based on incidental findings on chest radiographs or because of complaints of chest pain with dyspnea. In these patients, the heart lesion was either isolated or concomitant with lymphadenopathy, with no evidence of bone osteosclerosis [bib_ref] Isolated cardiac involvement of Rosai-Dorfman disease, Sarraj [/bib_ref] [bib_ref] Extranodal Rosai-Dorfman disease involving the heart: report of two cases, Maleszewski [/bib_ref] [bib_ref] Rosai-Dorfman disease of multiple organs, including the epicardium: an unusual case with..., Chen [/bib_ref] [bib_ref] Extranodal Rosai-Dorfman disease involving the right atrium in a 60-yearold male, Bi [/bib_ref]. During follow-up, the patients displayed no signs of disease progression without any treatment, with the exception of one patient who died following the development of pericardial and bilateral pleural effusions [bib_ref] Isolated cardiac involvement of Rosai-Dorfman disease, Sarraj [/bib_ref] [bib_ref] Extranodal Rosai-Dorfman disease involving the heart: report of two cases, Maleszewski [/bib_ref] [bib_ref] Rosai-Dorfman disease of multiple organs, including the epicardium: an unusual case with..., Chen [/bib_ref] [bib_ref] Extranodal Rosai-Dorfman disease involving the right atrium in a 60-yearold male, Bi [/bib_ref]. However, in this case, the patient presented with lesions located in the pericardium and in multiple bones. Osteosclerosis of long bones, especially the distal ends of the femurs and the proximal and distal tibia, and pericardial lesions are the most common radiographic findings in ECD [bib_ref] A Unique case of Erdheim-Chester disease with axial skeleton, lymph node, and..., Lim [/bib_ref]. Pericardial thickening and effusion occur in approximately 40% of patients with ECD who have bone lesions [bib_ref] Consensus guidelines for the diagnosis and clinical management of Erdheim-Chester disease, Diamond [/bib_ref]. In contrast to the indolent and self-limited clinical course of RDD without therapy, interferon (IFN-) and pegylated IFN- are the recommended first-line therapy for ECD. In this case, IFN- therapy led to significant symptomatic relief during the follow-up. Based on this patient's clinical course, ECD is the Currently, RDD is regarded as a reactive/non-neoplastic disorder of the histiocytes, and no genetic alterations have been identified to date [bib_ref] High prevalence of BRAF V600E mutations in Erdheim-Chester disease but not in..., Haroche [/bib_ref]. Conversely, the recent discovery of BRAF V600E (54%), NRAS (18%), and phosphoinositide 3-kinase (13%) mutations confirm the clonal features associated with ECD [bib_ref] High prevalence of BRAF V600E mutations in Erdheim-Chester disease but not in..., Haroche [/bib_ref]. BRAF/RAS/PI3KCA mutations result in activation of the RAS-ERK pathway, a process that may play an important role in the pathogenesis of ECD. Our case demonstrated an apparent BRAF V600E mutation, supporting the neoplastic features that are more consistent with ECD.
With regard to the distinct morphological features, immunohistochemical profile, molecular abnormalities, and the clinical setting, this case represents a rare non-LCH histiocytosis characterized by emperipolesis and a BRAF mutation involving the heart and multiple bones. Additional cases are needed to determine whether this case is an atypical ECD with emperipolesis, an RDD with a BRAF mutation, or a new entity of non-LCH between RDD and ECD. We consider this to be a unique case of ECD with atypical morphology, especially the presence of RDD-like cells with emperipolesis.
## Conflicts of interest
Conflict of interest relevant to this article was not reported.
[fig] Figure 1: (A) Ultrasound shows a massive circumferential pericardial effusion. (B) Computerized tomography shows pericardial effusion and pericardial soft tissue density. (C, D) Computerized tomography and positron emission tomography shows radiotracer uptake in the thoracic vertebra (C) and ilium (D). (E) Electrical capacitance tomography demonstrates symmetrical radiotracer uptake in the distal ends of the femurs and the proximal and distal tibia, as well as the ribs and vertebrae. [/fig]
[fig] Figure 2: (A) The lesion shows infiltration of non-foamy histocytes (arrow) in a marked fibroinflammatory background (H&E staining, 100). (B) The lesion shows granular histiocytes in a fibroinflammatory background. Emperipolesis (arrow) shows engulfed intact lymphocytes inside the cytoplasm of non-foamy histiocytes (H&E staining, 200). (C, D) Positive immunostaining of CD68 (C, 100) and CD163 (D, 100) in non-foamy histiocytes in a diffuse cytoplasmic pattern. (E) The histiocytes show strong cytoplasmic and nuclear staining for S-100. Engulfed lymphocytes (arrows) are well demonstrated in some histiocytes as emperipolesis (200). (F) The histiocytes are negative for Langerin (100). [/fig]
|
Copper-Catalyzed Oxidative Dehydrogenative Carboxylation of Unactivated Alkanes to Allylic Esters via Alkenes
We report copper-catalyzed oxidative dehydrogenative carboxylation (ODC) of unactivated alkanes with various substituted benzoic acids to produce the corresponding allylic esters. Spectroscopic studies (EPR, UV−vis) revealed that the resting state of the catalyst is [(BPI)Cu(O 2 CPh)] (1-O 2 CPh), formed from [(BPI)Cu(PPh 3 ) 2 ], oxidant, and benzoic acid. Catalytic and stoichiometric reactions of 1-O 2 CPh with alkyl radicals and radical probes imply that C−H bond cleavage occurs by a tert-butoxy radical. In addition, the deuterium kinetic isotope effect from reactions of cyclohexane and d 12 -cyclohexane in separate vessels showed that the turnoverlimiting step for the ODC of cyclohexane is C−H bond cleavage. To understand the origin of the difference in products formed from copper-catalyzed amidation and copper-catalyzed ODC, reactions of an alkyl radical with a series of copper−carboxylate, copper−amidate, and copper−imidate complexes were performed. The results of competition experiments revealed that the relative rate of reaction of alkyl radicals with the copper complexes follows the trend Cu(II)−amidate > Cu(II)−imidate > Cu(II)−benzoate. Consistent with this trend, Cu(II)−amidates and Cu(II)−benzoates containing more electron-rich aryl groups on the benzamidate and benzoate react faster with the alkyl radical than do those with more electron-poor aryl groups on these ligands to produce the corresponding products. These data on the ODC of cyclohexane led to preliminary investigation of copper-catalyzed oxidative dehydrogenative amination of cyclohexane to generate a mixture of N-alkyl and N-allylic products.
# ■ introduction
The oxidation of alkanes to alcohols or ketones and the dehydrogenation of alkanes to alkenes are both widely studied targets for C−H bond functionalization. For example, the oxidation of cyclohexane to a mixture of cyclohexanol and cyclohexanone is a large-scale commercial process for the production of adipic acid.The oxidation of propene at the allylic C−H bond to form acrolein 7 also is a well-known largescale C−H bond oxidation process, and the oxidation of allylic C−H bonds to allylic esters is being studied actively for applications in target-oriented synthesis. The dehydrogenation of light alkanes is being studied as a route to ethylene, propene, butene, butadiene, isobutene, and isoprene, with hydrogen as the single side productor with an oxidant to consume the hydrogen and make the reaction, called oxidative dehydrogenation (ODH), favorable thermodynamically.
Although alkane dehydrogenation and allylic oxidation are both known reactions, the combination of these two reactions in a single process is rare. One can envision that such a process could occur by initial dehydrogenation of an alkane to an alkene, followed by oxidation of allylic C−H bonds in the alkene product. Indeed, one form of such a reaction is the wellestablished synthesis of maleic anhydride from butane.However, the combination of dehydrogenation and selective oxidation of the alkene to an allylic alcohol derivative directly from an alkane is poorly developed.
Recently, two examples of copper-catalyzed oxidative dehydrogenative cross-coupling reactions of an aldehyde and toluene with cyclohexane to generate allylic esters have been reported. However, the yields of these reactions were generally low and occurred with limited substrate scope. Moreover, the mechanisms of these reactions were not studied in depth. Copper-catalyzed combinations of alkane dehydrogenation and aziridinationor epoxidation 23 also have been reported, but the epoxide and aziridine are just one component of a mixture of products, and they formed with a maximum of 3−4 turnovers. Thus, a high-yield combination of dehydrogenation and C−H bond oxidation of an alkane to form an allylic alcohol derivative that occurs with tolerance for a wide range of functional groups is not known.
Herein, we report the copper-catalyzed oxidative dehydrogenative carboxylation (ODC) of unactivated alkanes in the presence of carboxylic acid derivatives to form the corresponding allylic ester (Scheme 1). This reaction is related to the classic Kharasch−Sosnovsky reaction, 24 but the starting material is an alkane, rather than an alkene. The reactions occur by oxidative dehydrogenation of an alkane and oxidation of the resulting allylic C−H bond. Detailed mechanistic studies show that the tert-butoxy radical abstracts a C−H bond of cyclohexane to generate a transient cyclohexyl radical, and this radical is converted to cyclohexene by a copper−benzoate complex. The cyclohexene is then oxidized to form the allylic ester product. The relative rates for trapping of the radical by the ligand on copper versus conversion of the radical to an alkene control the selectivity for the formation of allylic vs alkyl ester products.
■ RESULTS AND DISCUSSION 1. Development of Intermolecular ODC of Cyclohexane. To extend our recently published copper-catalyzed amidation of cyclohexane to the acetoxylation or benzyloxylation of cycloalkanes, we conducted the reaction of cyclohexane with benzoic acid and tBuOOtBu in the presence of [(phen)Cu](μ 2 -I) 2 (phen = 1,10-phenanthroline).We envisioned that the combination of the Cu(I) and tBuOOtBu should generate tBuO - , which could generate cyclohexyl radical, and this radical could combine with [(phen)Cu-(O 2 CPh) 2 ] 26,27 to form cyclohexyl benzoate. Although benzoic acid did react with cyclohexane and tBuOOtBu in the presence of 2.5 mol% of [(phen)Cu](μ 2 -I), this combination of materials yielded the allylic ester cyclohex-2-en-1-yl benzoate (21%) and methyl benzoate (16%), not the alkyl benzoate (Scheme 2). Cyclohexyl benzoate was not detected. Thus, the reaction of cyclohexane with benzoic acid and tBuOOtBu in the presence of copper occurs by a combination of dehydrogenation and C−H bond carboxylation.
To increase the yield of the allylic ester from this reaction, we evaluated the reactivity of benzoic acid (0.5 mmol) and cyclohexane (10 equiv) with a series of copper salts and discrete copper complexes. The results of these experiments are presented in [fig_ref] Table 1: Reaction Development of Catalytic ODC of Cyclohexane a [/fig_ref]. Simple Cu(I) and Cu(II) halides catalyzed the coupling of benzoic acid and cyclohexane to produce cyclohex-2-en-1-yl benzoate in moderate to good yields (entries 1−7). For example, the combination of CuCl (5 mol%) and tBuOOtBu (3 equiv relative to benzoic acid) gave a high yield of product (76%). Reactions conducted with a higher 10 mol% loading of Cu occurred in a lower yield (54%) than did the reaction with 5 mol% copper (entry 4). A similar trend of lower yield with higher loadings of catalyst was observed for the copper-catalyzed amidation of cyclohexane.The lower yield of product from reactions containing higher concentrations of copper presumably results from quenching of the transient tert-butoxy radical by Cu(I) to form Cu(II)-OtBu species.The reaction requires both copper and oxidant (entry.
The reaction also occurred when catalyzed by Cu(I) complexes containing neutral bidentate nitrogen ligands, such as 4,7-dichloro-1,10-phenanthroline, 3,4,7,8-tetramethyl-1,10phenanthroline, 1,10-phenanthroline, 4,7-dimethyl-1,10-phenanthroline, 4,5-diazafluoren-9-one, and 1,10-phenanthroline-5,6-dione.The reactions catalyzed by these Cu(I) complexes ligated by dative nitrogen ligands formed cyclohex-2-en-1-yl benzoate (23−64%) in modest yields.Reactions catalyzed by the well-defined [(L1)CuCl] (L1 = Me 2 NCH 2 CH 2 NCH(2-O-C 6 H 4 ), [(BPI)Cu(PPh 3 ) 2 ], and [(BPI)CuCl] (BPI = bis(2pyridylimino)isoindole) reproducibly produced the product in 40−50% yields (entries 9−13). The reactions catalyzed by ligated copper complexes gave larger amounts of methyl benzoate than did reactions with unligated copper.Consistent with this observation, Kochi reported that nitrogen-ligated Cu(II) complexes oxidize alkyl radicals to alkenes more slowly than do simple Cu(II) salts. Although these reactions with ligated copper occurred in lower yield than those with simple copper halides, they did give substantial amounts of product and were valuable for studying the mechanism of this reaction (vide inf ra).
2. Scope of Intermolecular Oxidative Dehydrogenative Carboxylations of Alkanes. The scope of the ODC of cyclohexane with carboxylic acids to form allylic esters is presented in . The yields of these reactions are based on carboxylic acid. The mass balance consisted of unreacted carboxylic acid and methyl benzoate, the origin of which will be discussed later in the paper. The reaction is tolerant of halogens on the benzoic acid 4-X-C 6 H 4 -CO 2 H (X = F (1a), Cl (2a), Br (3a)), forming the corresponding allylic esters in 57−79% yields in these cases. The reaction is also tolerant of a halide (1c, 2c), methoxy (4c), and acetyl group (5c) in the ortho position. Carboxylic acids containing electron-donating substituents on the aromatic system, such as methyl (1b,d,e), tertbutyl (2b), methoxy (1f), 4-phenoxy (2f), and phthalimido (l) groups, generated the corresponding allylic ester products in 56−71% yields. Substrates containing electron-withdrawing substituents, such as acetyl (m), trifluoromethyl (j), cyano (p), and carboalkoxy (n) groups, also gave the corresponding products in moderate to good yields (52−76%). Even a thioether (h) is tolerated, despite the oxidizing conditions of the catalytic reaction; cyclohex-2-en-1-yl 4-(methylthio)benzoate was produced in 69% yield. The reactions with heteroaryl carboxylic acids, such as furan (q) and thiophene (r), also gave substantial yields of allylic esters; however, pyridine carboxylic acids did not yield allylic oxidation products.
Finally, vinyl and aliphatic carboxylic acids reacted to form allylic esters. Specifically, the ODC of cyclohexane with cyclohexanecarboxylic acid (s), (E)-2-methyl-3-phenylacrylic acid (t), octanoic acid (1u), and phenylacetic acid (2u) gave cyclohex-2-en-1-yl cyclohexanecarboxylate (62%), cyclohex-2en-1-yl (E)-2-methyl-3-phenylacrylate (57%), cyclohex-2-en-1yl octanoate (69%), and cyclohex-2-en-1-yl 2-phenylacetate (69%), respectively.
The reaction also occurred with smaller or larger cycloalkanes and, to an extent, with acyclic alkanes. Reactions of benzoic acids with cyclopentane and cycloheptane in the presence of 5 mol% of CuCl yielded the corresponding products in good yield (cyclopent-2-en-yl benzoate (1v, 75%), cyclohept-2-en-yl benzoate (2v 75%)), but the reaction with cyclooctane formed cyclooct-2-en-yl benzoate (3v, 12%) in modest yield. In addition to reactions of cyclic alkanes, reactions of linear alkanes (i.e., pentane) containing multiple C−H bonds were performed to assess the selectivity of the catalytic ODC. The reaction of pentane and benzoic acid in the presence of 2.5 mol% of 1-PPh 3 and tBuOOtBu produced two products: pent-en-2-yl benzoate 34 (1w, 26%) and pent-1-en-3yl benzoate(2w, 10%). The potential product of pen-2-en-1yl benzoate, which would be obtained from the oxidation of the pent-2-ene intermediate at the primary C−H bond, was not observed. This observation suggests that oxidation of a secondary allylic C−H bond is favored over oxidation of a primary allylic C−H bond. This relative reactivity is consistent with the relative C−H bond dissociation energies.The reaction of 2,2-dimethylpentane and benzoic acid produced 4,4dimethylpent-1-en-3-yl benzoate 37 (y, 16%) and methyl benzoate (80%) as the major byproduct.
3. Synthesis of Cu(I) and Cu(II) Complexes and Determination of the Resting State of the Catalyst. Although most of the catalytic reactions were performed with CuCl as catalyst, copper complexes ligated by the imidobipyridine ligand BPI did catalyze the reaction, and the molecular complex [(BPI)Cu(O 2 CPh)] (1-O 2 CPh) was amenable to isolation. The soluble, single-component Cu(II) species 1-O 2 CPh was prepared in 80% yield as a green solid by salt metathesis between [(BPI)CuCl] (1-Cl) and NaO 2 CPh in MeOH at room temperature for 3 h (Scheme 3). Elemental analysis of the product was consistent with the proposed atomic composition for 1-O 2 CPh. We suspect that the molecular structure of 1-O 2 CPh is similar to that of the derivatives of . Intermolecular ODC Cyclic and Acyclic Alkanes to Allylic Esters a a Yields were determined by 1 H NMR spectroscopy with MeNO 2 as internal standard, added after the reaction, and reported as an average of two reactions. 1 H NMR chemical shifts were compared to authentic allylic ester products. b 2.5 mol% of 1-PPh 3 . c 5 mol% of 1-PPh 3 , MeCN. phth = phthalimide.
## Scheme 3. syntheses of cu(i) and cu(ii) complexes
Journal of the American Chemical Society [(BPI)CuX] (X = 2,6-dimethoxybenzoate and 3,4-dimethoxybenzoate) (vide inf ra). To assign the oxidation state of the copper center, we performed X-band EPR measurement on 1-O 2 CPh in toluene at 25 K. The X-band EPR spectrum of 1-O 2 CPh revealed an axial pattern, consistent with a Cu(II) (S = 1/2) center.
To isolate a discrete Cu(I) complex, [(PPh 3 ) 2 Cu(OAc)] 38 was allowed to react with NaBPI. The reaction in toluene at room temperature formed [(BPI)Cu(PPh 3 ) 2 ] (1-PPh 3 ) in 58% yield as an orange crystalline solid (Scheme 3). Compound 1-PPh 3 was characterized by multinuclear ( 1 H,C, 31 P) NMR spectroscopy, FT-IR spectroscopy, and elemental analysis. With discrete Cu(I) and Cu(II) complexes in hand, we investigated the resting state of the catalyst.
The resting state of the copper species in the reaction between cyclohexane and benzoic acid catalyzed by 1-PPh 3 with tBuOOtBu as oxidant was determined by UV−vis spectroscopy, X-band EPR spectroscopy, and independent synthesis of copper complexes.A mixture of benzoic acid, cyclohexane, and tBuOOtBu with 5 mol% 1-PPh 3 in benzene was allowed to react for 2 h at 100°C. The UV−vis spectrum of this reaction mixture was identical to that of independently synthesized 1-O 2 CPh recorded in benzene.Likewise, the Xband EPR spectrum of the reaction mixture collected at 25 K was identical to that of an authentic sample of 1-O 2 CPh.To assess the identity of this complex further, a stoichiometric reaction of 1-PPh 3 with benzoic acid (1.5 equiv) and tBuOOtBu in the absence of cyclohexane was conducted at 100°C for 0.5 h in benzene. This reaction afforded 1-O 2 CPh in 64% isolated yield (Scheme 4), as determined by FT-IR, X-band EPR spectroscopy, elemental analysis, and comparison of the material to 1-O 2 CPh synthesized independently from the salt metathesis reaction of 1-Cl with NaO 2 CPh. These data from the spectroscopic measurement of the copper species in the catalytic reactions and of the species formed independently from stoichiometric reactions strongly indicate that a copper(II)−benzoate complex is the resting state of the catalyst.
## Elucidation of the mechanism by stoichiometric reactions, trapping experiments, and competition
Experiments. To assess the role of 1-O 2 CPh in the catalytic reaction, we performed stoichiometric reactions of 1-O 2 CPh with the reaction components. The reaction of 1-O 2 CPh with cyclohexane was conducted in the presence of tBuOOtBu at 100°C for 21 h in acetonitrile (Scheme 5). The products consisted of cyclohex-2-en-1-yl benzoate (44%) and methyl benzoate (49%). These results are consistent with competitive reactions of a cyclohexenyl radical and a methyl radical with 1-O 2 CPh to produce cyclohex-2-en-1-yl benzoate and methyl benzoate, respectively. The analogous reaction performed in the absence of tBuOOtBu gave no product from reaction of the cyclohexane. These results show that the copper−benzoate does not react directly with the alkane. Instead, a species generated from copper and tBuOOtBu reacts with the alkane.
To assess the sequence of bond-forming events in the catalytic ODC of cyclohexane, we conducted the reaction of cyclohexyl benzoate and cyclohexene (separately) (Scheme 6,A) with tBuOOtBu and the copper catalyst. These two reactions reveal whether formation of the alkene occurs before or after formation of the C−O bond. The reaction of cyclohexyl benzoate, benzoic acid, and tBuOOtBu with 2.5 mol% of 1-O 2 CPh at 100°C for 24 h did not form cyclohex-2-en-1-yl benzoate. Instead, this reaction generated methyl benzoate (16%). The detection of methyl benzoate indicates that tertbutoxy radical was generated from the reaction of tBuOOtBu with copper, but that this radical reacts more slowly with the cyclohexyl benzoate than it undergoes β-methyl scission to generate the methyl radical (which reacts with the copper− benzoate complex to form methyl benzoate).In contrast to the reaction of cyclohexyl benzoate, the reaction of cyclohexene with benzoic acid and tBuOOtBu in the presence of 2.5 mol% 1-O 2 CPh generated the allylic ester. This reaction formed cyclohex-2-en-1-yl benzoate in 58% yield and methyl benzoate in 27% yield after 24 h at 100°C (Scheme 6B). Moreover, the reaction of cyclohexane with tBuOOtBu in the presence of 1 mol% of 1-O 2 CPh (based on tBuOOtBu) in benzene-d 6 at 100°C for 20 h (Scheme 6C) formed cyclohexene in 12% yield, with respect to cyclohexane, as determined by 1 H NMR spectroscopy. These results clearly indicate that ODC of cyclohexane to cyclohex-2-en-1-yl benzoate proceeds by initial conversion of the cycloalkane to The mechanism for the initial conversion of cyclohexane to cyclohexene likely proceeds by abstraction of a hydrogen atom from cyclohexane by a tert-butoxy radical to generate a cyclohexyl radical, which undergoes oxidation to the alkene. The oxidation of alkyl radicals to olefins by copper−peroxide systems has been studied by and Walling.Their studies imply that oxidation of the cyclohexyl radical formed in the current system likely generates cyclohexyl cation, which undergoes deprotonation to form the alkene. The deprotonation could occur by the anionic Cu(I) complex [(BPI)Cu-(O 2 CPh)] − (1-O 2 CPh*) (Scheme 7). The yield of allylic ester would then be a function of the relative rate of oxidation of the alkyl radical versus reaction of the alkyl radical with the copper carboxylate.
After formation of cyclohexene, oxidation at the allylic position to form cyclohex-2-en-1-yl benzoate would occur through the mechanism of the Kharasch−Sosnovsky reaction. In this pathway, the allylic hydrogen is abstracted by a tert-butoxy radical, and the resulting allylic radical reacts with the copper carboxylate to form the allylic ester.
To detect for the possible formation of 1,3-cyclohexadiene or benzene from cyclohexene through a series of steps involving abstraction of the allylic hydrogen, oxidation of the allyl radical, and deprotonation of the allyl cation, the catalytic reaction of benzoic acid, cyclohexane, and tBuOOtBu in the presence of 1-PPh 3 (2.5 mol%) in d 6 -benzene (or d 3 -MeCN) at 100°C for 16 h was monitored by 1 H NMR spectroscopy. The result of the reaction revealed only cyclohex-2-en-1-yl benzoate and methyl benzoate as products in 40% and 14% yields, with respect to benzoic acid. Cyclohexadiene and benzene that could result from dehydrogenation of cyclohexene were not observed.
The stoichiometric and catalytic ODC of benzoic acid with cyclohexane forms methyl benzoate as the major side product. The observation of this product is consistent with the intermediacy of tert-butoxy radical. β-Methyl scission of a tertbutoxy radical is known to produce a methyl radical, and this radical would react with the resting-state 1-O 2 CPh to give methyl benzoate.
To evaluate the potential generation of tert-butoxy radical in the system, the standard catalytic reaction of benzoic acid in C 6 D 6 was performed in the absence of cyclohexane (Scheme 8A). Without a source of an alkyl radical besides the one formed by β-methyl scission of - OtBu, the reaction produced a quantitative yield of methyl benzoate and acetone (based on benzoic acid as limiting reagent). This high-yield formation of methyl benzoate from benzoic acid and tBuOOtBu in the presence of 1-PPh 3 further supports the intermediacy of a transient tert-butoxy radical in the catalytic reaction.As a final test of the potential intermediacy of tert-butoxy radical in the catalytic process, we conducted reactions in the presence of diphenylmethanol, a known trap for tert-butoxy radical,and in the presence of 9,10-dihydroanthracene, which forms anthracene via hydrogen atom abstraction by alkoxy radicals. The catalytic reaction of cyclohexane, benzoic acid, and diphenylmethanol in the presence of 1-O 2 CPh produced methyl benzoate (18%), benzophenone, and diphenylmethanol in a ratio of 1:10:7.3 (Scheme 8B). The same reaction between cyclohexane and benzoic acid at 100°C for 24 h in the presence of 9,10-dihydroanthracene produced anthracene as the exclusive product from the hydrocarbon reactants (Scheme 8C). The formation of benzophenone and anthracene is consistent with H-atom abstraction of the methine C−H bond of diphenylmethanol and a methylene C−H bond of dihydroanthracene by tert-butoxy radical to produce the organic products. The detection of methyl benzoate as an additional product, again, is consistent with β-methyl scission of a tertbutoxy radical under the catalytic conditions.
To assess the potential intermediacy of a cyclohexyl radical, the catalytic reaction of cyclohexane with benzoic acid was performed in the presence of CBr 4 . This reaction exclusively formed bromocyclohexane (Scheme 9). The observation of bromocyclohexane is further consistent with the formation of cyclohexyl radical by abstraction of a hydrogen atom from cyclohexane by a tert-butoxy radical in the catalytic reaction. In this case, the radical reacts with CBr 4 to form the alkyl bromide.In addition, a comparison of the rates of the catalytic reaction of octanoic acid with cyclohexane and cyclohexene revealed that the conversion of cyclohexene to cyclohex-2-en-1-yl octanoate is faster than the conversion of cyclohexane.After 1 h, the reaction of octanoic acid and cyclohexene cleanly produced 40% of cyclohex-2-en-1-yl octanoate, whereas the reaction of octanoic acid with cyclohexane produced only 4% of cyclohex-2-en-1-yl octanoate and 2% of methyl octanoate. This result clearly indicates that abstraction of the C−H bond from cyclohexane, not from cyclohexene, is the turnover-limiting step in the catalytic ODC.
6. Effects of the Electronic Properties of Copper− Benzoate and Copper−Amidate Complexes on the Reaction of Alkyl Radicals. The roles of copper in the catalytic ODC of cyclohexane are closely related to those of copper in the catalytic amidation of cyclohexane we reported recently.However, the two reactions form products containing different hydrocarbyl groups (alkyl vs allylic), and the difference between these groups likely stems from a difference in relative rates for reaction of the alkyl radical with the copper−benzoate and copper−amidate complexes. The alkyl radical can undergo electron transfer, or it can combine with a ligand at copper to form a product containing a new carbon−heteroatom bond (Scheme 11). Apparently, oxidation of the alkyl radical by the copper−benzoate is faster than reaction of the alkyl radical with the benzoate ligand, whereas oxidation of the alkyl radical by the copper−amidate is slower than reaction of the alkyl radical with the amidate ligand (Scheme 11).
This difference in relative rates could arise from a difference in redox potential of the benzoate and amidate complexes. A Cu(II)−benzoate complex, presumably, is less electron-rich than a Cu(II)−amidate complex. Therefore, the former complex could oxidize the alkyl radical to an alkyl cation faster than the latter complex. Alternatively, the difference in relative rates could arise from differences in the rates of reaction of alkyl radicals with the Cu(II)−carboxylate and Cu(II)−amidate complexes.
To reveal the origin of the difference in formation of alkyl and allyl products with amide and carboxylic acid reagents we conducted a series of reactions in which a methyl radical is generated in the presence of a copper−carboxylate, amidate, or imidate complex. Third, a competition reaction was performed with 1-O 2 Chept and 1-NHC(O)hept in the presence of cyclohexane and tBuOOtBu at 100°C in benzene (Scheme 13). Unlike a methyl radical, a cyclohexyl radical can form a carbon− heteroatom bond or convert to cyclohexene; carboxylation of the resulting alkene then forms an allylic ester. The reaction of the two copper complexes formed N-cyclohexyloctanamide in 92% yield and the allylic ester product cyclohex-2-en-1-yl benzoate in 30% yield. This result is consistent with faster reaction of an alkyl radical with a copper−amidate than with a copper−benzoate, but the origin of the absence of product indicates that the reaction of the alkyl radical with 1-NHC(O)Ph is faster than that with 1-phth. This trend is consistent with faster reaction of an alkyl radical with the more electron-rich anionic ligand on copper.
To gain more systematic data conerning the electronic effects on the rates of reactions of alkyl radicals with the copper complexes, we studied reactions with a series of substituted benzoate complexes. Reactions of a methyl radical generated from tBuOOtBu with a mixture of [(BPI)Cu(O 2 C[C 6 H 4 -4-OMe]) (1-OMe) and [(BPI)Cu(O 2 C[C 6 H 4 -4-CN]) (1-CN) were conducted at 100°C in benzene (Scheme 15A). The result showed that methyl radical reacted faster with the electron-rich 1-OMe to produce the corresponding methyl 4methoxybenzoate than with the more electron-poor 1-CN to produce methyl 4-cyanobenzoate. For example, at 24 h the reaction produced methyl 4-methoxybenzoate in 76% yield and methyl 4-cyanobenzoate in 10% yield.
To gain analogous information on the reaction of an alkyl radical with copper−amidates, the analogous experiment was . The reaction of tBuOOtBu with these complexes at 100°C in benzene (Scheme 15B) showed that the methyl radical reacts faster with the more electron-rich 1-NHOMe than with the more electron-poor 1-NHCF 3 to produce the corresponding product of N-methyl-4-methoxybenzamide (53%) at 24 h. The results of these competition reactions clearly demonstrate that alkyl radicals react faster with the more electron-rich copper−benzoate and amidate complexes than with the more electron-deficient copper−benzoate and amidate complexes to form the corresponding N-alkyl and O-alkyl products, respectively. the combination of 2,4-Me 2 and 2,6-Me 2 formed a higher yield of cyclohex-2-en-1-yl 2,4-dimethylbenzoate than cyclohex-2-en-1-yl 2,6-dimethylbenzoate (Scheme 16A′). The reaction with a combination of 2,4-Me 2 and 3,4-Me 2 formed the corresponding cyclohex-2-en-1-yl 3,4-dimethylbenzoate in higher yields than it formed cyclohex-2-en-1-yl 2,4-dimethylbenzoate (Scheme 16B′). Lastly, the reaction with the combination of 3,4-Me 2 and 2,6-Me 2 produced higher yield of cyclohex-2-en-1yl 3,4-dimethylbenzoate (40%) than of cyclohex-2-en-1-yl 2,6dimethylbenzoate (20%) at 24 h (Scheme 16C′). Similar results were obtained from competition reactions between cyclohexane and tBuOOtBu with a series of dimethoxybenzoate−Cu(II) complexes.The difference in rates of reaction of the carboxylate complexes as a function of the steric properties could result from the effect on the conformation of these copper−benzoate complexes and overlap of the aryl π system with the carbonyl group. The molecular structures of 2,6-OMe 2 and 3,4-OMe 2 were determined by X-ray diffraction [fig_ref] Figure 1: Molecular structures of [ [/fig_ref]. The aryl ring of the 2,6-dimethoxybenzoate is nearly orthogonal to the carbonyl group with a torsion angle of −76°, whereas the benzene ring of the 3,4-dimethoxybenzoate lies nearly in plane with the carbonyl group with a torsion angle of −163°. Thus, the aryl ring of 3,4-OMe 2 has more orbital overlap between the carbonyl and benzene π-systems than does that of the 2,6dimethoxylbenzoate. This distortion from planarity of the 2,6disubstituted benzoate makes it less nucleophilic (Scheme 17).This reduced nucleophilicity is then responsible for the difference in the reactivity of copper−benzoate with the allylic radical to generate allylic benzoate products.
8. Observation of Oxidative Dehydrogenative Amination: Effect of Ligand and Nitrogen Source. The step that distinguishes the reactivity of copper-catalyzed ODC of cyclohexane and copper-catalyzed amidation of cyclohexane is the reaction of the transient cyclohexyl radical with a copper− benzoate versus a copper−amidate or copper−imidate intermediate. Specifically, alkyl radicals react faster with copper−amidate and imidate complexes than they do with copper−benzoates. Moreover, more electron-rich copper− amidate and benzoate complexes react faster with alkyl radicals than more electron-deficient copper−amidate and benzoate complexes. Lastly, the slower rate of reactions of alkyl radicals with copper−benzoate allows electron transfer to occur faster than ligand transfer, unless the alkyl radical is not able to form an alkene (i.e., a methyl radical). When the alkyl radical cannot form an alkene, ligand transfer occurs to form an O-alkyl product. Based on these hypotheses, we investigated coppercatalyzed oxidative dehydrogenative amination (ODA) of cyclohexane with electron-deficient amides. More electrondeficient copper−amidate and imidate complexes would undergo slower reactions with an alkyl radical and faster electron transfer.
To this end, we conducted catalytic reactions of phthalimide and cyclohexane (10 equiv) in the presence of 2.5 mol% of [(phen)Cu(phth)], 1-PPh 3 , and [Cu(phth)] (Scheme 18).The reactions in the presence of 1-PPh 3 and [Cu(phth)] produced 18−25% of cyclohex-2-en-1-yl pthalimidate and roughly 70% of N-cyclohexylphthalimide. The reactions conducted with 1-PPh 3 and [Cu(phth)] apparently occur by competitive ligand transfer and electron transfer to form a mixture of N-cyclohexylphthalimide and cyclohex-2-en-1-yl pthalimidate, respectively. Conversely, the reaction catalyzed by [(phen)Cu(phth)] formed N-cyclohexylphthalimide (63%) and N-methylphthalimide (19%), and no cyclohex-2-en-1-yl pthalimidate, as determined by gas chromatography, mass spectrometry, and 1 H NMR spectroscopy.
The catalytic reaction in the presence of [(phen)Cu(phth)] exclusively produced N-cyclohexylphthalimide and N-methylphthalimide, presumably from reaction of the methyl and cyclohexyl radicals with the resting-state species [(phen)Cu-(phth) 2 ].The preference of [(phen)Cu(phth) 2 ] to react with cyclohexyl radical by ligand-transfer is presumably because [(phen)Cu(phth) 2 ] is more electron-rich than the complexes 1-phth and [Cu(phth) 2 ] n ; thus, reaction of the alkyl radical with [(phen)Cu(phth) 2 ] occurs faster than electron transfer. Previously, we reported the copper-catalyzed oxidative coupling of p-toluenesulfonamide with cyclohexane to generate the corresponding N-cyclohexyl-p-toluenesulfonamide.To achieve ODA of cyclohexane, we reasoned that replacing ptoluenesulfonamide with the more electron-deficient 4-CF 3benzenesulfonamide should decrease the rate of reaction of the alkyl radical with the ligand and increase the rate of oxidation of the alkyl radical. The reaction of 4-CF 3 -benzenesulfonamide with cyclohexane and tBuOOtBu in the presence of 2.5 mol% of 1-PPh 3 in acetonitrile produced N-cyclohexyl-4-CF 3benzenesulfonamide (50%) and N-(cyclohex-2-en-1-yl)-4-CF 3benzenesulfonamide (38%) (Scheme 19). This result demonstrates that electron-deficient sulfonamides can form substantial amount of product from ODA of cyclohexane by the electrontransfer pathway.
The selectivity of the reaction of cyclohexane with CF 3 -4benzenesulfonamide in the presence of 1-PPh 3 depended on solvent and supporting ligand (Scheme 19). The reaction in acetonitrile formed the mixture of alkyl and allylic amides just described, but the reaction in benzene generated exclusively cyclohexyl-4-CF 3 -benzenesulfonamide (57%). The distribution of products from the reaction of CF 3 -4-benzenesulfonamide and cyclohexane in the presence of "ligandless" copper generated from [Cu(Mes)] also depended on solvent, but with the opposite trend. The products of ODA were observed in benzene, but not in acetonitrile (Scheme 19). Like the reactions of phthalimide, the reactions of CF 3 -4-benzenesulfonamide with cyclohexane in the presence of phen-ligated copper provided only the N-alkyl product (Scheme 19). This result again highlights the effect of the electronic properties of the supporting ligand on the relative rates to form N-alkyl and Nallyl products. This effect of ligand (i.e., BPI vs phen) on this selectivity reflects an opportunity to design ligands that favor ODA of unactivated alkanes.
9. Proposed Mechanism for ODC of Cyclohexane. A proposed mechanism for the catalytic ODC of cyclohexane to cyclohex-2-en-1-yl benzoate is presented in Scheme 20. In this pathway, catalysis is initiated by the decomposition of tBuOOtBu by 1-PPh 3 to produce a tert-butoxy radical and [(BPI)CuOtBu], which rapidly reacts with benzoic acid to form 1-O 2 CPh and tBuOH. The tert-butoxy radical can undergo reversible, secondary internal return to regenerate tBuOOtBu or abstract a hydrogen atom from cyclohexane to generate a cyclohexyl radical. To form the unsaturated product, 1-O 2 CPh would oxidize the alkyl radical by one electron to form a carbocation, and the carbocation would undergo deprotonation by an anionic Cu(I) species [(BPI)Cu(O 2 CPh)] − to give cyclohexene and benzoic acid. The resulting cyclohexene would then undergo a second C−H abstraction by a tert-butoxy radical to give an allylic radical that reacts with 1-O 2 CPh to release the allylic ester product and regenerate a (BPI)Cu(I) species to complete the catalytic cycle. In a side reaction, the tert-butoxy radical would decompose to a methyl radical and acetone. The methyl radical would then combine with 1-O 2 CPh to produce methyl benzoate and a (BPI)Cu(I) species. The catalytic cycle for ODA of cyclohexane to form N-allyl products is presumably analogous to that of the catalytic ODC. This catalytic cycle would contain a copper−amidate resting state, based on mechanistic investigations of a related copper-catalyzed amidation of unactivated alkanes.
## ■ summary and conclusions
In summary, we have described a copper-catalyzed oxidative dehydrogenative carboxylation (ODC) of unactivated alkanes with a variety of benzoic acids to produce the corresponding allylic ester products. A measurement of kinetic isotope effects showed that the turnover-limiting step is C−H bond cleavage, and experiments to trap radical intermediates revealed that a transient tert-butoxy radical cleaves the C−H bond of the alkane to generate an alkyl radical. Reactions of alkyl radicals with a combination of Cu(II)−amidates and Cu(II)−benzoates revealed that the alkyl radical reacts faster with a Cu(II)− amidate than with a Cu(II)−benzoate to form N-alkyl products. Additional mechanistic investigations indicated that the electronic properties of the Cu(II)−X (X = amidate, benzoate) resting state contributes to the partitioning of the alkyl radical between ligand transfer to form the alkyl−heteroatom bond and electron transfer to oxidize the alkyl radical to an olefin, followed by oxidative carboxylation to produce an allylic ester. The reaction of the alkyl radical with a Cu(II)−amidate versus a Cu(II)−benzoate is the step that distinguishes coppercatalyzed amidation and copper-catalyzed ODC of the alkane.
This insight into the mechanism of ODC of cyclohexane led to preliminary observations of copper-catalyzed oxidative dehydrogenative amination of cyclohexane with electrondeficient nitrogen sources (i.e., phthalimide and an electrondeficient sulfonamide) to form N-allyl products. Current efforts are underway to discover reaction conditions to suppress ligand transfer and favor electron transfer to achieve high selectivity for a copper-catalyzed ODA of unactivated alkanes.
[fig] 7: Steric Effects on the Reaction of Alkyl Radicals with Copper−Benzoates. To elucidate the steric effect of aromatic ring of the carboxylate ligand in copper−benzoate complexes on the reactivity, we performed reactions of cyclohexane and tBuOOtBu in the presence of a series of copper−benzoates containing methyl groups in the ortho, meta, and para positions: [(BPI)CuX], with X = 2,6-dimethylbenzoate (2,6-Me 2 ), 2,4-dimethylbenzoate (2,4-Me 2 ), and 3,4-dimethylbenzoate (3,4-Me 2 )). The results of these competition experiments are summarized in Scheme 16. doi.org/10.1021/ja510093x | J. Am. Chem. Soc. 2014, 136, 17292−17301 [/fig]
[fig] Figure 1: Molecular structures of [(BPI)Cu(3,4-OMe 2 -C 6 H 4 )] (3,4-OMe 2 ) (top) and [(BPI)Cu(2,6-OMe 2 -C 6 H 4 )] (2,6-OMe 2 ) (bottom) are shown with 50% thermal ellipsoid. Hydrogen atoms are omitted for clarity. Selected bond lengths (Å) and angles (°) for 3,4-OMe 2 : Cu1−N1 = 2.019(2); Cu1−N3 = 1.892(2); Cu1−O1 = 1.9443(18); N1−Cu1−N3 = 90.55(9); N3−Cu1−O1 = 168.23(9). Selected bond lengths (Å) and angles (°) for 2,6-OMe 2 : Cu1′−N1′ = 2−0074(17); Cu1′−N3′ = 1.9017(17); Cu1′−O1′ = 1.9721(14); N1′−Cu1′−N3′ = 90.94(7); N3′−Cu1′−O1′ = 167.63(7). [/fig]
[table] Table 1: Reaction Development of Catalytic ODC of Cyclohexane a [/table]
|
Oleogranulomatous Mastitis: A Topical Subject
[bib_ref] Paraffinoma of the breast, Tinckler [/bib_ref] [bib_ref] Death following injection of paraffin into the breast, Pang [/bib_ref] [bib_ref] Complications from injectable materials used for breast augmentation, Peters [/bib_ref]
## Observation
A 39-year-old patient consulted in 2011 in the University Hospital of Strasbourg for tense and painful breasts after an injection of petrolatum in Chechnya 2 years ago to increase her breast size. The patient gradually developed bilateral edema and erythema. In addition, the patient described fever episodes evolving repeatedly and associated with a major weakness. The clinical breast examination found erythematosus plates in the upper and internal regions without ulceration or fistula consequences [fig_ref] Figure 1: a, preoperative clinical appearance of both breasts [/fig_ref].
The mammogram showed multiple rounded masses of fat density, confluent and disseminated over dense breasts [fig_ref] Figure 1: a, preoperative clinical appearance of both breasts [/fig_ref]. At the ultrasound, there were multiple thin-walled cystic formations of anechoic content. The fine-needle aspiration confirmed the presence of an oily liquid. The Aspiration followed by liposuction was tried to eliminate petrolatum, but without any result, the breasts being too fibrous and petrolatum too diffuse. After prolonged anti-inflammatory and antibiotic treatment, the patient agreed to a bilateral mastectomy. Given the extent of resection, immediate breast reconstruction was impossible without prior tissue expansion. Abdominal expansion by saline inflatable implants was performed [fig_ref] Figure 1: a, preoperative clinical appearance of both breasts [/fig_ref]. The prostheses volume was 1 L per side. Inflation was performed once a week. The prostheses were overinflated and left in place for 3 and a half months. The second intervention consisted of a bilateral mastectomy without preservation of the nipple-areolar plates, which were too adherent and inflammatory. The gland and the pectoralis major muscle were infiltrated, fibrous, and altered. The removed mammary glands weighed 930-950 g. Histological examination of the breasts confirmed the diagnosis of oleogranulomatous mastitis (OM). Immediate breast reconstruction was obtained by expansion and advancement of abdominal flap. To avoid excessive skin tension, inflatable breast expanders were introduced at the retro pectoral level. Three months later, expanders were removed and replaced by permanent silicone protheses with an anatomical shape of 520 g.
# Disclosure
The nipple-areolar complex was reconstructed later by transplantation of the earlobe and tattoo. The postoperative course was uneventful. Today, the patient has no longer pain or inflammation or fever. The cosmetic result is satisfactory, despite a persistent erythematous area in the upper pole of the breasts [fig_ref] Figure 1: a, preoperative clinical appearance of both breasts [/fig_ref].
# Discussion
Our observation is consistent with a recent injection of paraffin, as evidenced by history and radio-
## Akladios et al. - oleogranulomatous mastitis
graphic appearance of the breasts. Indeed, recent injections result in multiple small lights scattered throughout the mammary gland, although limited and noncalcified. [bib_ref] Direct injection of paraffin into the breast: mammographic, sonographic, and MRI features..., Erguvan-Dogan [/bib_ref] In contrast, former injections are characterized by dense masses with semicircular or "soap-bubble-like" calcifications. [bib_ref] Altered breast: paraffin injection with development of paraffinomas, Sinclair [/bib_ref] Considering the fact that paraffin injections were recent, we tried to avoid in the first-place mastectomy, using liposuction of oily masses. But, because of the inflammatory and hypervascular nature of the tissue on the one hand and of the small oily masses and their diffuse nature on the other hand, the attempt was unsuccessful.
For many authors, mastectomy is the only curative treatment. In Alagaratnam and Ng's 8 series of 43 patients, where the injections were recent or very old (realized between 3 and 41 years ago), 30 mastectomies had to be performed because of the importance of the masses and chronic ulcerations. Ho et al 9 reported 8 patients cared for OM (injections performed between 11 and 30 years ago). In one case, the injection of paraffin was followed by destructive ulceration of the anterior chest wall, with damage to both pectoral muscles and the ribs. The author also concluded that the only possible treatment for OM was mastectomy, which might be associated with first-stage reconstruction by rectus abdominal musculocutaneous flap. Ortiz-Monasterio and Trigos 10 propose second-stage breast reconstruction due to the risk of bleeding and infections.
In our case, mastectomy with conservation of the skin would have resulted in a thin skin covering the implant, with increased risk of exposure of the prosthesis. To reduce the risk of complications, we performed an original method for abdominal expansion to provide healthy skin to the implant. Reconstruction is cosmetically satisfactory, but a rash persists in the upper quadrants.
# Conclusion
OM is a challenge for surgeons. Therapeutic problems are indeed encountered because patients attach great importance to the cosmetic result, and paraffin injections have been devised for aesthetic reasons. Patients also pose psychological difficulties because they do not understand the seriousness of the problem and do not accept mutilation for a benign lesion. No treatment is able to modify the course of the OM, which may be accompanied by skin ulceration or parietal necrosis. We have shown that liposuction of paraffin is impossible. Mastectomy is the only treatment, possibly associated with immediate breast reconstruction or delayed. The injection of paraffin or petrolatum leads to real dilapidations, and for that reason must be firmly prohibited.
## Dr. cherif akladios, md, phd
Strasbourg University Hospital Departement of Obstetrics and Gynecology 1 Avenue Molière 67000 Strasbourg. France Email: [email protected]
[fig] Figure 1: a, preoperative clinical appearance of both breasts. Bilateral erythematous plates of the upper and internal regions are found without ulcers or fistula sequelae. Breasts are very tense. B, Mammographic appearance of both breasts viewed from the front showing multiple small lights scattered throughout the mammary gland, although limited and noncalcified with a predilection for retro grandular and prepectoral regions. C, an abdominal expansion by saline inflatable implants was performed (prostheses volume: 1 L per side) and left in place for 3 and a half months. D, Front final aspect and after reconstruction of nipple-areolar plates. there remains an erythematous area in the upper pole of the breast. [/fig]
|
Mixed Mycobacterium Avium-Intracellulare and Serratia Marcescens Cellulitis of the Breast in an HIV-Negative Patient with Breast Cancer: A Case Report
Mycobacterium avium-intracellulare (MAI) causes pulmonary infection in patients with chronic lung diseases or severe T-cell deficiency. Cutaneous manifestations caused by MAI are rare and the few cases reported describe mostly patients with hematologic malignancies who were treated with highly immunosuppressive agents. Herein, we report a case of a breast cancer survivor who developed chronic breast cellulitis due to MAI, following localized breast cancer treatment.
# Introduction
Mycobacterium avium-intracellulare (MAI), the most common nontuberculous mycobacterium in the world, is an opportunistic pathogen that is notorious for causing pulmonary infections in patients with chronic lung disease and disseminated disease in patients with acquired immune deficiency syndrome [bib_ref] Mycobacterium avium-intracellulare: cutaneous presentations of disseminated disease, Friedman [/bib_ref] [bib_ref] Mycobacterium avium complex in patients with HIV infection in the era of..., Karakousis [/bib_ref]. Skin and soft tissue infection (SSTI) secondary to MAI is a rare entity in the absence of mycobacterial bloodstream infection (BSI). To our knowledge, the following report is the first description of MAI-related SSTI in a patient with a solid tumor [bib_ref] Mycobacterium avium-intracellulare: cutaneous presentations of disseminated disease, Friedman [/bib_ref] [bib_ref] Primary cutaneous infection by Mycobacterium avium: a case report and literature review, Aboutalebi [/bib_ref] [bib_ref] Disseminated cutaneous mycobacterium avium complex in a person with AIDS, Endly [/bib_ref] [bib_ref] Cutaneous manifestations of mycobacterial infection in patients with AIDS, Inwald [/bib_ref] [bib_ref] Successful treatment of refractory disseminated Mycobacterium avium complex infection with the addition..., Nannini [/bib_ref]. MAI soft tissue manifestations in patients with solid tumors have not been described before. To our knowledge, this is the first patient case reported of MAI soft tissue infection in a solid tumor. Informed consent was obtained from the patient for this study.
## Case presentation
This is a 71-year-old female who was recently diagnosed with a 1.2 cm left breast ductal carcinoma in situ (estrogen receptor weakly positive/progesterone receptor negative). She underwent a left segmental mastectomy with sentinel lymph node biopsy followed by adjuvant radiation therapy and placement of a catheter evacuation device. On her three-month followup visit, the patient complained of left breast tenderness, erythema, and a foul-smelling nipple discharge. Her symptoms had been ongoing for two weeks during which she also experienced intermittent low grade fevers and malaise. Physical examination was notable for a markedly tender left breast with associated erythema, induration (11 cm x 9 cm), and skin thickening overlying the site of radiation. Ultrasound of the breast revealed an underlying subcutaneous seroma measuring 4 cm × 8 cm × 5.2 cm. Subsequent aspiration of the cavity recovered 40 ml of Open Access Case purulent fluid from which Serratia marcescens was eventually isolated. Of interest, the patient had an S. marcescens urinary tract infection six months ago. The patient was started on a sevenday regimen of ciprofloxacin after which she continued to experience purulent nipple discharge and diffuse breast erythema. Her course was further complicated by the development of a diffuse pruritic maculopapular rash that prompted the discontinuation of ciprofloxacin.
Four weeks after the aspiration, cultures also grew MAI. At that point, there was a modest improvement of the breast cellulitis, but persistent drainage of purulent fluid . Human immunodeficiency virus (HIV) serology results were negative. She was placed on an oral combination regimen, consisted of rifabutin (300 mg/day), clarithromycin (500 mg twice/day) and doxycycline (100 mg twice/day). Doxycycline, although an unconventional agent for the treatment of S. marcescens, was used instead of quinolones due to the patient's recent history of hypersensitivity to ciprofloxacin, as it has in vitro activity against S. marcescens [bib_ref] Antibiotic susceptibility of Serratia marcescens and Serratia liquefaciens, Traub [/bib_ref]. The patient's seroma was aspirated again, draining 10 ml of yellow fluid that was sent for cultures and returned negative . In addition to rifabutin and clarithromycin, the patient was placed on intravenous ertapenem (1000 mg/day), followed by a course of oral cefixime (400 mg/day), and she received local wound care with topical silver nitrate, leading to a complete resolution of the erythema and drainage. Follow-up for the next 12 months showed progressive healing, and follow-up cultures did not grow MAI or Serratia .
## Figure 1: breast seroma infected with mycobacterium aviumintracellulare and serratia marcescens
Lateral and close-up views of the skin manifestations attributed to Mycobacterium aviumintracellulare and Serratia marcescens at: A. Four weeks of treatment. B. 10 weeks of treatment. C. 12 months after the end of treatment.
# Discussion
SSTIs secondary to MAI infection are thought to occur via hematogenous seeding in the context of an underlying trauma. Some of the well-established cutaneous manifestations include scaling papules, subcutaneous/ulcerative nodules, granulomatous plaques, and abscesses [bib_ref] Primary cutaneous infection by Mycobacterium avium: a case report and literature review, Aboutalebi [/bib_ref] [bib_ref] Chronic granulomatous Mycobacterium avium skin pseudotumour, Nassar [/bib_ref]. Presumptively, in our patient, the recent breast surgery and radioadjuvant therapy served as the nidus for seeding S. marcescens and MAI. The few reports existing in the literature of cutaneous MAI infection in non-HIV infected cancer patients describe mostly leukemia patients who were treated with highly immunosuppressive agents [bib_ref] Mycobacterium avium-intracellulare: cutaneous presentations of disseminated disease, Friedman [/bib_ref] [bib_ref] Successful treatment of refractory disseminated Mycobacterium avium complex infection with the addition..., Nannini [/bib_ref].
A combination of surgical debridement and a three-drug regimen with macrolides (clarithromycin or azithromycin), ethambutol, and a rifamycin (rifampin or rifabutin) may be extrapolated from studies of HIV-infected patients. [bib_ref] Mycobacterium avium complex in patients with HIV infection in the era of..., Karakousis [/bib_ref] [bib_ref] An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases, Griffith [/bib_ref].
# Conclusions
While an optimal treatment for an extra-pulmonary MAI infection isolated to the soft tissue has not yet been defined, a combination of surgical debridement and a three-drug regimen with macrolides, ethambutol, and rifamycin may provide a successful outcome and should be considered.
# Additional information disclosures
Human subjects: Institutional review board at MD Anderson Cancer Center issued approval PA 14-0829.
|
Understanding how the immune system facilitates or controls HIV-1 disease progression has important implications for the design of effective interventions. We report that although B-cell dysregulations associated with HIV-1 disease progression are accompanied by an overall decrease in the percentage of total blood B-cells, we observe an increase in relative frequencies of cells presenting characteristics of both transitional immature and first-line marginal zone (MZ) B-cell populations, we designated as precursor MZ-like B-cells. B-cells with similar attributes have been associated with IL-10 expression and ''regulatory'' potential. As such, the relative frequencies of precursor MZ-like B-cells expressing IL-10 are increased in the blood of viremic HIV-1-infected individuals when compared to HIV-negative subjects. Importantly, in aviremic HIV-1 Elite-Controllers (EC), we found unaltered relative percentages of precursor MZ-like B-cells which presented normal IL-10 expression patterns. Furthermore, EC had increased relative frequencies of blood MZ-like B-cells expressing LTa. Thus in contrast to viremic HIV-1-infected individuals, EC present MZ-like B-cell populations which IL-10 and LT-a expression profiles may favour homeostasis of immune responses and lymphoid microenvironments. Citation: Chagnon-Choquet J, Fontaine J, Poudrier J, Roger M, et al. (2014) IL-10 and Lymphotoxin-a Expression Profiles within Marginal Zone-Like B-Cell Populations Are Associated with Control of HIV-1 Disease Progression. PLoS ONE 9(7): e101949.
# Introduction
It is well known that the contribution of the B-cell compartment to effective viral control is impeded in the vast majority of HIV-1infected individuals. Indeed, B-cell dysregulations are observed early, persist throughout the course of infection, and are not fully restored by therapy. These B-cell alterations favour the overall inflammatory burden and often lead to autoimmune manifestations and malignancies [bib_ref] B cells in HIV infection and disease, Moir [/bib_ref] [bib_ref] Influence of Dendritic cells on B-cell Responses during HIV Infection, Poudrier [/bib_ref].
We have previously shown that HIV-transgenic (Tg)-mice, which develop a negative factor (Nef)-dependant AIDS-like disease [bib_ref] Nef harbors a major determinant of pathogenicity for an AIDS-like disease induced..., Hanna [/bib_ref] , present B-cell dysregulations that are similar to those reported for HIV-infected individuals [bib_ref] The AIDS disease of CD4C/HIV transgenic mice shows impaired germinal centers and..., Poudrier [/bib_ref]. Strikingly, these animals present an enlarged splenic marginal zone (MZ), in which accumulated myeloid dendritic cells (mDCs) likely contribute to MZ expansion, polyclonal B-cell activation and breakage of tolerance through delivery of excessive signals such as B lymphocyte stimulator (BLyS/BAFF) [bib_ref] The AIDS disease of CD4C/HIV transgenic mice shows impaired germinal centers and..., Poudrier [/bib_ref] [bib_ref] The AIDS-like disease of CD4C/human immunodeficiency virus transgenic mice is associated with..., Poudrier [/bib_ref]. A similar B-cell profile was reported for BLyS/BAFF-Tg [bib_ref] B-cell stage and context-dependent requirements for survival signals from BAFF and the..., Mackay [/bib_ref] and autoimmune-regulatory-(AIRE)-deficient mice, in which BLyS/BAFF is elevated in serum and overexpressed by DCs [bib_ref] AIRE-deficient mice develop haematopoietic irregularities and marginal zone B-cell lymphoma, Hä Ssler [/bib_ref] [bib_ref] AIRE regulates T-cell-independent B-cell responses through BAFF, Lindh [/bib_ref]. Accordingly, DCs play a pivotal role in regulating B-cell development, activation and survival mainly through production of growth factors such as BLyS/BAFF [bib_ref] The who, how and where of antigen presentation to B cells, Batista [/bib_ref] [bib_ref] Dendritic cells, B cells and the regulation of antibody synthesis, Macpherson [/bib_ref] [bib_ref] Innate control of B cell responses, Cerutti [/bib_ref] , known to highly influence the transitional immature (TI) and MZ B-cell pools [bib_ref] Marginal zone B cells: virtues of innate-like antibody-producing lymphocytes, Cerutti [/bib_ref] [bib_ref] Human Marginal Zone B cells, Weill [/bib_ref].
MZ B-cells constitute early first-line defence against invading pathogens. In humans, they likely constitute a heterogeneous niche that is not restricted to the spleen, as they have been found in blood, lymphoid organs and mucosal-associated structures [bib_ref] Marginal zone B cells: virtues of innate-like antibody-producing lymphocytes, Cerutti [/bib_ref] [bib_ref] Human Marginal Zone B cells, Weill [/bib_ref]. Reminiscent of what had been observed in HIV-Tg mice [bib_ref] The AIDS disease of CD4C/HIV transgenic mice shows impaired germinal centers and..., Poudrier [/bib_ref] , we found that B-cell dysregulations in HIV-1-infected rapid and classic progressors were accompanied by increased relative frequencies within total blood B-cells of a population presenting features shared by both TI and MZ B-cells, which we designated ''precursor'' MZ-like B-cells. Concomitantly, these individuals presented increased BLyS/BAFF levels in blood and on surface of blood mDCs [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref]. In contrast, relative frequencies of precursor MZ-like B-cells as well as BLyS/BAFF expression status were unaltered in HIV-1 Elite-Controllers (EC) [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref]. Instead, decreased frequencies of more ''mature'' MZ-like B-cells were observed in the blood B-cell compartment of EC. B-cells are involved in regulating the development, proliferation and maintenance of CD4 + T-cell populations, through both contact and/or cytokine mediated effector and regulatory functions [bib_ref] Effector and regulatory B cells: modulators of CD4+ T cell immunity, Lund [/bib_ref]. Regulatory ''Breg'' potential has not yet been attributed to a particular B-cell population and relies rather on IL-10 expression/production and function. Both precursor and mature B-cells with MZ attributes as well as TI and memory populations have been ascribed such Breg potential [bib_ref] Immune Regulatory function of B cells, Mauri [/bib_ref] [bib_ref] What are Regulatory B cells?, Gray [/bib_ref].
Recently, increased percentages of Breg producing IL-10 were observed in chronically HIV-infected subjects [bib_ref] Regulatory B cell frequency correlates with markers of HIV disease progression and..., Siewe [/bib_ref]. This has prompted us to assess whether IL-10 expression profiles by precursor and mature MZ-like as well as TI and memory B-cell populations are modulated during HIV-1 infection. In addition, the fact that high levels of Lymphotoxin (LT)-a have been associated with autoimmune and inflammatory contexts [bib_ref] Lymphotoxin a revisited: general features and implications in rheumatoid arthritis, Calmon-Hamaty [/bib_ref] , and that increased LT-a to IL-10 B-cell expression ratios have been observed in patients with multiple sclerosis [bib_ref] Abnormal B cell cytokine responses a trigger of T cell-mediated disease in..., Bar-Or [/bib_ref] , prompted us to also assess B-cell LT-a expression profiles. We show increased frequencies of total B-cells expressing IL-10 in the blood of HIVinfected rapid and classic progressors as compared to those observed in HIV-uninfected donors. The most significant increase in cells expressing IL-10 is within the precursor MZ-like B-cell population. Furthermore, in contrast to viremic HIV-1-infected individuals, EC present MZ-like B-cell populations which IL-10 and LT-a expression profiles may favour homeostasis of immune responses and lymphoid microenvironments.
# Methods
# Ethics statement
Written informed consent was obtained from all subjects, and the research conformed to ethical guidelines and was approved by the Ethics Review board of the Centre de Recherche du CHUM. The reference number for the project is: SL 05.028.
## Subjects
Thirty male HIV-1-infected subjects were selected from the Montreal Primary HIV Infection (PHI) cohort: 13 were classified as rapid and 17 as classic progressors. The date of infection was estimated using criteria established by the Acute HIV Infection and Early Disease Research Program (NIAID, Bethesda, MD). Rapid progressors had blood CD4 + T-cell counts below 250 cells/ mm 3 within 2 years of infection. Blood samples were taken in acute (0-3 months) and/or early (5-8 months) phases of infection, and 3-6 and 9-12 months after initiation of antiretroviral therapy (ART). Classic progressors were ART-naive individuals whose blood CD4 + T-cell counts remained above 500 cells/mm 3 for the 2 year follow-up. Blood samples were obtained in the acute, early and chronic (24 months) phases of infection. Moreover, blood samples from 13 slow progressors were obtained from the Slow Progressors cohort, where patients were infected for 8 years or more with blood CD4 + T-cell counts above 500 cells/mm 3 and presented low (7 viremic patients) to undetectable (6 aviremic patients/EC) viral loads in absence of ART. Blood samples from 20 age-and sex-matched HIV-negative healthy volunteers were also obtained. HIV viral loads were determined in plasma using Versant HIV-1 RNA 3.0 Assay (Siemens Medical Solutions Diagnostics, Tarrytown, NY). Blood CD4 + T-cell counts were determined as reported [bib_ref] Persistent human immunodeficiency virus-1 antigenaemia affects the expression of interleukin-7Ralpha on central..., Mercier [/bib_ref]. None of the subjects had syphilis, hepatitis B or C.
## Evaluation of plasma il-10 and lt-a concentrations
Plasma levels of IL-10 were determined using the Cytometric Bead Array Human Inflammation Kit (BD-Biosciences, San Jose, CA, USA) according to the manufacturer's instructions. Data were acquired on a FACSAria and analysed with the FCAP software (BD-Biosciences). Plasma levels of .
## Evaluation of blood b-cell populations by flow-cytometry
Peripheral blood mononuclear cells (PBMCs) were isolated on Ficoll gradient, resuspended in heat-inactivated fetal bovine serum (hi-FBS) containing 10% dimethyl sulfoxyde and stored in liquid nitrogen. One million PBMCs per sample were used for staining. Live/dead exclusion was performed using Aqua-LIVE/DEAD Fixable Stain (Invitrogen Life technologies, Eugene, OR, USA). Non-specific binding sites were blocked using fluorescenceactivated cell sorting (FACS) buffer (16 PBS, 2% hi-FBS, and 0.1% sodium azide) supplemented with 20% hi-FBS and 10 mg mouse IgG (Sigma-Aldrich, St-Louis, MO, USA). The following conjugated mouse anti-human monoclonal antibodies were used: PacificBlue-anti-CD19, APC-Cy7-anti-CD10 (BioLegend, San Diego, CA, USA), AlexaFluor700-anti-CD27, FITC-anti-IgM, PE-anti-CD21 (BD-Biosciences), PerCP-eFluor710-anti-CD1c (eBioscience, San Diego, CA, USA), APC-anti-LT-a (LifeSpan BioSciences, Seattle, WA, USA), and rat anti-human PE-Cy7-anti-IL-10 (BioLegend). Intracellular labelling was performed using the Cytofix/Cytoperm Fixation/Permeabilization kit and perm/wash buffer (BD-Biosciences). Intracellular non-specific binding sites were blocked using perm/wash buffer containing 20% hi-FBS, 50% rat serum and 20 mg mouse IgG. Cells were kept at 4uC in 1.25% paraformaldehyde for 18 hours prior to analysis. Data acquisition of 10 5 events per sample was performed with LSRFortessa (BD-Biosciences), and analysis was done with FlowJo7.6.3 software (TreeStar, Ashland, OR, USA). All stainings were compared to that of fluorescence minus one (FMO) values and isotype controls (see [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] for details). Anti-mouse Ig(k) Compbeads and CS&T Beads (BD-Biosciences) were used to optimize fluorescence compensation settings and calibrate the LSRFortessa, respectively.
## Statistical analyses
The statistical significance of differences between groups was assessed with Fisher exact test for categorical variables and unpaired Student t test or one-way analysis of variance when continuous variables were normally distributed or with Mann-Whitney U test otherwise. Wilcoxon signed rank test was used for pairwise comparisons of different phases of infection within each group. Spearman rank test was used to determine correlations between continuous variables. Outliers' samples were withdrawn from statistical analyses when the values were greater than the mean plus three times standard deviation. Analyses were performed using GraphPad Prism 5.03 for Windows (GraphPad Software Inc, La Jolla, CA, USA).
# Results
## Socio-demographic and clinical characteristics of hiv-1-infected individuals
The socio-demographic and clinical characteristics of the HIV-1-infected individuals are shown in [fig_ref] Table 1: Sociodemographic and clinical characteristics of HIV-infected individuals [/fig_ref] , and the longitudinal assessment of blood CD4 + T-cell counts and viral loads are depicted in [fig_ref] Figure 2: IL-10 expression by blood B-cell populations [/fig_ref] -B. There were no significant correlations between blood CD4 + T-cell counts or viral loads and B-cell populations, LT-a and IL-10 plasma levels studied here either within groups or among all subjects during early or chronic infection (data not shown).
## Longitudinal monitoring of blood b-cell populations in hiv-1-infected individuals with different rates of disease progression
We have previously reported that the frequencies of blood precursor and mature MZ-like B-cell populations are different in the context of HIV disease control or progression [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref]. To further the characterization of these B-cells, we have adapted our staining cocktail for the detection of intracellular expression of IL-10 and LT-a (see below). The analyses of B-cell populations using this modified staining strategy confirm previous observations and are depicted in [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref]. The mean percentages of total live blood CD19 + B-cells in all HIV-1-infected individuals were lower throughout the course of infection than those observed in HIVnegative donors, reaching statistical significance in both rapid and classic progressors [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref]. Upon analysis for variations in the relative frequencies of different B-cell populations, we observed that the mean relative percentages of mature activated (CD19 + CD27 + IgM 2 CD21 l uCD10 2 ) and precursor MZ-like Bcells (CD19 + CD27 + IgM hi CD21 l uCD1c + CD10 + ), were significantly elevated in rapid and classic progressors [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] , E). There was a trend for increased relative frequencies of TI B-cells (CD19 + CD27 2 IgM + CD21 + CD10 + ) in chronically HIV-1-infected classic progressors [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref]. In contrast, the relative frequency of the more mature MZ-like B-cell population (CD19 + CD27 + IgM hi CD21 hi CD1c + CD10 2 ) within total blood B-cells remained unaltered in both rapid and classic progressors [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] , left and middle panels), whereas the percentages were significantly decreased in both viremic and aviremic slow progressors, the latter also referred herein as EC [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] , right panel). The relative frequencies of resting switched memory B-cells (CD19 + CD27 + IgM 2 CD21 hi CD10 2 ) were decreased in chronically infected classic progressors [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] , middle panel). Naïve resting B-cells (CD19 + CD27 2 IgM + CD21 hi CD10 2 ), were slightly lower in all HIV-1-infected individuals (data not shown).
Longitudinal monitoring of IL-10 expression by blood B-cell populations in HIV-1-infected individuals with different rates of disease progression Blood B-cell populations presented in [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] were assessed ex vivo for intracellular IL-10 expression based on the same strategy of flow-cytometry analysis depicted in [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref]. [fig_ref] Figure 2: IL-10 expression by blood B-cell populations [/fig_ref] shows a representative profile of IL-10 expression on total live B-cells from HIV-and HIV+ donors, and distribution within IgM and CD27 quadrants of expression. Note that the majority of IL-10 expressing B-cells are IgM + [fig_ref] Figure 2: IL-10 expression by blood B-cell populations [/fig_ref]. Because IL-10 expression might be influenced by the fluctuations in B-cell population frequencies shown in [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] , data are expressed as percentages of IL-10 expressing cells within each B-cell population. Frequencies of total B-cells expressing IL-10 in all viremic HIV-1-infected individuals were significantly higher than those observed in HIVnegative donors [fig_ref] Figure 2: IL-10 expression by blood B-cell populations [/fig_ref]. The percentages of total B-cells [fig_ref] Figure 2: IL-10 expression by blood B-cell populations [/fig_ref] , G, right panels). The frequencies of naïve resting B-cells expressing IL-10 remained unaltered in all HIV-infected individuals (data not shown). We did not detect changes in IL-10 expression levels following analysis of geometric mean fluorescence intensities (geoMFI) (data not shown). [fig_ref] Figure 3: Percentage of IL-10 expressing cells for each B-cell population, within each patient... [/fig_ref] shows a recapitulative analysis of B-cell IL-10 expression profiles. The heterogeneity of CD19 expression may be explained by the presence of blasts or other B-cell populations that were not specifically analyzed in the present study. The contribution to IL-10 expression from B-cell populations not analyzed herein will require further investigation. Data were plotted in order to compare the percentage of IL-10 expressing cells between each B-cell population, within each patient group at various time points [fig_ref] Figure 3: Percentage of IL-10 expressing cells for each B-cell population, within each patient... [/fig_ref]. Amongst the B-cell populations analyzed, the precursor MZ-like B-cells were the main expressors of IL-10 during the early phase of infection in both the rapid and classic progressor groups [fig_ref] Figure 3: Percentage of IL-10 expressing cells for each B-cell population, within each patient... [/fig_ref] , left panel; B, middle panel). This was also true for classic progressors in the chronic phase of infection [fig_ref] Figure 3: Percentage of IL-10 expressing cells for each B-cell population, within each patient... [/fig_ref]. All B-cell populations contributed equally to IL-10 expression in the slow progressor and HIV-negative donor groups [fig_ref] Figure 3: Percentage of IL-10 expressing cells for each B-cell population, within each patient... [/fig_ref]. Of note, plasma levels of IL-10 were found to be elevated throughout follow-up in all HIV-1-infected individuals, including ART-treated rapid progressors and aviremic slow progressors/EC [fig_ref] Figure 2: IL-10 expression by blood B-cell populations [/fig_ref].
Overall, frequencies of total blood B-cells expressing IL-10 in all viremic HIV-1-infected individuals were significantly higher than those observed in HIV-negative donors. Amongst the B-cell populations analyzed, the relative percentage of IL-10 expressing cells is significantly increased within the precursor MZ-like, TI and memory populations. The most significant increase in cells expressing IL-10 was within the precursor MZ-like population. Importantly, we found unaltered percentages of total and of MZlike B-cells expressing IL-10 in aviremic slow progressors/EC.
## Longitudinal monitoring of lt-a expression by blood b-cell populations in hiv-1-infected individuals with different rates of disease progression
Blood B-cell populations presented in [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] were assessed ex vivo for intracellular LT-a based on the same strategy of flowcytometry analysis depicted in [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref]. [fig_ref] Figure 4: LT-a expression by blood B-cell populations [/fig_ref] shows a representative profile of total live B-cells expressing LT-a in HIVand HIV+ donors, and distribution within IgM and CD27 quadrants of expression. The data are presented as percentage of LT-a expressing cells within each B-cell population. There was a significant increase in frequencies of total B-cells expressing LT-a in the blood of both classic and slow progressors when compared to HIV-negative donors [fig_ref] Figure 4: LT-a expression by blood B-cell populations [/fig_ref]. There was a trend for increased frequencies of total B-cells expressing LT-a following therapy in rapid progressors [fig_ref] Figure 4: LT-a expression by blood B-cell populations [/fig_ref]. Similar patterns of LT-a expression were observed in mature activated, precursor MZ-like, mature MZ-like and TI B-cell populations [fig_ref] Figure 4: LT-a expression by blood B-cell populations [/fig_ref] , E-G). The relative frequencies of resting switched memory B-cells expressing LT-a were not significantly affected [fig_ref] Figure 4: LT-a expression by blood B-cell populations [/fig_ref] , as were those of naïve resting B-cells (data not shown). We did not detect changes in LT-a expression levels following analysis of geoMFI (data not shown). A recapitulative analysis of B-cell LT-a expression profiles is provided in [fig_ref] Figure 4: LT-a expression by blood B-cell populations [/fig_ref].
Data were plotted in order to compare the percentage of LT-a expressing cells between each B-cell population, within each patient group at various time points [fig_ref] Figure 5: Percentage of LT-a expressing cells for each B-cell population, within each patient... [/fig_ref]. Amongst the B-cell populations analyzed, the precursor MZ-like and TI B-cells were the main contributors of LT-a expression during the acute, early and chronic phases of infection in all viremic HIV-1-infected individuals [fig_ref] Figure 5: Percentage of LT-a expressing cells for each B-cell population, within each patient... [/fig_ref] -B; C, left panel). The precursor MZ-like Bcells were the main contributors of LT-a expression in the aviremic slow progressor/EC and HIV-negative groups [fig_ref] Figure 5: Percentage of LT-a expressing cells for each B-cell population, within each patient... [/fig_ref]. Plasma levels of LT-a in all HIV-infected individuals were similar to those observed in HIV-negative donors [fig_ref] Figure 2: IL-10 expression by blood B-cell populations [/fig_ref]. Overall, increased relative frequencies of precursor MZ-like B-cells expressing LT-a appear associated with a better control of disease progression.
# Discussion
The percentages of total live blood CD19 + B-cells were lower in all HIV-1-infected individuals when compared to those observed in HIV-negative donors. However, despite the fact that precursor MZ-like B-cells may represent only a small fraction of circulating B-cells, their increased relative frequencies in the blood of HIV-1infected rapid and classic progressors, may significantly influence the ability to mount efficient B-cell responses and disease progression. Indeed, given the frequent auto-reactive and crossreactive repertoires of MZ type populations [bib_ref] Marginal zone B cells: virtues of innate-like antibody-producing lymphocytes, Cerutti [/bib_ref] [bib_ref] Human Marginal Zone B cells, Weill [/bib_ref] , the increased CD1c and CD10 expression were used for further characterisation of blood MZ and TI B-cell populations respectively, as reported [bib_ref] Human Marginal Zone B cells, Weill [/bib_ref]. Quadrants were set based on the expression values obtained with fluorescence minus one (FMO) and isotype controls. Mature activated B-cells are defined as CD19+CD27+IgM-CD21loCD1c-CD10-, resting switched memory B-cells are CD19+CD27+IgM-CD21hiCD10-, precursor marginal-zone (MZ)-like B-cells are CD19+CD27+IgM+CD21loCD1c+ CD10+, mature MZ-like B-cells are CD19+CD27+IgM+CD21hiCD1c+CD10-and transitional immature (TI) B-cells are CD19+CD27-IgM+CD21hiCD1c-CD10+. The graphs present (B) percentages of total B-cells (mean events gated: 932061750), and the relative frequencies of (C) mature activated (mean events gated: 360667), (D) resting switched memory (mean events gated: 6326301), (E) precursor MZ-like (mean events gated: 145636), (F) mature MZ-like (mean events gated: 3276233) and (G) TI (mean events gated: 9446174) B-cell populations in the blood of rapid progressors (left panels; 5-8 months PI (n = 11), 3-6 months ART (n = 6) and 9-12 months ART (n = 5)), classic progressors (middle panels; 0-3 months PI (n = 12), 5-8 months PI (n = 17), and 24 months PI (n = 13)), and viremic and aviremic slow progressors (EC) (right panels; viremic (n = 6); aviremic (n = 5)). The same values for HIV-negative donors in the left, middle and right panels are used as a control group (n = 7). Cell population frequencies were compared using the Wilcoxon signed rank test and the Mann-Whitney U test for pairwise comparisons of different phases of infection within each group and between the study groups, respectively. Data shown are mean 6SEM. * p,0.05. PI, post-infection; ART, antiretroviral therapy. doi:10.1371/journal.pone.0101949.g001 rapid progressors (left panels; 5-8 months PI (n = 11), 3-6 months ART (n = 6) and 9-12 months ART (n = 5)), classic progressors (middle panels; 0-3 months PI (n = 12), 5-8 months PI (n = 17), and 24 months PI (n = 13)), and viremic and aviremic slow progressors (EC) (right panels; viremic (n = 6); aviremic (n = 5)). The same values for HIV-negative donors in the left, middle and right panels are used as a control group (n = 7). Data are expressed as percentages of IL-10 expression within each B-cell population. Cell population frequencies were compared using the Wilcoxon signed rank test and the Mann-Whitney U test for pairwise comparisons of different phases of infection within each group and between the study groups, respectively. Data shown are mean 6SEM. * p,0.05. PI, postinfection; ART, antiretroviral therapy. doi:10.1371/journal.pone.0101949.g002 relative frequency of these cells may be associated with polyclonal B-cell activation and hyperglobulinemia previously reported for these individuals [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref] , leading to breakage of tolerance and autoimmune manifestations. Moreover, their increased relative proportions may reflect the host response to lymphopenia observed in rapid progressors [bib_ref] Idiopathic CD4+ T lymphocytopenia is associated with increases in immature/transitional B cells..., Malaspina [/bib_ref] and the overall inflammation and high levels of BLyS/BAFF [bib_ref] Excess BAFF rescues self-reactive B cells from peripheral deletion and allows them..., Thien [/bib_ref] [bib_ref] B-cell tolerance breakdown in Sjögren's syndrome: focus on BAFF, Varin [/bib_ref] [bib_ref] The role of B Lymphocyte Stimulator (BLyS) in systemic lupus erythematosus, Cancro [/bib_ref] [bib_ref] BAFF and selection of autoreactive B cells, Liu [/bib_ref] [bib_ref] B lymphocyte stimulator protein-associated increase in circulating autoantibody levels may require CD4+..., Stohl [/bib_ref] [bib_ref] Plasma levels of B-lymphocyte stimulator increase with HIV disease progression, Rodriguez [/bib_ref] previously found in the plasma and on surface of blood mDCs in these individuals [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref]. In support of our findings, it has been shown that human blood and tonsil MZ-like B-cells naturally bind to gp120 through C-type lectins, and that this is increased by BLyS/BAFF in vitro, subsequently leading to polyclonal IgG and IgA responses, of which a fraction is gp120-specific [bib_ref] Marginal zone B cells: virtues of innate-like antibody-producing lymphocytes, Cerutti [/bib_ref]. Furthermore, the fact that the individuals presented herein experience microbial translocation [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref] is in line with human MZ B-cells being particularly solicited in response to encapsulated bacteria [bib_ref] IgM+ IgD+CD27+ B cells are markedly reduced in IRAK-4-, MyD88-, and TIRAPbut..., Weller [/bib_ref]. In contrast, relative blood levels of precursor MZ-like B-cells in EC were similar to those observed in HIV-negative donors [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref]. Moreover, EC had low-inflammatory profiles, unaltered levels of BLyS/ BAFF and undetectable elements of microbial translocation [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref] , suggesting they are able to maintain a certain degree of immune integrity.
Interestingly, both viremic and aviremic slow progressors/EC had decreased relative frequencies of more mature MZ-like B-cells within their blood B-cell compartment when compared to both rapid and classic progressors, and HIV-negative donors. This finding could suggest that the capacity to recruit this population to peripheral sites may be beneficial to the ''control'' of disease progression. Intriguingly, a similar decrease in blood CD27 + Ig-M + IgD + MZ type B-cells has been recently reported in patients with MyD88, IRAK4, and TIRAP deficiencies [bib_ref] IgM+ IgD+CD27+ B cells are markedly reduced in IRAK-4-, MyD88-, and TIRAPbut..., Weller [/bib_ref]. Whether there is a relationship with the reduced frequencies of mature MZlike B-cells we observe in slow progressors warrants further investigation.
Although our sample size was limited, we were able to accurately and reproducibly measure ex vivo IL-10 expression by blood B-cells. We found that in the context of detectable viremia, the percentages of total blood B-cells expressing IL-10 were increased when compared to HIV-negative individuals [fig_ref] Figure 2: IL-10 expression by blood B-cell populations [/fig_ref]. Amongst the B-cell populations analysed, the percentage of IL-10 expressing cells was significantly increased within the precursor MZ-like, TI and switched memory populations. Strikingly, the most significant increase in cells expressing IL-10 was within the precursor MZ-like population. Importantly, the relative increase in precursor MZ-like B-cells expressing IL-10 correlated with high levels of serum IgG1 and IgG3 in the early phase of HIV-1infection in classic progressors (p = 0.0184; p = 0.0499, respectively), and with high levels of serum IgA in chronic classic progressors (p = 0.0479). These are likely to reflect polyclonal B-cell activation associated with the overall inflammatory condition we observed in HIV-1 progressors [bib_ref] High expression levels of B Lymphocyte Stimulator (BLyS) by dendritic cells correlate..., Fontaine [/bib_ref] [bib_ref] HIV infection affects blood myeloid dendritic cells after successful therapy and despite..., Fontaine [/bib_ref] , as B-cell IL-10 production has been shown to be modulated by BLyS/BAFF [bib_ref] B cell-activating factor enhances interleukin-6 and interleukin-10 production by ODN-activated human B..., Yehudai [/bib_ref] and various signals, such as delivered via TLR, CD40 and the B-cell receptor (BCR) [bib_ref] Immune Regulatory function of B cells, Mauri [/bib_ref] [bib_ref] What are Regulatory B cells?, Gray [/bib_ref].
Our findings could also suggest an attempt from the host to dampen the overwhelming inflammatory burden by soliciting regulatory functions from various B-cell populations, such as switched memory, TI and MZ-like B-cells, which have been variably associated with Breg capacity [bib_ref] Immune Regulatory function of B cells, Mauri [/bib_ref] [bib_ref] What are Regulatory B cells?, Gray [/bib_ref]. Of course, we cannot rule out that precursor MZ-like B-cells are more sensitive to the overall HIV disease process, but it may be that they represent a population more prone to Breg potential. Frequencies of populations with similar characteristics and Breg attributes have been shown to be increased in the blood of systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) patients, albeit these cells were refractory to in vitro stimulation, produced less IL-10 and lacked suppressive activity [bib_ref] Immune Regulatory function of B cells, Mauri [/bib_ref]. Likewise, it is possible that in the context of HIV disease progression, populations ascribed to Breg potential may have exhausted and/or dysregulated regulatory capacities.
Nevertheless, the overall outcome of increased frequencies of B-cells expressing IL-10 may well be to sustain chronic B-cell activation and dysregulation, and may lead to imbalanced Treg/ Teffector ratios [bib_ref] Effector and regulatory B cells: modulators of CD4+ T cell immunity, Lund [/bib_ref] associated with HIV disease progression [bib_ref] Microbial translocation, immune activation, and HIV disease, Klatt [/bib_ref]. Furthermore, high level of Breg activity was recently shown to be involved in suppression of antiviral T effector functions in vitro during HIV [bib_ref] Regulatory B cell frequency correlates with markers of HIV disease progression and..., Siewe [/bib_ref] and chronic hepatitis B [bib_ref] IL-10-Producing Regulatory B Cells in the Pathogenesis of Chronic Hepatitis B Virus..., Das [/bib_ref] infections, thus impending viral eradication.
In the context of HIV disease control, such as encountered in EC, the frequencies of total and MZ-like B-cells expressing IL-10, remained unaltered. Although increased frequencies of resting switched memory and TI blood B-cells expressing IL-10 were found in EC, these were lower than in the MZ-like populations. These findings suggest that EC are capable, to a certain extent, to regulate their IL-10 expression status and possibly ''Breg'' activity, and this may help in maintaining homeostasis of immune responses associated with HIV control.
Although high levels of LT-a have been associated with autoimmune and inflammatory conditions [bib_ref] Lymphotoxin a revisited: general features and implications in rheumatoid arthritis, Calmon-Hamaty [/bib_ref] , plasma levels of LT-a in all HIV-1-infected subjects were similar to those found in HIV-negative donors. Amongst the B-cell populations analyzed, the precursor MZ-like B-cells were the main contributors to LT-a expression in all subjects regardless of HIV-1 infection and rate of disease progression. The percentages of IL-10 expressing cells were significantly greater than those of LT-a within most B-cell populations except for precursor MZ-like and TI B-cells, which frequencies of IL-10 and LT-a expression were similar. Importantly, EC had increased frequencies of MZ-like B-cells expressing LT-a resulting in higher LT-a to IL-10 expression ratios. Studies in the murine system have elegantly shown the importance of LTa in the organization and maintenance of lymphoid structures, as well as in the modulation of immune responses [bib_ref] Lymphotoxin a revisited: general features and implications in rheumatoid arthritis, Calmon-Hamaty [/bib_ref] , through a process involving a positive CXCL13 feedback loop [bib_ref] A chemokine-driven positive feedback loop organizes lymphoid follicles, Ansel [/bib_ref]. Interestingly, plasma levels of CXCL13 were higher in viremic individuals compared to HIV-negative donors [bib_ref] Altered expression of the receptor-ligand pair CXCR5/CXCL13 in B cells during chronic..., Cagigi [/bib_ref] and EC (Chagnon-Choquet J unpublished data), suggesting that the higher relative percentages of MZ-like B-cells expressing LT-a in EC may rapid progressors (left panels; 5-8 months PI (n = 11), 3-6 months ART (n = 6) and 9-12 months ART (n = 5)), classic progressors (middle panels; 0-3 months PI (n = 12), 5-8 months PI (n = 17), and 24 months PI (n = 13)), and viremic and aviremic slow progressors (EC) (right panels; viremic (n = 6); aviremic (n = 5)). The same values for HIV-negative donors in the left, middle and right panels are used as a control group (n = 7). Data are expressed as percentages of LT-a expression within each B-cell population. Cell population frequencies were compared using the Wilcoxon signed rank test and the Mann-Whitney U test for pairwise comparisons of different phases of infection within each group and between the study groups, respectively. Data shown are mean 6SEM. * p,0.05. PI, postinfection; ART, antiretroviral therapy. doi:10.1371/journal.pone.0101949.g004 help to maintain CXCL13 levels and a certain lymphoid homeostasis.
The first-line MZ-like B-cells described herein likely contribute directly to the generation of HIV-specific Abs. Indeed, human MZ B-cells express IGHV1-2*02 [bib_ref] Mouse marginal zone B cells harbor specificities similar to human broadly neutralizing..., Pujanauski [/bib_ref] , which has been repeatedly found to encode mutated HIV broadly neutralizing Abs (bNAbs), such as VRC01 [bib_ref] Structural basis for broad and potent neutralization of HIV-1 by antibody VRC01, Zhou [/bib_ref]. Interestingly, the recent characterization of transient Gp41-specific IgA in mucosal genital fluids from patients within the first weeks after HIV transmission, suggest these Abs might have originated from first-line B-cell populations [bib_ref] HIV-1 gp41 envelope IgA is frequently elicited after transmission but has an..., Yates [/bib_ref]. Of note, BLyS/BAFF was elevated immediately preceding the appearance of these Abs. Interestingly, repeated treatment of mice with BLyS/BAFF increased the MZ compartment, and generated an increased response to Env immunization and bNAbs from these animals [bib_ref] BLyS-Mediated Modulation of Naive B Cell Subsets Impacts HIV Env-Induced Antibody Responses, Dosenovic [/bib_ref]. Understanding the dynamics of BLyS/ BAFF and its role in homeostasis of immune responsiveness appears pivotal to the design of vaccine strategies soliciting protective B-cell responses. Based on our observations, the capacity to contain BLyS/BAFF expression levels seems associated with control of disease progression, whereas excessive BLyS/BAFF may promote immune dysregulation and disease progression [bib_ref] Dendritic Cell Status Modulates the Outcome of HIV-Related B Cell Disease Progression, Poudrier [/bib_ref].
The findings reported herein are in line with growing evidence suggesting that first-line B-cell responses are involved in the battle against HIV [bib_ref] Innate immunity in acute HIV-1 infection, Borrow [/bib_ref] , and with the importance of MZ type B-cells in health and disease [bib_ref] Marginal zone B cells: virtues of innate-like antibody-producing lymphocytes, Cerutti [/bib_ref] [bib_ref] Human Marginal Zone B cells, Weill [/bib_ref]. Although further characterization is required to identify the exact nature of the MZ-like B-cells described herein, they certainly deserve further interest. [fig_ref] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals [/fig_ref] , and slow progressors (right panel; viremic (n = 6), aviremic (n = 5)). (B) Viral loads (log copies/ ml) were quantified by in vitro signal amplification nucleic acid probe assay of HIV-1 RNA (bDNA) in the plasma of rapid progressors (left panel; 0-3 months PI (n = 13), 5-8 months PI (n = 13), 3-6 months ART (n = 7), 9-12 months ART (n = 6)), classic progressors (middle panel; 0-3 months PI (n = 17), 5-8 months PI (n = 17), 24 months PI (n = 11)), and slow progressors (right panel; viremic (n = 6), aviremic (n = 6)). (C) Concentrations of IL-10 measured longitudinally in the plasma of rapid progressors (left panel; 0-3 months PI (n = 12), 5-8 months PI (n = 13), 3-6 months ART (n = 9), 9-12 months ART (n = 7)), classic progressors (middle panel; 0-3 months PI (n = 17), 5-8 months PI (n = 17), 24 months PI (n = 11)) and slow progressors (right panel; viremic (n = 7), aviremic (n = 5)). The same values for HIV-negative donors (n = 20) in the left, middle and right panels are used as a control group. (D) Concentrations of LT-a measured longitudinally in the plasma of rapid progressors (left panel; 0-3 months PI (n = 10), 5-8 months PI (n = 12), 3-6 months ART (n = 6), 9-12 months ART (n = 4)), classic progressors (middle panel; 0-3 months PI (n = 14), 5-8 months PI (n = 10), 24 months PI (n = 9)) and slow progressors (right panel; viremic (n = 4), aviremic (n = 4)). The same values for HIV-negative donors (n = 18) in the left, middle and right panels are used as a control group. Cell populations, viral loads and plasma concentrations were compared using the Wilcoxon signed-rank test and Mann-Whitney U test for pairwise comparisons of different phases of infection within each group and between the study groups, respectively. Data shown are mean 6SEM. Significance levels are shown as * p,0.05, ** p,0.001, *** p,0.0001. PI, post-infection; ART, antiretroviral therapy. (TIF) [fig_ref] Figure 3: Percentage of IL-10 expressing cells for each B-cell population, within each patient... [/fig_ref] Contribution of each blood B-cell population to IL-10 expression. Percentages of IL-10 expression within each B-cell population; 'mature' marginal zone (MZ)-like (purple), 'precursor' MZ-like (cherry red), mature activated (yellow), transitional immature (TI) (blue) and resting switched memory (orange) B-cells, for rapid progressors (left panel; 5-8 months PI (n = 11), 3-6 months ART (n = 6), 9-12 months ART (n = 5)), classic progressors (middle panel; 0-3 months PI (n = 12), 5-8 months PI (n = 17), 24 months PI (n = 13)), and slow progressors (right panel; viremic (n = 6), aviremic (n = 5)). The same value for HIV-negative donors in the left, middle and right panels are used as a control group (n = 7). Cell population frequencies were compared using the Mann-Whitney U test between the study groups. Data shown are mean 6 SEM. * p,0.05. PI, postinfection; ART, antiretroviral therapy. (TIF) [fig_ref] Figure 4: LT-a expression by blood B-cell populations [/fig_ref] Contribution of each blood B-cell population to LT-a expression. Percentages of LT-a expression within each B-cell population; 'mature' marginal zone (MZ)-like (purple), 'precursor' MZ-like (cherry red), mature activated (yellow), transitional immature (TI) (blue) and resting switched memory (orange) B-cells, for rapid progressors (left panel; 5-8 months PI (n = 11), 3-6 months ART (n = 6), 9-12 months ART (n = 5)), classic progressors (middle panel; 0-3 months PI (n = 12), 5-8 months PI (n = 17), 24 months PI (n = 13)), and slow progressors (right panel; viremic (n = 6), aviremic (n = 5)). The same value for HIV-negative donors in the left, middle and right panels are used as a control group (n = 7). Cell population frequencies were compared using the Mann-Whitney U test between the study groups. Data shown are mean 6SEM. * p,0.05. PI, postinfection; ART, antiretroviral therapy. (TIF)
## Supporting information
[fig] Figure 1: Longitudinal analysis of blood B-cell populations of HIV-1 infected individuals. (A) Representative plot showing gating strategy on live PBMCs. Total CD19+ B-cells were selected based on expression of CD27 and/or IgM, and levels of CD21. [/fig]
[fig] Figure 2: IL-10 expression by blood B-cell populations. (A) Representative plot showing gating strategy on live total CD19+ B-cells from HIVand HIV+ donors, expressing IL-10. Frequencies of cells expressing IL-10 within (B) total, (C) mature activated, (D) resting switched memory, (E) precursor marginal zone (MZ)-like, (F) mature MZ-like and (G) transitional immature (TI) B-cell populations in the blood of [/fig]
[fig] Figure 3: Percentage of IL-10 expressing cells for each B-cell population, within each patient group. Percentage of IL-10 expressing cells within mature activated, resting switched memory, precursor marginal zone (MZ)-like, mature MZ-like and transitional immature (TI) B-cells within (A) rapid progressors (5-8 months PI (n = 11), 3-6 months ART (n = 6) and 9-12 months ART (n = 5), (B) classic progressors (0-3 months PI (n = 12), 5-8 months PI (n = 17), and 24 months PI (n = 13)), (C) slow progressors (viremic (n = 6); aviremic (n = 5)), and HIV-negative individuals (n = 7). Percentages were compared using the Mann-Whitney U test between the B-cell populations. Data shown are mean 6SEM. *p,0.05. PI, postinfection; ART, antiretroviral therapy. doi:10.1371/journal.pone.0101949.g003 [/fig]
[fig] Figure 4: LT-a expression by blood B-cell populations. (A) Representative plot showing gating strategy on live total CD19+ B-cells from HIVand HIV+ donors, expressing LT-a. Frequencies of cells expressing LT-a within (B) total, (C) mature activated, (D) resting switched memory, (E) precursor marginal zone (MZ)-like, (F) mature MZ-like and (G) transitional immature (TI) B-cells expressing LT-a in the blood of [/fig]
[fig] Figure 5: Percentage of LT-a expressing cells for each B-cell population, within each patient group. Percentage of LT-a expressing cells within mature activated, resting switched memory, precursor marginal zone (MZ)-like, mature MZ-like and transitional immature (TI) B-cells within (A) rapid progressors (5-8 months PI (n = 11), 3-6 months ART (n = 6) and 9-12 months ART (n = 5), (B) classic progressors (0-3 months PI (n = 12), 5-8 months PI (n = 17), and 24 months PI (n = 13)), (C) slow progressors (viremic (n = 6); aviremic (n = 5)), and HIV-negative individuals (n = 7). Percentages were compared using the Mann-Whitney U test between the B-cell populations. Data shown are mean 6SEM. * p,0.05, ** p,0.001, *** p,0.0001. PI, postinfection; ART, antiretroviral therapy. doi:10.1371/journal.pone.0101949.g005 [/fig]
[fig] Figure S1 Flow: -cytometry gating strategy based on fluorescence minus one and isotype controls. Dot plots showing gating strategy based on fluorescence minus one and isotype control for (A) CD19, (B) CD21, (C) CD27 and IgM, (D) CD1c and CD10, (E) IL-10 and (F) LT-a. (TIF) Figure S2 Longitudinal variations of blood CD4+ T-cell counts, viral loads and IL-10 and LT-a concentrations of HIV-1 infected individuals. (A) Blood CD4+ T-cell counts (cell/mm3) were determined by flow-cytometry in rapid progressors (left panel; 0-3 months PI (n = 13), 5-8 months PI (n = 13), 3-6 months ART (n = 7), 9-12 months ART(n = 5)), classic progressors (middle panel; 0-3 months PI (n = 16), 5-8 months PI (n = 16), 24 months PI (n = [/fig]
[table] Table 1: Sociodemographic and clinical characteristics of HIV-infected individuals.Age, CD4 and viremia are expressed as mean 6SD. Sex and race were compared using Fisher exact test. Pairwise comparisons of CD4 and viremia for early phases were performed using unpaired Student's t tests. Comparisons among all groups (age at first visit, CD4, viremia in the chronic phase, nadir CD4) were performed with the one-way analysis of variance test. n, numbers; NS, not significant; na, not available. a P = 0.004 and 0.050 for the comparison of age between rapid and viremic slow progressors, and classic and viremic slow progressors, respectively, as determined by the Mann-Whitney test. b P = 0.008 for the comparison of CD4+ T cells/mm 3 in chronic phase between rapid progressors and aviremic slow progressors, as determined by the Mann-Whitney test. c P = 0.0008 and 0.001 and 0.020 nadir CD4 for the comparison between rapid and classic progressors, rapid and viremic slow progressors, and rapid and aviremic slow progressors, respectively, as determined by the Mann-Whitney test. d Fifty copies/ml corresponds to the detection threshold of the viral load test. [/table]
|
Parental educational level and childhood wheezing and asthma: A prospective cohort study from the Japan Environment and Children’s Study
BackgroundThe influence of mothers' and fathers' educational levels in separate evaluations of asthma has not been fully investigated. This study aims to examine the associations of the mother's and fathers' educational levels with childhood wheeze and asthma adjusting for crude and pre-and post-natal modifiable risk factors.MethodsWe conducted a prospective cohort study using data from the Japan Environment and Children's Study, which recruited pregnant women from 2011 to 2014. The mother's and father's educational levels were surveyed by a questionnaire during the pregnancy, and childhood wheezing and doctor-diagnosed asthma were estimated using a 3-year questionnaire. Multilevel logistic regression analysis was performed to evaluate the association between the mother's and father's educational levels and childhood wheezing and asthma, adjusted for pre-and post-natal factors.ResultsA total of 69,607 pairs of parents and their single infants were analyzed. We found 17.3% of children had wheezing and 7.7% had asthma. In crude analyses, lower educational level of parents was associated with an increased risk of childhood wheezing and asthma. After full adjustment, a lower educational level of mothers was associated with an increased risk of childhood asthma (junior high school (reference: high school); odds ratio (OR): 1.17, 95%
[formula] a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 [/formula]
# Introduction
Asthma is a common non-communicable disease affecting people of all ages around the world, and it has a substantial burden of disease like premature death and reduced quality of life. The prevalence of childhood asthma in Japan is at the middle-ranked level in the world, although a very recent survey indicates that asthma prevalence is declining. Wheezing is a frequent symptom for preschool children, which often disappear by the time they reach school age, but some of them remain symptomatic, developing persistent wheezing. Further, adult-onset asthma may be affected by early childhood wheezing with pathogenesis changes.
The social determinants of health are one of the principal causes of health inequalities, and parents' low socioeconomic status negatively influences their child's physical and mental health. Several studies reported that lower mother's educational level (no evaluation of father's educational level) was associated with childhood asthma, but no association was reported after adjusting for post-natal factors including breastfeeding. Swedish and Dutch studies which defined the parental educational level as the highest education attained within each couple found that lower educational level was associated with significantly higher childhood asthma risk, but the adjustments did not include post-natal factors. A study that measured both mothers' and fathers' educational levels reported that only lower paternal educational level had a significant relationship without postnatal factor adjustment. However, one Japanese study reported no significant relationship of both mother's and father's educational levels with pre-and postnatal factors adjustment.
Thus, parental lower educational attainment broadly links to childhood asthma, but the influence is attenuated by modifiable factors, especially post-natal factors. Furthermore, the influence of mothers' and fathers' educational levels on childhood asthma has not been fully investigated. This study aims to examine the crude and pre-and post-natal modifiable risk factors and adjusted associations of mothers' and fathers' educational levels with childhood wheezing and asthma by using birth cohort data from the Japan Environment and Children's Study (JECS).
# Methods
## Participants
The JECS is an ongoing nationwide prospective birth cohort study in Japan that aims to identify environmental factors of children's health and development. Fifteen regional centers were selected to cover all the Japanese geographical areas (Hokkaido, Miyagi, Fukushima, Chiba, Kanagawa, Koshin, Toyama, Aichi, Kyoto, Osaka, Hyogo, Tottori, Kochi, Fukuoka, and South Kyushu/Okinawa). To maximize representativeness, baseline recruitment was performed in collaboration with local governments and co-operating health care providers. Follow-up of the children from these pregnancies is ongoing until they reach 13 years of age.
Pregnant women in the early stages of their pregnancy who resided in the study area were recruited, and a total of 103,060 pregnancies throughout Japan were covered in this study between January 2011 and March 2014. It is difficult to correctly indicate the coverage of the children (the number of live births registered in JECS divided by the number of all live births within the study areas) for the whole study period because we recruited women in early pregnancy and later expanded the study areas. However, in 2013, when recruitment was largely stabilized, the child coverage was approximately 45%. Discounting pregnancies in the same woman, the study involved 97,413 unique pregnancies. After excluding miscarriage, stillbirth, multiple births, positive physical abnormalities (infants' medical records at one month of age), not responding to the three-year questionnaire (response rate of unique pregnancy without before mentioned exclusion criteria: 82.3%), and not responding to wheezing queries of the International Study of Asthma and Allergies in Childhood (ISAAC)in the three-year questionnaire, the number of final analyzed participants was 69,067 children. We used the dataset of jecs-ta-20190930 from the JECS, which is the registry of the JECS is the University Hospital Medical Information Network (UMIN) 000030786 (UMIN Clinical Trials Registry).
# Ethics statement
The JECS protocol was approved by the Institutional Review Board on epidemiological studies of the Ministry of the Environment and by the Ethics Committees of all participating institutions. The JECS was conducted in accordance with the Declaration of Helsinki and other nationally valid regulations. Written informed consent was obtained from all participating mothers and fathers.
## Outcomes
Wheezing (primary outcome) and doctor-diagnosed (secondary outcome: additional analysis) asthma at the 3-year questionnaire were outcomes. In the primary outcome, we used the Japanese translation of the International Study of Asthma and Allergies in Childhood (ISAAC) questionnaire for 6-7-year children. Caregivers were asked, "Has your child ever had wheezing or whistling in the chest at any time in the past?" If yes, then they were asked "Has your child had wheezing or whistling in the chest in the last 12 months (yes or no)?" If yes, we defined the children as wheezing positive at 3-years. In the secondary outcome, caregivers were also asked, "Has your child had any diseases which a medical doctor diagnosed since 2-years age?" If they had asthma and we checked whether they had any other diseases or "no diseases," we defined the children as negative of doctor-diagnosed asthma. If any diseases and diseases were unchecked, the outcome of doctor-diagnosed asthma was considered as a missing value.
## Mother's and father's educational levels
Questionnaires were administered to enrolled mothers during the first (T1) and second/third trimester (T2). The T2 questionnaire included questions about the educational attainments of mothers and fathers. They were categorized as � 9 years (junior high school: EDC1), 10-12 years (high school: EDC2), 13-15 years (technical junior college, technical/vocational college, or associate degree: EDC3), or � 16 years (bachelor's degree or postgraduate degree: EDC4).
## The other independent variables
Based on previous studies, the following variables were selected as covariates: child sex, gestational age, season of birth, parity, type of delivery, mother and father age, pre-pregnancy body mass index (BMI), household income, marital status, mother and father smoking status during pregnancy, mother and father allergy, breast milk, nursery, lower respiratory infection, mold at home, pet, and passive smoke.
The T1 questionnaire for mothers included questions regarding the mother's date of birth, marital status, mother and father smoking habit, and past history of mother doctor-diagnosed allergy-related diseases. Marital status was classified as married, unmarried, divorced, or bereavement. Smoking habit was categorized as never smoked, quit smoking before pregnancy, quit smoking during pregnancy, or continued smoking during pregnancy. Mothers responded to the question, "Have you ever been diagnosed by a physician for asthma, allergic rhinitis/hay fever, atopic dermatitis, allergic conjunctivitis, or food allergy (including oral allergy syndrome)?". We defined anyone with at least one allergic disease as allergy positive. A questionnaire for fathers during pregnancy also included birth day and the past history of doctor-diagnosed allergyrelated diseases, and they were categorized in the same manner as mother allergy.
The following information was also collected from medical records transcripts: child's date of birth, parity, gestational period, height, and pre-pregnancy weight, from which BMI was
## Plos one
Parental educational level, childhood wheezing and asthma calculated. Mother and father ages at the time of childbirth were calculated from their birth date and the child's birth date.
Breastfeeding was categorized as � 1 month, 2 to 5 months, and �6 months using a 1-year questionnaire. Using a 3-year questionnaire, nursery admission was defined if the child went to a nursery and was under 2 years of age.
Regarding the child's environment after birth, mold growth, pet at home, and passive smoke were collected from the 1.5-year questionnaire. Mold growth positive was defined if they answered positive to any of the queries regarding mold growth in the living room, child's bedroom, or caregiver's bedroom if the child slept there. Pet at home positive was defined if they had any dogs, cats, small mammals, or birds. Caregivers were asked, "Has anyone smoked around your child?", and the child's exposure to passive smoke was categorized as none, sometimes, and often.
The number of lower respiratory infections was defined as adding the positive numbers of 1-year (recall period: from birth), 1.5-year (recall period: from one year old), and 2-year questionnaires (recall period: from 1.5 years old).
# Statistical analysis
The associations between the two outcomes and child's sex, gestational age, season of birth, parity, type of delivery, mother's and father's age, pre-pregnancy BMI, household income, marital status, mother's and father's smoking status during pregnancy, mother's and father's allergy, breast milk, nursery, lower respiratory infection, mold at home, pet, passive smoke, and mother's and father's educational levels were investigated using Fisher's exact test or chisquare test (supplemental analysis).
To impute missing values for participants with missing data (7.1% of all data except for 15 regional units that had no missing data), the information was replaced using multiple imputations (25 imputed datasets) based on the assumption that data were missing at random. The variables included in the imputation model were as follows: 15 regional centers, wheezing, doctor-diagnosed asthma, child sex, gestational age, season of birth, parity, type of delivery, mother's and father's age, pre-BMI, household income, marital status, mother's and father's smoking status during pregnancy, mother's and father's allergy, breast milk, nursery, lower respiratory infection, mold at home, pet, passive smoke, and mother's and father's educational levels. Using the imputed datasets, the crude odds ratios (ORs) of each variable for childhood wheezing and doctor-diagnosed asthma were calculated (supplemental analysis).
Next, we conducted multilevel logistic regression analyses in which individual-level factors were at the first level and 15 regional centers at the second level to estimate ORs for childhood wheezing (primary outcome) and doctor-diagnosed asthma (secondary outcome: additional analysis) with 95% confidence intervals (95% CIs). First, mothers' and fathers' educational levels were separately introduced (crude model). In Model 1, mothers' and fathers' educational levels was simultaneously introduced. In Model 2, child sex, gestational age, season of birth, parity, type of delivery, mother's and father's age, pre-pregnancy BMI, marital status, mother's and father's allergy, household income and mother's and father's smoking status breast milk, nursery, the number of lower respiratory infections, mold, pet at home, and passive smoke were added. Then, to confirm the combination exposure effects of the mother and father's educational levels, the combination variable was introduced in crude and full adjusted models (additional analysis).
Two-sided P-values <0.05 were considered statistically significant. All analyses were conducted using Stata statistical software version 15.0 for Windows (StataCorp, College Station, TX, USA).
# Results
Percentages of childhood wheezing and doctor-diagnosed asthma positives at 3-year were 17.4% and 7.6%, respectively. Both mother's and father's educational levels had statistically significant relationships with wheezing and doctor-diagnosed asthma in the pairwise deletion analyses (S2 and S3 Tables).
In crude general logistic regressions analysis of childhood wheezing after multiple imputation among mothers (S4 .displays the results of a multilevel analysis of the associations of mothers' and fathers' educational levels with childhood wheezing. In the crude models, the significance was the same as in the general logistic regression except for the father's EDC3 (fEDC3). In the final displays the results of a multilevel analysis of the associations of mothers' and fathers' educational level with doctor-diagnosed asthma. In the crude models, the significance was the
## Associations of mothers' and fathers' educational levels with childhood wheezing
## Associations of mothers' and fathers' educational level with doctordiagnosed asthma
## Associations of the combination of mother's and father's educational level with childhood wheezing
## Associations of the combination of mother's and father's educational level with doctor-diagnosed asthma
# Discussion
This prospective birth cohort study presented associations of lower educational level with increasing childhood wheezing and asthma, and higher educational level with decreasing them in crude analyses, and lower educational level of mothers was associated with increased childhood asthma in the fully adjusted model. However, contrary to expectations, higher educational level, especially the mother's, was associated with increasing childhood wheezing and asthma after adjusting for pre-and post-natal factors. This is the first study reporting that associations of the combination of mother' and father's educational levels on childhood wheezing and asthma in a large-scale prospective birth cohort. Since lower educational attainment is often linked to limited health literacy and undesirable health behavior, the deteriorative effect of lower educational attainment of mother, father, or parents, and the protective effect of higher educational attainment in crude analyses are reported. Thus, the results of childhood wheezing and asthma deteriorative effects of lower educational attainment of mothers and fathers, and the protective effect of higher educational attainment in crude analysis in this study are consistent with previous reports. Because early intervention in high-risk groups may reduce asthma development, health education for parents risk reduction such as smoking cessation, prevention of exposure to environmental tobacco smoke, should be performed from pregnancy.
In a fully adjusted model including pre-and post-natal factors, mother's EDC3 and 4 were significantly associated with childhood wheezing, and mothers' EDC1 and EDC3 with. Crude and adjusted odds ratios of the combination of mother's and father's education for doctor-diagnosed asthma in multilevel logistic regression analysis (multiple imputation, N = 69,067). childhood asthma. Furthermore, in the fully adjusted model introducing the combination of mother's and father's educational levels, mEDC3+fEDC1 and +fEDC4 were significantly associated with childhood wheezing, and mEDC3+fEDC3 with childhood asthma. Two studies evaluated the effect of mothers' and fathers' educations simultaneously reported that they had no statistically significant difference after pre-and post-natal factor adjustment. One possible reason for the discrepancy was the difference in the education category between our present study and two previous studies. One previous study categorized education levels as � junior high school and � high school or more (2 categories), the other as primary school, junior high school, high school, and university (4 categories). Thus, two categories of previous Japanese studies may lack detecting differences, which could be clear in 4 categorizations. Because educational level is mandatory through junior high school in Japan, the category in this study started from junior high school. Moreover, the age of childhood in those studies was older than ours (6-7 years and 10 years old). These factors may have caused the discrepancy. The reasons for higher childhood wheezing and asthma risk among mothers with higher educational attainment after pre-and post-natal factors adjustment could not be elucidated in this study. In the occupation setting, the mental health status of women may not be as good as that of men, and for women in Japan, advantaged occupational positions did not link to favorable psychological health or even linked to poor psychological health. However, one study reported that advantaged occupational positions among men had favorable psychological health. An association of maternal psychosocial stress with childhood asthma development was reported, and higher educational attainment has a greater chance to have advanced positions. Thus, the psychosocial stress of mothers with higher educational attainment may be one of the reasons for the higher childhood wheezing and asthma risks.
Another possible factor affecting the significance of mothers' higher educational level was the difference in reporting rate among mothers' educational levels. The significant association between mothers' junior high school education and childhood wheezing disappeared in Model 1, but the association with childhood asthma was retained until Model 2. Mothers with higher education more closely monitor their children's behaviors and are more health conscientious. Compared to doctor-diagnosed asthma, wheezing involves mild symptoms. This may have caused the report of childhood wheezing among mothers with junior high school education, and the association disappeared in Model 1.
There are several limitations to this study. First, many variables came from self-administered questionnaires. The outcomes may have been underreported, especially mild wheezing, as previously mentioned. Further, doctor-diagnosed asthma was defined based on a selfadministered questionnaire to caregivers. Among preschool children, wheezing is very heterogeneous, and the first manifestation of persistent asthma is clinically indistinguishable from transient wheezing. Therefore, doctor-diagnosed asthma was heterogeneous and may have contained transient wheezing. However, we believe doctor-diagnosed asthma was a significant indicator because it reflected severe or recurrent wheezing needing a hospital visit. Second, the generalization of our results to other country contexts may be limited because every nation has an original education system and different rates of entering higher educational institutions. Third, educational attainment could change after childbirth. This may cause educational level misclassification. Fourth, since 16 categories of the combination of the mother's and father's educational levels made small number categories including EDC1, the statistical power of those categories was reduced. Fifth, this birth cohort reflected Japanese pregnant women generally, but this analysis was limited to those who answered a 3-year questionnaire. Thus, the participants in this study tended to have higher health consciousness. Sixth, the adjustment was made by many known asthma risk factors in our present study, but all possible risk factors were not included (e.g., air pollution). However, major risk factors such as smoking, family history of allergy were included. Furthermore, because the concept of parental educational levels can contain disparities of health literacy, undesirable health behavior, and exposure to contaminants, adjustment of such factors may reduce the comprehensive effect of 'educational level' which suggests socio-economic disparity.
# Conclusions
The crude analysis in this prospective birth cohort study demonstrated that mothers' and fathers' lower education levels were associated with increased child wheezing and asthma, and higher educational level was associated with decreases in the two outcomes, though mECD3 had a significant crude higher OR for childhood asthma. Pre-and post-natal factor adjustment analysis also indicated that mothers' lower educational level was associated with increased childhood asthma. Parental educational level to reduce childhood asthma should target mothers and fathers with lower educational attainment. Furthermore, pre-and post-natal factor adjustment analysis, higher educational level, especially mothers', was associated with increasing childhood wheezing and asthma. Thus, there may be mediating factors that were not included in the full models. Further studies were needed to clarify how the mother's higher educational level affects childhood asthma, and to elucidate the later year effect of mothers' and fathers' educational levels on asthma, including hospitalization and medication.
## Supporting information
|
Progress on 3+1D Glasma simulations
We review our progress on 3+1D Glasma simulations to describe the earliest stages of heavy-ion collisions. In our simulations we include nuclei with finite longitudinal extent and describe the collision process as well as the evolution of the strongly interacting gluonic fields in the laboratory frame in 3+1 dimensions using the colored particle-in-cell method. This allows us to compute the 3+1 dimensional Glasma energy-momentum tensor, whose rapidity dependence can be compared to experimental pion multiplicity data from RHIC. An improved scheme cures the numerical Cherenkov instability and paves the way for simulations at higher energies used at LHC.
# Introduction
QCD matter under extreme temperatures and densities in the form of the quark-gluon plasma is experimentally accessible in relativistic heavy-ion collisions. The possibility to conduct these experiments for a wide range of collision energies and baryon chemical potentials allows for the successive exploration of the QCD phase diagram. The highest collision energies have been achieved at LHC and RHIC, and lower collision energies are being explored in the Beam Energy Scan programs of RHICand upcoming programs at GSI FAIRand JINR NICA. The matter created in such collisions is initially very far from an ideal thermodynamic equilibrium. With the experimental progress also an improved theoretical understanding of the collision process from first principles is desirable.
The Color Glass Condensate (CGC) frameworkprovides such a theoretical basis for describing nuclear matter at ultrarelativistic energies. It is a classical effective field theory where hard partons within the nuclei act as sources for a e-mail: [email protected] b e-mail: [email protected] (corresponding author) soft gluonic fields. In the simplest version, describing very large nuclei, the distribution of the color charges is given by the McLerran-Venugopalan (MV) model. The preequilibrium stage that is created right after the collision is characterized by longitudinal color flux tubes and has been termed the Glasma. In more sophisticated models like the IP-Glasma, the color charge distribution is based on fits to deep-inelastic-scattering data. In combination with a subsequent hydrodynamical evolution, the IP-Glasma model is able to correctly reproduce many observables, including azimuthal anisotropies or event-by-event multiplicity distributions. Nevertheless, the underlying Glasma evolution is commonly based on a boost-invariant formulationwhere incoming nuclei are assumed to be Lorentzcontracted to infinitesimally thin discs. This assumption is justified for observables close to mid-rapidity at very high energies, but is a severe conceptual limitation when studying rapidity-dependent quantities or collisions at lower energies.
It is possible to break boost invariance by introducing fluctuations on top of boost invariant background fields. Boost invariance is also broken by the JIMWLK evolution. Apart from recent attempts, Glasma simulations using JIMWLK-based initial conditions still have to be performed in an effectively boost invariant manner. The generated rapidity dependence of the Glasma can reproduce observables like gluonand charged hadron multiplicities. However, deviations from boost invariance may already arise at the classical level if one considers nuclei with finite extent in the beam direction. A three-dimensional formulation has been introduced by using an extended source that is not quite aligned with the light cone, however in a way that violates the covariant conservation condition at order g 2 . An alternative approach to include such finite extent corrections has been developed for proton-nucleus collisionswhich however is difficult to generalize to nucleusnucleus collisions.
In the following we review our progress towards an alternative approach of a 3+1D simulation for heavy-ion collisions including finite longitudinal extent of the nuclei [bib_ref] Simulations of the Glasma in 3+1D, Ipp [/bib_ref]. The loss of boost invariance requires us to keep track of the hard color sources throughout the subsequent evolution after the collision. This is achieved using the colored particlein-cell method (CPIC), which has been originally developed to study aspects of the evolution of the quark-gluon plasma. We apply this technique to study the collision process itself within the CGC framework. The simulation is performed in the laboratory frame and follows the nuclei throughout the collision process. Using this approach, we demonstrate that already a classical leading-order CGC simulation can give rise to a rapidity dependency consistent with experimental findings. Recent algorithmic developments will allow us to scrutinize our findings at even higher energies.
## The 2+1d glasma
From the viewpoint of the laboratory frame, a high energy nucleus is highly Lorentz-contracted along the beam axis and its dynamics are slowed down by time dilation. At collision energies available to RHIC and LHC, nuclei therefore appear to be almost infinitesimally thin, static discs moving at highly relativistic velocities. In CGC effective theorythe partons of high nuclei are split into hard and soft degrees of freedom. At leading order, hard partons are described in terms of classical color currents J μ and soft partons in terms of classical color fields A μ , whose dynamics are governed by the Yang-Mills (YM) field equations. For example, the classical color current associated with a nucleus moving along the negative x 3 = z axis (shown as nucleus "A" in is given by
[formula] J μ A (x) = δ μ − ρ a A (x + , x T )t a ,(1) [/formula]
where we have used light cone coordinates x ± = (x 0 ± x 3 )/ √ 2 and transverse coordinates x T = (x, y). The matrices t a are the traceless, hermitian generators of the color gauge group SU(N c ). The color charge density is given by ρ A (x + , x T ) and is assumed to be highly peaked around x + = 0 to account for the high Lorentz contraction of the nucleus. Additionally, ρ A (x + , x T ) is static in the sense that it does not explicitly depend on x − , which is a consequence of time dilation.
The color current J μ induces a classical color field A μ , which is to be determined from the non-linear YM equations.
[formula] D μ F μν = ∂ μ F μν + ig A μ , F μν = J ν ,(2) [/formula]
Nucleus A Nucleus B Glasma z y x Three-dimensional simulation of the collision of two nuclei. The distribution of the energy density between the two nuclei "A" and "B" right after the collision reveals the flux tube structure of the Glasma that develops between them. The simulation only covers a small part of the full collision in the transverse plane spanned by x and y. Adapted fromwhere D μ denotes the gauge covariant derivative. The non-Abelian field strength tensor is defined as
[formula] F μν = ∂ μ A ν − ∂ ν A μ + ig A μ , A ν ,(3) [/formula]
where g is the YM coupling constant. If J μ is given by Eqs.
(1), the YM Eqs.
(2) can be solved analytically by employing covariant gauge ∂ μ A μ = 0. In this gauge choice, the field Eqs. simplify to
[formula] −Δ T A − A (x) = ρ a A (x + , x T )t a ,(4) [/formula]
where Δ T is the two-dimensional Laplace operator in the transverse plane. The other components of A μ A (x) vanish. This can be formally solved by inverting the Laplace operator:
[formula] A μ A (x + , x T ) = δ μ − d 2 k T (2π) 2 ρ a A (x + , k T ) k 2 T + m 2 e ik T ·x T t a .(5) [/formula]
Infrared divergences are avoided by including a small regulator m, which dampens the long-range behavior of the color field. The length scale m −1 is usually identified with the confinement radius. It is also possible to regulate ultraviolet modes with a cutoff . The color field of the nucleus consists of purely transverse color-electric and -magnetic fields located near x + = 0. The color field of a high energy nucleus is therefore analogous to the Lorentz-boosted electromagnetic field of an ultrarelativistic charge. In CGC effective theory the color currents of nuclei are stochastic fields whose distribution is described by a probability functional W . Expectation values of observables (expressed as functionals of the color fields O[A μ ]) are computed by averaging over all possible realizations of ρ:
[formula] O = Dρ O[A μ [ρ]] W [ρ].(6) [/formula]
A simple and popular model for W [ρ] is the MV model, which assumes the functional to be Gaussian:
[formula] W [ρ] = 1 Z exp − d 2 x T dx + ρ a (x + , x T )ρ a (x + , x T ) 2g 2 μ 2 (x + , x T ) ,(7) [/formula]
where Z is a normalization constant and the function μ 2 (x + , x T ) describes the variance of the randomly distributed color charges. The description of nucleus "B" (see , which moves along the positive z axis is completely analogous.
A collision of two CGCs can be described by solving
[formula] D μ F μν (x) = J ν A (x) + J ν B (x),(8) [/formula]
where the source is given by the combined color current of the colliding nuclei. We assume nucleus "A" to be centered around x + = 0 and nucleus "B" to be centered around x − = 0 such that the collision occurs at x + = x − = 0. A Minkowski diagram of the collision scenario is shown in [fig_ref] Figure 2 A: Minkowksi diagram of a boost invariant heavy ion collision [/fig_ref]. In the boost invariant limit, the color charge densities become effectively two-dimensional due to Lorentz contraction:
[formula] x − x − x + x + x 0 = t x 0 = t x 3 = z x 3 = z (I) (I) (II) (II) (III) (III) (IV) (IV) ρ B (x T )δ(x − ) ρ B (x T )δ(x − ) ρ A (x T )δ(x + ) ρ A (x T )δ(x + ) A µ =0 A µ = 0 α i B = 1 ig V B ∂ i V † B α i B = 1 ig V B ∂ i V † B α i A = 1 ig V A ∂ i V † A α i A = 1 ig V A ∂ i V † A A µ (τ,x T ) A µ (τ,x T )ρ (A,B) (x) ≈ δ(x ± )ρ (A,B) (x T ).(9) [/formula]
This approximation is only correct in the limit of infinite collision energy. In this case, the space-time shown in [fig_ref] Figure 2 A: Minkowksi diagram of a boost invariant heavy ion collision [/fig_ref] separates into four distinct regions. Invoking causality, the solution A μ (x) to Eq. (8) in regions I -III is given by the superposition of the single nucleus solutions
[formula] A μ A (x) and A μ B (x). [/formula]
The solution in the future light cone, which describes the Glasma, generally does not exist in closed form and has to be determined perturbatively or numerically. In the boost invariant limit however, the gauge field can be determined along the boundary of the light cone. Using proper time τ = √ 2x + x − and (space-time) rapidity η s = ln x − /x + /2 and employing temporal gauge A τ = 0 for τ ≥ 0, the color field at τ = 0 + is given byA
[formula] i (τ = 0 + , x T ) = α i A (x) + α i B (x),(10)A η (τ = 0 + , x T ) = ig 2 α i A (x), α i B (x) .(11) [/formula]
Here, the color fields
[formula] α i (A,B) are the light cone (LC) gauge A ∓ = 0 solutions A i (A,B) (x ± , x T ) = θ(x ± )α i (A,B) (x T ),(12)α i (A,B) (x T ) = 1 ig V (A,B) (x)∂ i V † (A,B) (x),(13) [/formula]
where the lightlike Wilson lines are given by
[formula] V † (A,B) (x T ) = lim x ± →∞ V † (A,B) (x ± , x T ),(14)V † (A,B) (x ± , x T ) = P exp ig x ± −∞ dx ± A ∓ (A,B) (x ± , x T ) .(15) [/formula]
The initial conditions Eqs. (10) and (11) describe a highly anisotropic initial state consisting of purely longitudinal color-electric and -magnetic flux tubes. Since these initial conditions do not depend on rapidity η s , the Glasma and any observables remain boost invariant for τ > 0. For charge densities which are not δ-shaped as in Eq. (9) and instead have finite longitudinal length along x ± , there are no rigorous derivations of generalized initial conditions. In order to allow for charge densities which explicitly break boost invariance, we have to move to a fully 3+1 dimensional description of the Glasma. In particular, it is necessary to solve the YM Eqs.in a different way.
## Simulating the glasma in 3+1d
The motivation for relaxing Eq. (9) is to go beyond the approximation of infinite collisional energy and be able to describe observables such as the energy momentum tensor of the Glasma in a rapidity dependent manner. A simple generalization is given by
[formula] ρ (A,B) (x) ≈ λ(x ± )ρ (A,B) (x T ),(16) [/formula]
where λ is a normalized function, which determines the longitudinal shape of the charge density. The width of λ should be directly related to the Lorentz-contracted diameter of the nucleus. It should be noted that Eq. (16) is a special case where the color structure does not depend on the longitudinal coordinate x ± and more general color charge densities
[formula] ρ (A,B) (x ± , x T ) are possible as well. [/formula]
A direct consequence of allowing for finite longitudinal extent is that the collision event is not just a single point in the Minkowski diagram (see [fig_ref] Figure 2 A: Minkowksi diagram of a boost invariant heavy ion collision [/fig_ref] , but an extended spacetime region in which the nuclei are able to interact. The nonperturbative nature of the color fields of the nuclei and the extended interaction time generally make analytical calculations in this space-time region intractable. Our approach to the 3+1D Glasma is therefore to simulate not just the evolution in the future light cone, i.e. region IV in [fig_ref] Figure 2 A: Minkowksi diagram of a boost invariant heavy ion collision [/fig_ref] , but the whole collision. This setup is formulated in the laboratory frame using (t, z) coordinates and the time evolution is performed in the direction of t instead of proper time τ . The initial conditions of such a time evolution in t are specified in the following way: at some initial time t 0 < 0 sufficiently far away from the collision time t = 0, the color charge densities of the nuclei are non-overlapping in z. In LC gauge, the color field of each nucleus is given by
[formula] A i=x,y (A,B) (t 0 , x) = 1 ig V (AB) (t 0 , x)∂ i V † (AB) (t 0 , x),(17) [/formula]
where x = (x, y, z) is a three-dimensional coordinate vector. Equation (17) solves the classical YM Eqs. (2) with the current given by Eq. (16). Since the fields vanish exponentially fast between the two nuclei, the superposition of both color fields
[formula] A μ (t 0 , x) = δ μ i=x,y A i A (t 0 , x) + A i B (t 0 , x) ,(18) [/formula]
is a valid solution to the YM Eqs. (8) at time t 0 . Evolving this initial condition for the YM field to t > 0 yields a genuinely 3+1D description of the rapidity dependent Glasma. Equations (17) and (18) are compatible with the temporal gauge condition A 0 (t, x) = 0, ∀t ∈ R, which is a convenient gauge choice for an evolution along x 0 = t.
One of the main differences to the standard 2+1D Glasma is that in the laboratory frame the system not only consists of the color fields but also the color currents of the nuclei. In order to solve the YM Eqs. which include color currents, we make use of the CPIC method. CPIC allows for consistent simulations of color charged point particles coupled to color fields on a lattice. For a comprehensive description of our numerical methods we refer to .
The numerical treatment of the fields follows the common real-time lattice gauge theory approach: by discretizing Minkowski space-time as a hypercubic lattice with spacings a i and time-step Δt, the YM Eqs. can be written in the standard leapfrog scheme
[formula] E i (t + Δt 2 , x) = − j Δt (a j ) 2 U i, j (t, x) + U i,− j (t, x) ah +Δt j i (t, x) + E i (t − Δt 2 , x),(19)U i (t + Δt, x) = exp iΔt E i (t + Δt 2 , x) U i (t, x),(20) [/formula]
where is the chromo-electric field on the lattice and [X ] ah denotes the anti-hermitian, traceless part of a matrix X . The plaquette variables U i, j (t, x) are defined as
[formula] U i (t, x) exp iga i A i (t, x) are the gauge links, E i (tU i, j (t, x) = U i (t, x)U j (t, x +â i )U † i (t, x +â i )U † j (t, x).(21) [/formula]
The color currents of the nuclei j i (t, x), which enter on the right hand side of Eq., require careful treatment. The main idea of using the CPIC method to describe collisions in the CGC framework is to replace the continuous color charge distributions ρ of the nuclei by a large number of auxiliary particles with time-dependent color charges Q k (t) such that the original color charge distribution is sufficiently well approximated on a lattice:
[formula] ρ(t, x) ≈ k Q k (t)δ (3) (x − x k (t)),(22) [/formula]
where k is the particle index. Similar to the boost invariant case, we assume these auxiliary particles to be recoil-less. Thus, the trajectories of the particles x k (t) are fixed and not part of the dynamics of the system. The time-dependence of the color charges Q k (t) is determined from the discretized continuity Eq. Adapted from [bib_ref] Simulations of the Glasma in 3+1D, Ipp [/bib_ref] which is the discrete analogue of gauge covariant continuity equation
[formula] ρ(t + Δt 2 , x) − ρ(t − Δt 2 , x) Δt = i j i (t, x) − U † i (t, x −â i ) j i (t, x −â i )U i (t, x −â i ) a i ,(23)x x x+â i +â j x+â i +â j x+â i x+â i x+â j x+â j (a) Standard plaquette term x x x+Δt x+Δt x−Δt x−Δt x+â i +â j x+â i +â j x+Δt+â i +â j x+Δt+â i +â j x−Δt+â i +â j x−Δt+â i +â j x+Δt+â i x+Δt+â i x−Δt+â j x−Δt+â j (b) Time-averaged plaquette-like termD μ J μ (t, x) = 0.(24) [/formula]
In our setup, the color charge Q k (t) of each particle is mapped to its nearest grid point on the spatial lattice in each time-step. Whenever the nearest grid point of a particular point charge changes from one lattice site x to a neighbouring lattice site y, parallel transport is applied to the color charge accordingly:
[formula] Q k t + Δt 2 = U † x→y (t)Q k t − Δt 2 U x→y (t),(25) [/formula]
where U x→y (t) is the appropriate gauge link connecting x and y. At the same time, the movement of the particle generates a color current j i (t, x) in accordance with Eq. (23). In CPIC, this treatment of particles is known as the nearestgrid-point scheme. By evolving the color charges of the particles in this manner, the discretized field Eqs.andare solved consistently in the sense that the discrete Gauss law
[formula] i E i (t + Δt 2 , x) −Ẽ i (t + Δt 2 , x −â i ) = ρ(t + Δt 2 , x),(26) [/formula]
remains satisfied throughout the simulation and gauge covariance on the lattice is retained. In Eq. (26) the parallel transported electric field is given byẼ
[formula] i t + Δt 2 , x −â i = U † i t, x −â i E i t + Δt 2 , x −â i U i t, x −â i .(27) [/formula]
We find that numerically stable 3+1D Glasma simulations using the leapfrog scheme Eqs.and (20) require high lattice resolution with particularly fine lattice spacing along the beam axis z. In practice, this can be computationally prohibitive. This problem is related to a subtle numerical instability inherent to the leapfrog scheme known as the numerical Cherenkov instability, which also affects traditional electromagnetic plasma simulations. This unphysical instability is caused by lattice artifacts which modify the propagation of wave modes on the lattice: high frequency modes propagate at a lower phase velocity compared to low frequency modes (i.e. numerical dispersion). In contrast, the auxiliary particles move at the speed of light by design. This situation is reminiscent of charged particles moving through a medium in which the in-medium speed of light is lower than the particle velocity, which leads to the well-known phenomenon of Cherenkov radiation. The particular lattice discretization used in Eqs.andleads to a similar, although purely numerical generation of Cherenkov radiation. Due to the numerical instability the nuclei are not able to propagate stably unless a very small lattice spacing is chosen along z, which reduces the mismatch in propagation velocities. Fortunately, these numerical problems can be solved by modifying the lattice discretization of the color fields. In [bib_ref] Simulations of the Glasma in 3+1D, Ipp [/bib_ref] we show how an improved numerical scheme restores the correct propagation of wave modes along the beam axis such that artificial Cherenkov radiation is effectively avoided. This modification mainly amounts to replacing specific spa-tial plaquette terms (see [fig_ref] Figure 3: Two schematic diagrams of terms used in the lattice discretization of the... [/fig_ref] with time-averaged generalizations of these terms. [fig_ref] Figure 3: Two schematic diagrams of terms used in the lattice discretization of the... [/fig_ref] shows one example of such a time-averaged term. In comparison to the leapfrog scheme Eqs.and, which is an explicit finite difference method, our numerical method is in the form of a system of semi-implicit equations, which is solved in an iterative manner. Our semi-implicit method therefore trades computational performance for numerical stability and accuracy.
# Results
Here we review the most important results that have been obtained from simulations of collisions of nuclei with finite longitudinal extent using the standard leapfrog scheme. In these simulations we use initial conditions of the form Eq., where λ is a Gaussian of width L along z. We also assume μ 2 to be constant in the transverse plane. The longitudinal thickness L has been set to L = m N R/ √ s N N with collision energy √ s N N , nuclear radius R and nucleon mass m N ≈ 1GeV. The saturation momentum Q s grows with collision energy as Q 2 s ≈ √ s N N 0.25 GeV 2. The MV model parameter μ can be determined from 0.75 g 2 μ Q s with coupling g ≈ 2. The ultraviolet modes are regulated by = 10 GeV, and we vary the infrared regulator m in the range from 0.2 to 0.8 GeV to check for its dependency. The simulations presented use the gauge group SU(2) instead of SU(3), which should give qualitatively comparable results. The simulations have been performed on a lattice with 2048 × 192 2 cells with finer resolution along the longitudinal direction. The simulation box corresponds to a volume of (6 fm).
The main observables can be obtained from the components of the energy-momentum tensor T μν which are extracted from the gluon fields of the simulation. We average over 15 collision events to obtain the expectation value T μν . Due to the symmetries of the MV model, certain components vanish, and the remaining contributions of the averaged energy-momentum tensor are given by
[formula] T μν = ⎛ ⎜ ⎜ ⎝ ε 0 0 S L 0 p T 0 0 0 0 p T 0 S L 0 0 p L ⎞ ⎟ ⎟ ⎠ ,(28) [/formula]
where ε is the energy density in the laboratory frame, p L and p T are the longitudinal and transverse pressure components and S L is the longitudinal component of the Poynting vector. The energy density ε as given in the laboratory frame is depicted in . By diagonalizing the energy-momentum tensor, we obtain the local rest frame energy density ε loc , which can be expressed in terms of proper time τ = √ t 2 − z 2 and space-time rapidity η s = ln [(t − z) / (t + z)] /2. The black solid lines show the rapidity profiles as extracted from of our simulations, which can be fitted to a Gaussian shape (dashed continuing lines). The profiles have been extracted at τ 0 = 1 fm/c where the Glasma turns into the QGP. Already at times τ 0 0.3 fm/c, the system enters a free-streaming evolution with longitudinal velocity v z ≈ z/t and the shape of the profile does not change anymore. The width of the profiles depends on the energy √ s N N and becomes flatter with increasing energy as expected from the recovery of boost invariance at higher energies. We also find a strong dependency on the infrared regulator m, where higher values of m make the rapidity profiles narrower. While this strong dependence on the infrared regulator seems unexpected, it may indicate that the screening length λ D ∝ 1/m plays an important role in generating a deviation from boost invariance. It is not only the longitudinal thickness L, but the dimensionless ratio L/λ D , which seems to determine the shape of the profiles. Interestingly, the simulation results agree well with measured rapidity profiles of pion multiplicities at RHICwhich are indicated by the blue data points in . At these energies, the experimental results also agree with the Gaussian rapidity profile as obtained by the hydrodynamic Landau modelwhere the width is given by σ Landau = √ ln γ with the Lorentz gamma factor γ . For other particle species, the experimental momentum rapidity distribution of particle multiplicities deviates from the Landau picture but can still be fitted to a Gaussian distribution with slightly larger width. However, it should be noted that especially at higher energies the Landau model predicts rapidity profiles which are too narrow as measured by the ALICE collaboration. Whether our simulations of the 3+1D Glasma can describe these wider profiles at LHC energies will be the topic of future work.
It is important to note that shows the rapidity profiles of two different quantities: the local energy density ε loc (τ, η s ) as a function of space-time rapidity η s and the distribution of particle multiplicities d N/dy as a function of momentum rapidity y. This comparison is justified because our simulations showthat the 3+1D Glasma settles into a state of free-streaming flow v z ≈ z/t and vanishing longitudinal pressure p L ≈ 0, similar to the 2+1D Glasma. Freestreaming flow implies that -ignoring a subsequent hydrodynamical phase -we can identify momentum rapidity y with space-time rapidity η s , similar to the case of the Bjorken model. It also implies that T μν becomes diagonal in the (τ, η s ) frame. Due to vanishing longitudinal pressure, the energy-momentum tensor in the rest frame is anisotropic and reads
[formula] T μν = ⎛ ⎜ ⎜ ⎝ ε loc 0 0 0 0 ε loc /2 0 0 0 0 ε loc /2 0 0 0 0 0 ⎞ ⎟ ⎟ ⎠ ,(29) [/formula]
where we used 2 p T = ε loc due to conformal symmetry. In contrast to the Bjorken model, there are a priori no restrictions on the rapidity dependence of the energy density ε loc (τ, η s ). This can be checked using energy-momentum conservation
[formula] ∇ μ T μν = 0,(30) [/formula]
which simply reduces to
[formula] ∂ τ ε loc (τ, η s ) = −ε loc (τ, η s )/τ.(31) [/formula]
Equation (31) leads to ε loc ∝ 1/τ , but does not impose any additional constraints on the rapidity dependence of the energy density. On the other hand, the Bjorken model requires that T μν is isotropic in the rest frame, which combined with Eq. (30), then leads to boost invariance. As a first approximation, it is therefore reasonable to compare the space-time rapidity profile of ε loc to the momentum rapidity profile of charged particles as shown in . For a more quantitative comparison, our results should be used as input for a subsequent hydrodynamic simulation, which may slightly increase the width of the profiles.
To better understand how rapidity dependence of observables develops in our simulations, we can look at the
[formula] t [fm/c] z [fm] p T (t,z)/p T (0, 0) Fig. 5 [/formula]
Space-time distribution of the normalized transverse pressure p T (t, z) / p T (0, 0). The parameters are the same as in with m = 0.2 GeV. The transverse pressure corresponds to longitudinal chromo-magnetic and -electric fields and thus to the the longitudinal component of the energy density ε L (t, z) . Contrary to the boost-invariant case, this quantity falls off steeply along the boundary of the light cone transverse pressure p T (z, t) in . From the energymomentum tensor, one finds that the transverse pressure is linked to the longitudinal fields of the Glasma, whereas longitudinal pressure involves both, transverse and longitudinal field components
[formula] p T = 1 2 E 2 L + B 2 L ,(32)p L = 1 2 E 2 T + B 2 T − E 2 L − B 2 L ,(33) [/formula]
where the square implies a summation over color indices. In the boost invariant case, the initial conditions of the Glasma are specified at τ = 0. This corresponds to the boundary of the forward light cone, where the longitudinal fields would be constant. In contrast, in our simulation the longitudinal fields are peaked around the collision region at t ∼ z ∼ 0 and decrease rather quickly along the light cone boundaries. Accordingly, there is less Glasma being produced at larger values of rapidity η s which produces the Gaussian profiles. It is interesting to see that our weak coupling results are in qualitative agreement with strong coupling results from holographic models of heavy-ion collisions that also exhibit similar transverse pressure distributions.
# Conclusions and outlook
In this paper we reviewed our progress on 3+1D Glasma simulations. Our simulations allow to explore the creation of the Glasma in heavy-ion collisions beyond the commonly assumed boost-invariant case. We do this by introducing a finite longitudinal extent for the incoming nuclei corresponding to realistic Lorentz contractions as found for example at RHIC. Without the usual simplifications of boost-invariance, we have to keep the color currents of the hard partons in the Glasma simulation. This is achieved using CPIC in the lab-oratory frame. Using the MV model, we demonstrated that our approach can give rise to Gaussian rapidity profiles in the energy density. These profiles depend on the energy of the incoming nuclei, but also on an infrared regulator. For energies used at RHIC we obtain qualitative agreement of these profiles. This is remarkable as it shows that boost invariance can be broken already at the classical level if the longitudinal structure is properly taken into account. This nicely complements findings from holographic models where similar profiles can be found. Algorithmic improvements in the form of a new semiimplicit solver [bib_ref] Simulations of the Glasma in 3+1D, Ipp [/bib_ref] will allow for further explorations of our boost-invariance breaking simulations. By modifying the standard Wilson gauge action we achieve a dispersion-free propagation along the longitudinal direction which cures the numerical Cherenkov instability which has plagued previous simulations. This sets the basis for more accurate and larger simulations valid for larger ranges of rapidity which are necessary for a comparison of rapidity profiles at LHC collision energies. One crucial aspect that shall be studied in this context is the role of longitudinal color fluctuations. These are usually approximated as an infinitely thin stack of uncorrelated sheets of color charge. Dispersion-free propagation will allow for the fine-grained simulation of collisions and the study of the effect of internal longitudinal color structures on the creation of the Glasma. In principle, it should be straightforward to also include more realistic sub-nucleonic color structure in the transverse direction as is the case in the IP-Glasma model. Here, one is essentially limited by the large computational requirements of such three-dimensional simulations.
On a more conceptual level, it would be interesting to better understand the relation between the boost-invariance breaking that we find at the leading classical order and a similar breaking that can be found at next-to-leading order from the JIMWLK evolution. It would be a highly desirable but presumably very non-trivial task to generalize the JIMWLK evolution Eqs. to be applicable to threedimensional color distributions. Another extension to our work would be the deviation from the eikonal approximation and the inclusion of dynamical colored particles. This could also be a way to accommodate three-dimensional extensions of the calculation of quantities like energy loss or momentum broadening from the Glasma.
[fig] Figure 2 A: Minkowksi diagram of a boost invariant heavy ion collision. The colliding nuclei move along x + and x − (red arrows) and are assumed to be infinitesimally thin along the beam axis z. The gray hyperbolas in region IV are contour lines of constant proper time τ and the straight dashed lines are lines of constant rapidity η s . Adapted from [39] [/fig]
[fig] Figure 3: Two schematic diagrams of terms used in the lattice discretization of the YM equations: the standard spatial plaquette term (a) as defined in Eq.(21), which is used in the leapfrog scheme, and a timeaveraged generalization of this term (b), which is used in our semi-implicit method. While (a) only contains gauge links defined at a common equal-time slice of discretized Minkowski space, the generalized term (b) includes terms from both future and past equal-time slices. [/fig]
[fig] Figure 4 15, Figure 4: Comparison between simulation and experimental results for the space-time rapidity profile[36]. The thick black solid lines show simulation results of the local rest frame energy density ε loc (τ 0 , η s ) atτ 0 = 1 fm/c for various values of the infrared regulator m. The following widths of the Gaussian profiles have been extracted: (a) m = 0.2 GeV with σ η = 2.34, (b) m = 0.4 GeV with σ η = 1.66 and (c) m = 0.8 GeV with σ η = 1.28. For comparison, the experimentally obtained profile of π + multiplicity d N/dy at RHIC is given by the blue data points [51], with a width of σ exp = 2.25. The thin red line corresponds to the profile predicted by the Landau model with σ Landau = √ ln γ ≈ 2.shows the local energy density ε loc (τ 0 , η s ) as a function of space-time rapidity η s for √ s N N = 200 GeV. [/fig]
|
Unusual presentation in Haiti of a recurrent giant cell tumor of bone affecting the distal radius: A case report
# Introduction
Cooper and Travers first described giant cell tumor of bone (GCTB) in 1818. GCTB represents 4-5% of all bone tumors, and 20% of all benign bone tumors [bib_ref] Giant cell tumor of the bone, Athanasou [/bib_ref] [bib_ref] Giant-cell tumor: a study of 195 cases, Dahlin [/bib_ref]. It is a benign aggressive tumor most common in young adults between 20 and 40 years of age and more prevalent in females [bib_ref] Local recurrence of GCTB after intralesional treatment with and without adjuvant therapy, Knochentumoren [/bib_ref]. GCTB typically arises in the metaphysis and extends into the epiphysis [bib_ref] Giant cells tumor of the bone: treatment and outcome of 214 cases, Balke [/bib_ref].
Patients present with symptoms such as pain, swelling, joint effusion and limited range of motion (ROM). Local recurrence is strongly increased by curettage alone or with soft tissue extension, which is present in 20-25% of all GCTB [bib_ref] Giant cell tumor of bone: risk factors of recurrence, Klenke [/bib_ref]. Treatment options include en bloc resection, curettage with physical and chemical adjuvants, and denosumab [bib_ref] Giant cell tumor of bone: risk factors of recurrence, Klenke [/bib_ref] [bib_ref] Denosumab treatment for giant-cell tumor of bone: a systematic review of the..., Luengo-Alonso [/bib_ref].
We present a case of a recurrent Campanacci Grade II GCTB of the right distal radius in a right-hand dominant female patient from Haiti [bib_ref] Giant-cell tumor of the bone, Campanacci [/bib_ref]. The patient was initially evaluated at Hôpital Bernard Mevs in Port au Prince and her surgery was performed at the Hôpital Adventiste d'Haïti in Carrefour. The purpose of this case is to demonstrate that complex limb-salvage surgery is possible in underserved regions of the world when care is coordinated between an international and local team. The importance of adequate follow-up and post-operative care with continued com- . Multiple areas of the skin had breakdown, but no tumor was exposed. The volar and ulnar skin was spared. munication between the teams is essential. This work has been reported in line with the SCARE criteria [bib_ref] Guideline: updating consensus surgical CAse REport (SCARE) guidelines, Agha [/bib_ref].
## Presentation of case
A 22-year-old female presented with a recurrent right wrist mass treated surgically three years prior to presentation. She had undergone a curettage resection without adjuvant and placement of iliac crest autograft for a GCTB of the distal radius. The mass had recurred over a period of 2.5 years and was 18 × 25 cm at the time of presentation. The mass involved the majority of the wrist and only the volar and ulnar aspects were spared. There was some superficial skin breakdown, but no tumor was exposed [fig_ref] Figure 1: A-C Clinical images of the patient after recurrence of the tumor in... [/fig_ref]. The mass was firm, non-mobile and minimally tender. She had full active finger ROM and thumb flexion, but lacked extensor pollicis longus (EPL), abductor pollicis longus (APL) and extensor pollicis brevis (EPB) function (Video 1). Her wrist dorsiflexion was 20 degrees and her palmar flexion was 40 degrees. Her supination was 55 degrees and pronation was 80 degrees. Radial and ulnar deviation was 10 and 15 degrees, respectively. She lacked sensation over her superficial radial nerve, but otherwise had full sensation. She had 2+ radial and ulnar pulses and a normal Allen's test.
Radiographs demonstrated an expansile, lytic lesion of the distal radius with internal trabeculations with a thin cortical rim. There was encasement of the proximal carpus with dislocation of the distal ulna [fig_ref] Figure 2: A-B Initial anteroposterior [/fig_ref]. Although an MRI was unable to be obtained in Haiti, a CT scan was available which demonstrated that the volar structures were free, whereas the EPL, APL and EPB went directly into the tumor. The extensor tendons to the fingers were traversed through the tumor but were felt to be uninvolved based on the patient's clinical exam. A chest CT was negative for metastatic disease. The patients care was then coordinated with an international team. A biopsy was obtained two weeks after presentation under direction of an orthopaedic oncologist. The specimen was evaluated in the United States and was consistent with recurrent GCTB. One month after the biopsy, an international surgical team traveled to Haiti to perform the definitive surgery.
Limb-salvage surgery was performed by a fellowship-trained orthopaedic oncologist (LZ) and hand surgeon (JD) with over thirty years of combined surgical experience. They both had made multiple trips to Haiti to perform surgery and had previously coordinated post-operative care with the local orthopaedic surgeons. A local Anesthesiologist used a propofol drip to provide general anesthesia and the patient was given 1 g of cefazolin pre-operatively. An en bloc resection of the distal radius with the proximal row of the carpus was performed. The EPL, APL, EPB, flexor carpi radialis, extensor carpi radialis longus and brevis and superficial radial nerve were resected proximal to the tumor due to tumor involvement. The tumor had grown around the extensor tendons, which were freely mobile and not involved. The tendons were cut distal to the tumor and were easily pulled through the tunnel that had formed for later repair. The volar structures were free of tumor and were preserved, and the tumor was able to be resected en bloc [fig_ref] Figure 3: A-D Intraoperative photos after resection of the tumor but prior to reconstruction [/fig_ref]. A centralization of the ulna was then performed with fusion to the capitate and third metacarpal using an Acumed (Hillsboro, OR) wrist fusion plate [fig_ref] Figure 3: A-D Intraoperative photos after resection of the tumor but prior to reconstruction [/fig_ref]. The extensor indicis proprius was transferred to the EPL to restore thumb extension. The remaining tendons were redundant due to being stretched out by the tumor. The flexor tendons were cut, and both the flexor and extensor tendons were shortened and repaired side-to-side using a Krakow suture technique with appropriate tension [fig_ref] Figure 3: A-D Intraoperative photos after resection of the tumor but prior to reconstruction [/fig_ref].
The skin was closed with a tension-free repair and a bulky, sterile dressing and volar wrist splint was placed [fig_ref] Figure 3: A-D Intraoperative photos after resection of the tumor but prior to reconstruction [/fig_ref]. Histopathologic examination in the United States confirmed a GCTB without evidence of malignant transformation and negative margins [fig_ref] Figure 4: A-D and 4-E After resection of the tumor, radial and volar [/fig_ref].
Passive ROM of all digits was initiated immediately after surgery. At 3 weeks gentle active and active-assisted ROM was allowed and formal rehabilitation therapy started at 2 months. The patient was monitored for recurrence and metastatic spread every three months for the first two years with a physical examination and radiographs of both the forearm and chest. She will then be followed every six months until the fifth year.
At 3 months, the patient lacked 5 degrees of extension of the fingers. She had full flexion of the proximal and distal interphalangeal joints of the fingers but lacked flexion at the thumb interphalangeal joint. She had approximately 5 degrees of flexion at the metacarpophalangeal joints and was able to grip and pinch [fig_ref] Figure 5: A-C Three months after surgery the patient had full flexion of her... [/fig_ref]. She had 45 degrees of functional supination and 80 degrees of functional pronation by utilizing her shoulder. There was early evidence of consolidation of the fusion and her weight bearing was progressed to as tolerated [fig_ref] Figure 5: A-C Three months after surgery the patient had full flexion of her... [/fig_ref]. Two years after surgery the patient had no evidence of recurrent or metastatic disease, was pain free, and had regained full extension of her fingers and maintained her flexion (Video 2). The patient feels her life is back to normal and she is planning on attending school to become a nurse.
# Discussion
GCTB involving the distal radius in underserved regions requires a focused medical decision to ensure the lowest possibility of recurrence and to restore as much function as possible. Surgical techniques include intralesional curettage with use of adjuvants, en bloc resection, and use of denosumab [bib_ref] Denosumab treatment for giant-cell tumor of bone: a systematic review of the..., Luengo-Alonso [/bib_ref] [bib_ref] Giant cell tumor of the extremity: a review of 349 cases from..., Errani [/bib_ref]. Reconstruction options for an en bloc resection include a vascularized or non-vascularized fibula, iliac crest autograft, and intercalary or osteoarticular allograft [bib_ref] Clinical characteristics and risk factors analysis of lung metastasis of benign giant..., Yang [/bib_ref]. Allograft is currently unavailable in Haiti and use of autograft would increase blood loss and surgical time in this resource-limited environment. Therefore, a translocation or centralization of the ulna would be optimal.
Translocation of the distal ulna into areas of a radial defect was described by Hey-Groves in 1922 for fractures of the distal radius with a severe defect. In 1934, Greenwood [bib_ref] Reconstruction of the forearm after loss of radius, Greenwood [/bib_ref] and Watson-Jones [bib_ref] Reconstruction of the forearm after loss of the radius, Watson-Jones [/bib_ref] described severing the distal ulna and transferring the ulna shaft to the radius, creating a one-bone forearm. More recently and specifically for recurrent Campanacci Grade III GCTB, Seradge [bib_ref] Distal ulnar translocation in the treatment of giant-cell tumors of the distal..., Seradge [/bib_ref] in 1982, and Puri et al. [bib_ref] Ulnar translocation after excision of a Campanacci grade-3 giant-cell tumour of the..., Puri [/bib_ref] in 2010 described a standard technique for wide resection of the distal radius with translocation of the ulna onto the remaining radius. Bhan et al. [bib_ref] Ulnar translocation after excision of giant cell tumour of distal radius, Bhan [/bib_ref] and Puri et al. [bib_ref] Ulnar translocation after excision of a Campanacci grade-3 giant-cell tumour of the..., Puri [/bib_ref] have observed the risk of non-union of the ulnoradial interface with translocation.
Centralization of the ulna to the carpus was first described by Sayre [bib_ref] A contribution to the study of club-hand, Sayre [/bib_ref] in 1893 and modified by Lidge [bib_ref] Congenital radial deficient club hand. in: proceedings of the american academy of..., Lidge [/bib_ref] in 1969. This proce-dure was initially indicated for Type 3 or Type 4 radial longitudinal deficiencies. More recently Singh et al. [bib_ref] Wide resection and ulnar centralization for campanacci grade 3 giant cell tumour..., Singh [/bib_ref] described centralization of the ulna for use in Grade III GCTB with good functional results.
In this case, a centralization of the ulna was chosen for reconstruction. The main advantages to this procedure include high rates of union and obviating the risk of non-union at a second site. Additionally, this procedure avoided further dissection and a longer procedure which would be necessary for both an ulnar translocation and autograft reconstruction [bib_ref] Reconstruction of the forearm after loss of the radius, Watson-Jones [/bib_ref] [bib_ref] Carpus translocation into the ipsilateral ulna for distal radius recurrence giant cell..., Ververidis [/bib_ref] [bib_ref] Retrospective analysis of giant cell tumor lower end radius treated with en..., Vyas [/bib_ref]. The main disadvantage is loss of pronation and supination at the wrist by virtue of the ulno-humeral joint lacking rotation. At two years, the patient is able to compensate for her loss of motion through her shoulder and has a good functional result.
# Conclusion
This case supports other reports that centralization of the ulna is an effective method of reconstruction after en bloc resection of the distal radius due to GCTB. When resources are limited, treatment options should be carefully considered in order to optimize outcomes. With adequate coordination and bidirectional collaboration, complex limb-salvage surgery can be successfully conducted in low-and middle-income countries like Haiti. The importance of continued collaboration and communication with the local team is essential in order to manage complications and provide proper follow-up.
# Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
# Ethical approval
This case report is exempt from ethical approval at our institution.
## Consent
Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.
## Registration of research studies
1. Name of the registry: Research Registry 2. Unique identifying number or registration ID: researchreg-istry6403 3. Hyperlink to your specific registration (must be publicly accessible and will be checked): https://www.researchregistry. com/browse-the-registry#home/registrationdetails/ 5fe7e690fe46e3001b6cce62/
[fig] Figure 1: A-C Clinical images of the patient after recurrence of the tumor in pronation (A), neutral (B) and supination (C) [/fig]
[fig] Figure 2: A-B Initial anteroposterior (A) and lateral (B) radiographs demonstrating involvement of the distal radius with encasement of the proximal carpus and dislocation of the ulna. The expansile, lytic lesion has multiple internal trabeculations with a thin cortical rim. [/fig]
[fig] Figure 3: A-D Intraoperative photos after resection of the tumor but prior to reconstruction (A), after centralization of the ulna with placement of a wrist fusion plate (B), after reconstruction of the tendons (C) and after wound closure (D). [/fig]
[fig] Figure 4: A-D and 4-E After resection of the tumor, radial and volar (A) and dorsal and ulnar (B) aspects of the resected specimen are noted with no evidence of exposed tumor. A sagittal view of the tumor during pathologic examination (C) demonstrates cystic areas inside the tumor. The cut radius is identified at the bottom of the photos. Low-power (D) and high-power (E) histopathologic examination of the resected specimen demonstrates multiple multi-nucleated osteoclast-like giant cells in a background of mononuclear cells. [/fig]
[fig] Figure 5: A-C Three months after surgery the patient had full flexion of her proximal and distal interphalangeal joints of her fingers with the ability to pinch (A), and near full extension of her fingers (B). Anteroposterior and lateral (C) radiographs demonstrate no evidence of recurrent tumor or hardware failure and early consolidation of the fusion. [/fig]
|
Cereal products derived from wheat, sorghum, rice and oats alter the infant gut microbiota in vitro
## Figure s1
Family level taxonomic composition of the initial (at 0 hours) gut microbiota. The 12 relative abundance at the family level was determined using QIIME and GraphPad Prism 13 (V7). Bacterial identifications that were not assigned to a family are categorised as 14 "Unassigned". Bacterial groups with less than 2% relative abundance in all biological 15 samples (1-6) are categorised as "Other".. Bellamy's organic baby rice cereal and Real good food-Organic baby oat cereal. NP-not 47 provided.
48 The relative abundance of bacterial families that were found to be significantly
|
How I Treat TP53-Mutated Acute Myeloid Leukemia and Myelodysplastic Syndromes
# Introduction
Tumor protein p53 (TP53), a tumor suppressor gene, is the most frequently mutated gene in cancer [bib_ref] Mutational Landscape and Significance across 12 Major Cancer Types, Kandoth [/bib_ref]. TP53 is the "guardian of the genome" via several functions such as regulation of metabolic functions, apoptotic pathways, cellular senescence, and DNA repair [bib_ref] TP53 Mutations in Acute Myeloid Leukemia: Still a Daunting Challenge? Front, Molica [/bib_ref]. Myelodysplastic Syndromes (MDS) and Acute Myeloid Leukemia (AML) with TP53 mutation generally have poor prognosis, almost irrespective of the treatment administered, including after allogeneic Hematopoietic Stem Cell Transplantation (aHSCT). Due to their specific characteristics, experts in the fields have recently argued that TP53-mutated AML/MDS require special consideration [bib_ref] Molecular Characterization of Mutant TP53 Acute Myeloid Leukemia and High-Risk Myelodysplastic Syndrome, Grob [/bib_ref] [bib_ref] Comparison of Acute Myeloid Leukemia and Myelodysplastic Syndromes with TP53 Aberrations, Dutta [/bib_ref] [bib_ref] Personalized Medicine for TP53 Mutated Myelodysplastic Syndromes and Acute Myeloid Leukemia, Cluzeau [/bib_ref].
## Landscape of tp53 mutations
An analysis of the TP53 whole coding sequence found that about 85% of mutations were clustered between codons 125 and 300, which is the location of the DNA-binding domain. The mutation types are mostly missense (90%). Outside this region, the majority of mutations are nonsense or frameshift.
Hotspot mutations are localized in the DNA-binding domain and R175H is one of the most frequent hotspot mutants [bib_ref] Quantitative Analysis of Residual Folding and DNA Binding in Mutant P53 Core..., Bullock [/bib_ref]. The majority of TP53 mutations have an impact on protein structure and are classified as "contact" (R248 and R273) or "structural" (R175, G245, R249, and R282) [bib_ref] Structures of P53 Cancer Mutants and Mechanism of Rescue by Second-Site Suppressor..., Joerger [/bib_ref]. These mutations are associated with a loss of function.
In its target genes p53 also modulate transcription by interfering with so called response elements (REs) in the promoters or introns [bib_ref] Transcriptional Control of Human P53-Regulated Genes, Riley [/bib_ref]. As a consequence, a loss of function has been reported in all the mutants.
Finally, several hotspot mutants have been shown to impact cellular transformation [bib_ref] Mutant P53 DNA Clones from Human Colon Carcinomas Cooperate with Ras in..., Hinds [/bib_ref] , defining gain-of-function (GOF) mutations. Mutant p53 proteins may also exert dominantnegative effects (DNE) over wild-type p53.
## Tp53 mutations in aml and mds
TP53 mutations are seen in 5-10% of de novo MDS and AML patients, 25-40% of therapy-related MDS and AML patients [bib_ref] Clinical Effect of Point Mutations in Myelodysplastic Syndromes, Bejar [/bib_ref] [bib_ref] Clinical and Biological Implications of Driver Mutations in Myelodysplastic Syndromes, Papaemmanuil [/bib_ref] , and 50% of patients with a complex karyotype [bib_ref] TP53 Mutation Status Divides Myelodysplastic Syndromes with Complex Karyotypes into Distinct Prognostic..., Haase [/bib_ref]. Complex karyotypes are usually monosomal, very frequently include 5q deletion, and generally include 17p deletion, the latter being sometimes difficult to detect in complex cytogenetic rearrangements. They largely predominate in elderly patients [bib_ref] Implications of TP53 Allelic State for Genome Stability, Clinical Presentation and Outcomes..., Bernard [/bib_ref]. In these patients, TP53 mutation is generally biallelic, with the loss of one allele and the mutation of the remaining one. TP53 mutation is however also seen in about 20% of low risk MDS patients with 5q deletion, where it is generally monoallelic. This MDS subset predominates in females, and the median age is lower than that of AML or MDS with a complex karyotype and biallelic TP53 mutation.
As in other cancers, the majority of TP53 mutations in AML and MDS, are missense and localized in the DNA-binding domain [bib_ref] The CBio Cancer Genomics Portal: An Open Platform for Exploring Multidimensional Cancer..., Cerami [/bib_ref]. It is unclear if the deleterious effect is only due to loss of function (especially in case of biallelic mutation) [bib_ref] A Dominant-Negative Effect Drives Selection of TP53 Missense Mutations in Myeloid Malignancies, Boettcher [/bib_ref] , or could include dominant negative effect or gain of function. TP53 mutations usually confer a poor prognosis in MDS and AML. This is especially the case for biallelic mutations, generally corresponding, as mentioned above, to mutations of one allele and loss of the other through 17 deletion, in complex monosomal karyotypes. Monoallelic mutations, generally corresponding to point mutation have less impact on outcome. Nevertheless, in the case of MDS with no excess of blasts, where it occurs in 20-25% of the cases, monoallelic TP53 mutation is associated with a poorer response to treatment and survival [bib_ref] Implications of TP53 Allelic State for Genome Stability, Clinical Presentation and Outcomes..., Bernard [/bib_ref] [bib_ref] TP53 Mutations in Low-Risk Myelodysplastic Syndromes with Del(5q) Predict Disease Progression, Jädersten [/bib_ref].
## Treatment of aml and mds with tp53 mutation by intensive chemotherapy
In AML, eligibility for classical intensive chemotherapy (IC), generally with an anthracycline and cytosine arabinoside (AraC), is mainly based on age and comorbidities. However, it also includes disease characteristics such as cytogenetic complexity and complex karyotypes, especially monosomal karyotypes, which are associated with poorer complete response (CR) rates, and very short responses [bib_ref] Consensus-Based Definition of Unfitness to Intensive and Non-Intensive Chemotherapy in Acute Myeloid..., Ferrara [/bib_ref] [bib_ref] Improved Survival of Men 50 to 75 Years Old with Acute Myeloid..., Juliusson [/bib_ref] [bib_ref] Validation of the "Fitness Criteria" for the Treatment of Older Patients with..., Borlenghi [/bib_ref] [bib_ref] Genetic Characteristics and Outcomes by Mutation Status in a Phase 3 Study..., Lindsley [/bib_ref] [bib_ref] TP53 Gene Mutation Is Frequent in Patients with Acute Myeloid Leukemia and..., Bowen [/bib_ref] [bib_ref] Clinical Implications of Subclonal TP53 Mutations in Acute Myeloid Leukemia, Prochazka [/bib_ref] [bib_ref] TP53 Alterations in Acute Myeloid Leukemia with Complex Karyotype Correlate with Specific..., Rücker [/bib_ref].
In patients with a complex karyotype, those with TP53 mutations appear to respond particularly poorly. Their outcome is not improved, contrary to some other types of AML, by the addition of the CD33 inhibitor gemtuzumab [bib_ref] Effect of Gemtuzumab Ozogamicin on Survival of Adult Patients with De-Novo Acute..., Castaigne [/bib_ref] [bib_ref] Routine Use of Gemtuzumab Ozogamicin in 7+3-Based Inductions for All "Non-Adverse, Hui [/bib_ref] , by the use of encapsulated anthracycline-AraC molecules (CPX 351), the addition of lomustine, or by maintenance treatment with CC486 [bib_ref] Genetic Characteristics and Outcomes by Mutation Status in a Phase 3 Study..., Lindsley [/bib_ref] [bib_ref] TP53 Gene Mutation Is Frequent in Patients with Acute Myeloid Leukemia and..., Bowen [/bib_ref] [bib_ref] Effect of Gemtuzumab Ozogamicin on Survival of Adult Patients with De-Novo Acute..., Castaigne [/bib_ref] [bib_ref] Midostaurin plus Chemotherapy for Acute Myeloid Leukemia with a FLT3 Mutation, Stone [/bib_ref] [bib_ref] Improved Survival by Adding Lomustine to Conventional Chemotherapy for Elderly Patients with..., Pigneux [/bib_ref] [bib_ref] CPX-351 (Cytarabine and Daunorubicin) Liposome for Injection Versus Conventional Cytarabine Plus Daunorubicin..., Lancet [/bib_ref] [bib_ref] Real-Life Experience with CPX-351 and Impact on the Outcome of High-Risk AML..., Chiche [/bib_ref] [bib_ref] Oral Azacitidine Maintenance Therapy for Acute Myeloid Leukemia in First Remission, Wei [/bib_ref] [bib_ref] TP53 Mutations in Older Adults with Acute Myeloid Leukemia, Yanada [/bib_ref]. In TP53-mutated AML patients who achieve CR with IC, allo SCT is therefore recommended whenever possible, but is associated with a high relapse risk in this situation, as shown below [fig_ref] Table 1: Outcome of TP53-mutated myeloid malignancies patients treated with intensive chemotherapy [/fig_ref]. MDS, in general, are not treated with IC, with the exception of some candidates to allo SCT, in order to reduce the blast infiltration before transplant. Response to IC is mainly observed in the absence of a complex karyotype, which most TP53-mutated patients carry, as described above.
## Hypomethylating agents (hma) alone
A HMA alone has been used in the treatment of AML until recently and is now considered unfit for IC (based on age, comorbidities and/or the presence of a complex karyotype), while in MDS it remains the reference treatment in most higher risk cases. In AML, Azacitidine (AZA) yields response rates of 30 to 35% including about 30% of CR/CR with incomplete blood count recovery (CRi) [bib_ref] International Phase 3 Study of Azacitidine vs. Conventional Care Regimens in Older..., Dombret [/bib_ref] , Decitabine (DAC) response rates of about 20% including 13% CR/CRi [bib_ref] Improved Survival by Adding Lomustine to Conventional Chemotherapy for Elderly Patients with..., Pigneux [/bib_ref] , and both drugs yield a median overall survival (OS) of about 10.5 months, which may be better than that obtained with low dose cytarabine, especially in patients with a complex karyotype. In higher risk MDS, median OS with AZA or DAC ranges from 18 to 24 months [bib_ref] Low-Dose Decitabine versus Best Supportive Care in Elderly Patients with Intermediate-or High-Risk..., Lübbert [/bib_ref].
In TP53-mutated AML patients, AZA was shown to yield a 20-30% CR rate and a median OS at 7 months [bib_ref] Cytogenetics and Gene Mutations Influence Survival in Older Patients with Acute Myeloid..., Döhner [/bib_ref] ; and DAC a 20-30% CR rate and a median OS of 6-12 months [bib_ref] TP53 and Decitabine in Acute Myeloid Leukemia and Myelodysplastic Syndromes, Welch [/bib_ref] [bib_ref] Decitabine in TP53-Mutated AML, Welch [/bib_ref] [bib_ref] Treatment with a 5-Day versus a 10-Day Schedule of Decitabine in Older..., Short [/bib_ref].
Using a 10-day regimen of DAC was shown to improve the outcome in one study, although this result was disputed [bib_ref] TP53 and Decitabine in Acute Myeloid Leukemia and Myelodysplastic Syndromes, Welch [/bib_ref].
Furthermore, TP53 mutation has an unfavorable impact on the outcome of higher-risk MDS treated with AZA, with lower response rates but, more importantly, shorter response duration and a median OS of 12.4 months [bib_ref] Prognostic Value of TP53 Gene Mutations in Myelodysplastic Syndromes and Acute Myeloid..., Bally [/bib_ref] [fig_ref] Table 2: Outcome of TP53-mutated myeloid malignancy patients treated with hypomethymating agents [/fig_ref].
## Hma in combination with other drugs in aml and mds with tp53 mutation
The combination of AZA and Venetoclax is superior to AZA alone in AML in the elderly as a whole but also for patients with comorbidities who are not eligible for intensive treatment. It has become the reference treatment of AML in this group, and combinations with other agents have also been tested. However, in higher-risk MDS no combination has so far demonstrated a benefit over AZA or DAC alone in a phase 3 trial, but a recent phase 1/2 trial reported encouraging results on the combination of AZA and venetoclax [fig_ref] Table 3: Outcome of TP53-mutated myeloid malignancy patients treated with a combination of hypomethymating... [/fig_ref].
## Hma + venetoclax
AZA combined with venetoclax (VEN) has become the standard of care for AML patients ineligible for IC, based on its superiority over AZA alone [bib_ref] Azacitidine and Venetoclax in Previously Untreated Acute Myeloid Leukemia, Dinardo [/bib_ref]. In TP53-mutated AML patients, while an increase of ORR was observed with HMA + VEN (55% vs. 0% in patients treated by AZA alone), responses were short with these combinations, and no significant OS benefit was observed (median OS was 6 months independent of treatment arm) [bib_ref] Venetoclax Combined with Decitabine or Azacitidine in Treatment-Naive, Elderly Patients with Acute..., Dinardo [/bib_ref] [bib_ref] Outcomes of TP53-Mutant Acute Myeloid Leukemia with Decitabine and Venetoclax, Kim [/bib_ref] [bib_ref] Venetoclax and Hypomethylating Agents in TP53-Mutated Acute Myeloid Leukaemia, Aldoss [/bib_ref]. In high-risk MDS patients, a recent phase 1/2 trial reported that the combination of AZA + VEN was safe, and demonstrated encouraging results in this group of patients. Therefore, HMA + VEN may be a useful combination to bridge a patient to allo SCT, but is insufficient in itself.
## Azacitidine + eprenetapopt
Eprenetapopt (APR-246) is a small molecule and its mechanism of action has not been completely elucidated. It has been shown to selectively induce apoptosis of TP53mutant cancer cells. After conversion to methylene quinuclidinone (MQ), it covalently binds to mutant p53 to restore wild-type conformation, resulting in cell cycle arrest and apoptosis [bib_ref] PRIMA-1 Induces P53-Mediated Apoptosis by Upregulating Noxa in Esophageal Squamous Cell Carcinoma..., Furukawa [/bib_ref]. Aside from this mechanism, it may induce depletion in glutathione (GSH), increase oxidative stress [bib_ref] PRIMA-1 and PRIMA-1Met (APR-246): From Mutant/Wild Type P53 Reactivation to Unexpected Mechanisms..., Perdrix [/bib_ref] [bib_ref] APR-246/PRIMA-1MET Inhibits Thioredoxin Reductase 1 and Converts the Enzyme to a Dedicated..., Peng [/bib_ref] [bib_ref] Inhibiting the System XC-/Glutathione Axis Selectively Targets Cancers with Mutant-P53 Accumulation, Liu [/bib_ref] [bib_ref] Inhibition of the Glutaredoxin and Thioredoxin Systems and Ribonucleotide Reductase by Mutant..., Haffo [/bib_ref] , and induce ferroptosis. Several preclinical and clinical studies have reported a synergistic effect with AZA [bib_ref] Synergistic Effects of PRIMA-1Met (APR-246) and 5-Azacitidine in TP53-Mutated Myelodysplastic Syndromes and..., Maslah [/bib_ref].
These early observations have led to two clinical trials, conducted in the US and France, to assess the efficacy and safety of a combination of AZA + eprenetapopt in intermediate, high, and very-high IPSS-R risk myeloid malignancies (AML and MDS) (only 20 to 30% of blasts in the US trial) [bib_ref] APR-246) and Azacitidine in TP53-Mutant Myelodysplastic Syndromes, Sallman [/bib_ref] [bib_ref] Eprenetapopt Plus Azacitidine in TP53-Mutated Myelodysplastic Syndromes and Acute Myeloid Leukemia: A..., Cluzeau [/bib_ref].
The US phase 2 trial enrolled fifty-five patients (40 MDS/11 AML) with a median age of 66 years. The Overall Response Rate (ORR) was 71% with a CR rate of 44%, with 38% undetectable TP53 measurable residual disease (Next Generation Sequencing (NGS)) with TP53 NGS negativity. Median CR duration was 7.3 months; with a median follow up of 10.5 months, median OS was 10.8 months.
In the French phase 2 trial, fifty-two patients (34 MDS/18 AML) with a median age of 74 years were enrolled. The ORR was 58%, with a CR rate of 37% and an undetectable TP53 measurable residual disease (MRD) of 30% (NGS). Median CR duration was 11.7 months; with a median follow up of 9.7 months, median OS was 12.1 months.
Regarding safety, no additional hematological toxicity was reported compared to azacitidine alone, however, neurological side effects including ataxia, cognitive impairment, acute confusion, isolated dizziness, and facial paresthesia were reported in 40% of patients (6% were grade 3 or 4, and all were fully reversible without recurrence after dose reduction).
Patients with TP53-mutated MDS fared better than AML patients (for AML patients, a phase 1 study evaluating a combination of AZA + VEN + eprenetapopt in TP53-mutated AML is ongoing (clinicaltrials.gov NCT: NCT04214860)).
Based on the results of these phase 2 trials, a phase 3 randomized clinical trial was conducted to compare AZA + eprenetapopt to AZA monotherapy in TP53 MDS patients (clinicaltrials.gov NCT: NCT03745716). The results did not demonstrate superiority for the combination when compared to azacitidine alone. Despite these results, eprenetapopt is now being investigated in the transplant setting, especially to decrease the tumor burden before and after transplant by providing a significant reduction in MRD.
## Azacitidine + magrolimab
Magrolimab (MAGRO) is a first-in-class investigational monoclonal antibody against CD47 and a macrophage checkpoint inhibitor that is designed to interfere with the recognition of CD47 by the SIRPα receptor on macrophages, thus blocking the "don't eat me" signal used by cancer cells to avoid being ingested by macrophages. Overexpression of CD47 is an adverse prognostic factor and could be a therapeutic antibody target on human acute myeloid leukemia stem cells. CD47 blockade induces tumor phagocytosis and eliminates leukemia stem cells in AML models [bib_ref] CD47 Is an Adverse Prognostic Factor and Therapeutic Antibody Target on Human..., Majeti [/bib_ref]. AZA has been shown to both increase expression of CD47 and the pro-phagocytic signal calreticulin in myeloid malignancies. AZA synergizes with MAGRO by inducing "eat me" signals on AML, to enhance phagocytosis [bib_ref] An Insight into Their Roles in the Disease Progression of MDS and..., Boasman [/bib_ref].
In a phase 1b trial of AZA + MAGRO in AML patients, where the MAGRO schedule was 30 mg/kg IV weekly or Q2W associated to AZA 75 mg/m 2 /d days 1-7 on 28-day cycles, results for 64 patients were reported. The ORR was 63% including 42% CR and 12%
CRi. Forty-five percent of patients obtained a complete cytogenetic response (CCR) and 35% obtained MRD negativity. The median duration of response was 9.6 months.
The study population was enriched with TP53-mutated patients (29/43 (67%)), whose ORR was 69% including 45% CR and 14% CRi, with a median duration of response of 7.6 months, 44% of CCR, and 29% MRD negativity, and a median OS of 12.9 months (vs. 18.9 months in TP53 wild-type patients) [bib_ref] The First-in-Class Anti-CD47 Antibody Magrolimab Combined with Azacitidine Is Well-Tolerated and Effective..., Sallman [/bib_ref].
A phase 3 clinical trial evaluating AZA + VEN vs. AZA + MAGRO in TP53-mutated AML (clinicaltrials.gov NCT04778397) and a phase 1/2 study evaluating AZA + VEN + MAGRO (clinicaltrials.gov NCT04435691) are currently enrolling.
## Aza+ sabatolimab
Sabatolimab is a T-cell immunoglobulin and mucin domain-containing protein 3 (TIM3) inhibitor. TIM3 is a member of the TIM family, that was originally identified as a receptor expressed on T cells. Recent data have demonstrated that TIM3 works as a co-inhibitory receptor (checkpoint receptor), exhibited by dysfunctional or 'exhausted' T cells. Sabatolimab was tested in combination with HMAs in AML and higher-risk MDS. Preliminary results of AZA + sabatolimab trial, presented at the EHA (European Hematology Association) 2021 meeting suggested promising results in a limited number of TP53-mutated patients, with an ORR of 71.4% including 28.6% CR, and 14.3% marrow CR.
## New combinations needing investigation
Due to the high frequency of TP53 mutations in cancer, and their resulting poor prognosis, many drugs are being tested in vitro for their ability to inhibit mutated p53, reconform it, and/or activate some of its targets.
Drugs already used in other diseases may be particularly interesting to develop in TP53-mutated AML and MDS.
As an example, Chen et al. recently showed that Arsenic Trioxyde (ATO) rescues structural p53 mutations through a cryptic allosteric site [bib_ref] Arsenic Trioxide Rescues Structural P53 Mutations through a Cryptic Allosteric Site, Chen [/bib_ref].
Sixty percent of TP53 mutations are conformational in MDS and AML and could be restored by ATO. The Groupe Francais des Myelodysplasies (GFM) obtained a 19% response rate with ATO in MDS. Roboz et al. [bib_ref] Arsenic trioxide and low-dose cytarabine in older patients with untreated acute myeloid..., Roboz [/bib_ref] reported 34% CR in AML with the combination of low dose aracytine and ATO (including 30% in secondary or therapy related AML and 30% in patients with unfavorable cytogenetics). However, no analysis of TP53 mutation was carried out in either study. Combinations of AZA + ATO (+/− VEN) warrant evaluation in AML and MDS with TP53 mutation.
Other drugs aiming specifically at mutated TP53 are being developed. ReACp53, a cell-penetrating peptide, designed to inhibit p53 amyloid formation, has shown rescue p53 function in cancer cell lines and in organoids, and is in evaluation in some solid TP53-mutated cancers [bib_ref] A Designed Inhibitor of P53 Aggregation Rescues P53 Tumor Suppression in Ovarian..., Soragni [/bib_ref]. COTI-2, a novel thiosemicarbazone derivative, normalized wild-type p53 target gene expression and restored DNA-binding properties to the p53mutant protein [bib_ref] COTI-2, A Novel Thiosemicarbazone Derivative, Exhibits Antitumor Activity in HNSCC through P53-Dependent..., Lindemann [/bib_ref]. Niclosamide, an anthelmintic drug, inhibited ASAP2, a member of the ArfGAP family, which is overexpressed in pancreatic ductal adenocarcinoma characterized by four main driver genes KRAS, TP53, CDKN2A, and SMAD4. Niclosamide was able to bypass the effect of TP53 mutations in other models and need to be explored [bib_ref] The Novel Driver Gene ASAP2 Is a Potential Druggable Target in Pancreatic..., Fujii [/bib_ref]. Other drugs in development for solid cancers could be also interesting such as Ataxia Telangiectasia Mutated (ATM) inhibitors, ataxia telangiectasia mutated, and Rad3-related (ATR) inhibitors. The Rad3 gene is required for cell viability and excision repair of damaged DNA, and is also called the FRAP-related protein 1 (FRA 1). ATR is encoded in humans by the ATR gene. Furthermore, Wee1 inhibitors and Checkpoint kinase (CHK) inhibitors may be potential candidates [bib_ref] Phase II Study of WEE1 Inhibitor AZD1775 Plus Carboplatin in Patients with..., Leijen [/bib_ref] [bib_ref] Chk1/2 Inhibition Overcomes the Cisplatin Resistance of Head and Neck Cancer Cells..., Gadhikar [/bib_ref]. Wee 1 kinase is gatekeeper of the G2-M cell cycle checkpoint that allows DNA repair before mitotic entry. Targeting Wee 1 for inhibition potentiates chemotherapy because Wee 1 is highly expressed and active in several cancer types that are dependent on a functional G2-M checkpoint for DNA repair. More recently, investigational bispecific dual-affinity retargeting antibody flotetuzumab was assessed in patients with relapsed refractory AML. Seven out of 15 patients achieved complete remission and a median overall survival of 10.3 months [bib_ref] TP53 Abnormalities Correlate with Immune Infiltration and Associate with Response to Flotetuzumab..., Vadakekolathu [/bib_ref]. These results emphasized the central role of immunotherapy in this highly chemoresistant AML subtype.
## Allogeneic stem cell transplantation
Allogeneic stem cell transplantation (aHSCT) remains the only curative option for patients with TP53-mutated AML or MDS. However, the presence of the TP53 mutation in AML or MDS is associated with a high risk of relapse pretransplant, and is the main cause of death following transplantation [fig_ref] Table 4: Outcome of TP53-mutated myeloid malignancy patients after allogeneic stem cell transplantation [/fig_ref]. A retrospective analysis of 30 patients who underwent aHSCT for AML (n = 19) or MDS (n = 11) with chromosome 17 abnormalities, reported a poor outcome for patients with the TP53 mutation [bib_ref] Acute Myeloid Leukemia or Myelodysplastic Syndrome with Chromosome 17 Abnormalities and Long-Term..., Britt [/bib_ref]. Patients experienced short relapse-free survival, with a median relapse-free survival of 6-and 4-months posttransplant for the AML and MDS patients, respectively, (p > 0.5). The non-relapse mortality (NRM) was 10.6% (2/19 patients) and 9.1% (1/11 patients) for AML and MDS patients, respectively, (p > 0.5). The median overall survival post-transplant was 18 months for AML and 11 months for MDS patients respectively (p > 0.5). In a recent retrospective report from the European Society for Blood and Marrow Transplantation (EBMT), the relapse rate for patients with a 17p abnormality was as high as 61% at 2 years [bib_ref] Allogeneic Stem Cell Transplantation in Adult Patients with Acute Myeloid Leukaemia and..., Poiré [/bib_ref]. The subsequent 2 year OS and leukemia free survival (LFS) were 28 and 24%, respectively. The study enrolled 125 AML patients carrying a 17p abnormality who received an aHSCT in CR1. In multivariate analysis only, the presence of a −5/5qin addition to abn (17p) was significantly and independently associated with worse OS, LFS, and higher relapse incidence. The type of conditioning was not significantly associated with the outcome. An American retrospective study by the Center for International Blood and Marrow Transplant Research (CIBMTR) confirmed these results in a cohort of 1514 patients transplanted for MDS [bib_ref] Prognostic Mutations in Myelodysplastic Syndrome after Stem-Cell Transplantation, Lindsley [/bib_ref]. In this cohort, 289 patients had TP53 mutations. In a multivariate analysis, TP53 mutations were associated with shorter survival and a shorter time to relapse. The 3 years OS was 20% and the median OS was 0.7 years. In this study, a TP53 variant allele fraction (VAF) of 10% or higher was not significantly associated with survival and neither was the presence of multiple TP53 mutations. Interestingly, another study found the mutation type had an impact on transplant outcome. Indeed, patients with only truncating mutations had shorter survival than patients exhibiting missense mutations. Another study identified the TP53 allelic state as an independent prognostic factor for TP53-mutated AML patients following aHSCT [bib_ref] TP53 Mutation Status Divides Myelodysplastic Syndromes with Complex Karyotypes into Distinct Prognostic..., Haase [/bib_ref]. In this retrospective cohort analysis of 36 patients with TP53, the authors showed a trend for longer overall survival in monoallelic patients as compared to multi-hit patients following HSCT. These results emphasized the importance of analyzing the TP53 state in future transplantation studies. In an attempt to identify different prognostic factors in TP53-mutated AML transplanted patients, biological samples of a cohort of 83 consecutive patients transplanted for TP53 AML or MDS at a single center were analyzed between February 2011 and March 2017 [bib_ref] Prognostic Factors Influencing Survival after Allogeneic Transplantation for AML/MDS Patients with TP53..., Ciurea [/bib_ref]. The median age was 60 years. In this study, 71% of the patients received a myeloablative conditioning, and 59% of the grafts were peripheral blood stem cells. Of note, 16% of the patients received subcutaneous azacitidine maintenance following HSCT. A multivariate analysis showed the melphalan-based regimen (hazard ratio [HR], 6.5; 95% CI, 2.1-20; p = 0.001), Karnofski Performans Status (KPS) ≤80% (HR, 2.8; 95% CI, 0.96-8; p = 0.06), and treatment-related AML/MDS (HR, 4.2; 95% CI, 1.5-12; p = 0.007). Regarding overall survival predictive factors, HCT-CI > 4 (HR, 3.9; 95% CI, 1.2-13; p = 0.03), KPS ≤80% (HR, 3.04; 95% CI, 1.6-5.9; p = 0.001), and disease not in CR1/2 (HR, 4.1; 95% CI, 1.5-11; p = 0.004) were associated with worse OS. In this study, azacitidine maintenance did not affect the patient's outcome. The study also reports better progression-free survival for patients receiving higher doses of busulfan, emphasizing the need to reduce the leukemic-cell burden before transplant. It has been previously shown that MRD negativity at transplant correlates with a better post-transplant outcome [bib_ref] Pre-and Post-Transplant Quantification of Measurable ('Minimal') Residual Disease via Multiparameter Flow Cytometry..., Zhou [/bib_ref]. A retrospective analysis from the Moffitt Cancer Center on 47 patients who received an aHSCT for TP53 mut MDS/AML, reported that patients who achieved a clearance of the TP53 mutation (NGS) at the time of aHSCT after receiving HMA treatment, had a significantly better OS than those who did not achieve clearance (median OS of 21.73 months vs. 6.44 months, p = 0.042) [bib_ref] Impact of TP53 Gene Mutation Clearance and Conditioning Intensity on Outcome in..., Chan [/bib_ref]. In the pre-transplant setting, the addition of eprenetapopt to demethylating agents offers more efficient treatment options to these usually chemoresistant AML subsets. As presented above, in a phase Ib/II trial, eprenetapopt combined with azacitidine yielded complete remission, including complete molecular remission [bib_ref] Prognostic Value of TP53 Gene Mutations in Myelodysplastic Syndromes and Acute Myeloid..., Bally [/bib_ref]. In this study 40% of the patients proceeded to aHSCT. The median time to aHSCT was 5.6 months (range, 1.7-9.7 months). The good complete response rate did not translate into an improved survival rate after aHSCT, as the median OS for patients who were bridged to aHSCT was 14.7 months (95% CI, 8.6 to 20.9). Moreover, in this study, aHSCT was not significantly associated with survival (hazard ratio, 1.01; p = 0.98). However, a subgroups analysis revealed several factors that may influence the outcome of TP53 AML patients treated with a combination of azacitidine and eprenetapopt. The number of cycles patients received before transplant impacted aHSCT survival. Indeed, OS was significantly better in the group of patients receiving at least four cycles of combination therapy prior to aHSCT compared with those who received less (16.1 months; 95% CI, 10.4 to not reported [NR] v 9.3 months; 95% CI, 8 to NR months, respectively; p = 0.01).
In the post-transplantation setting, the possibility of post-transplantation maintenance with eprenetapopt in combination with azacytidine for TP53 AML patients has been assessed in a phase 2 clinical trial (NCT03931291). In a cohort of 33 patients who received aHSCT for AML/MDS with TP53 mutations, 1-year relapse-free survival was 58% with an OS at 76%. Median RFS and OS were 12.1 and 19.3 months, respectively. These truly encouraging results confirm the necessity of post transplantation intervention for TP53-mutated AML and MDS.
Another factor accounting for the high relapse incidence following aHSCT is the highly immunosuppressive microenvironment surrounding the TP53-mutant leukemic cell. Among the immune escape mechanisms is the expression of inhibitory PDL1 [bib_ref] Bone Marrow Central Memory and Memory Stem T-Cell Exhaustion in AML Patients..., Noviello [/bib_ref]. A recent report demonstrated a significant increase of PDL1 expression at the surface of hematopoietic stem cells of patients with TP53 mutations. TP53-mutated AML patients also exhibited reduced numbers of bone marrow-infiltrating OX40 + cytotoxic T cells and helper T cells. Further adding to the immunosuppressive characteristics of the microenvironment was the increased frequency of highly suppressive Tregs. A donor lymphocyte infusion is one of the available options to bypass T-cell exhaustion [bib_ref] Donor Lymphocyte Infusion for Prevention of Relapse after Unmanipulated Haploidentical PBSCT for..., Gao [/bib_ref].
A better understanding of the mechanisms underlying the immune escape of TP53mutated AML is essential to improve the outcome for these patients following aHSCT. A greater response before transplantation, combined with increased alloreactivity post transplantation using donor lymphocyte infusions with or without chemical maintenance with eprenetapopt, sabatolimab, azacytidine, or venetoclax are some of the options currently under investigation.
# Conclusions
TP53-mutated AML and MDS are some of the myeloid malignancies with the poorest prognosis. The prognoses are heterogeneous, depending on mono or biallelic mutations. The emergence of new drugs with several mechanisms of action are encouraging. These new drugs induce a greater response and may cure some patients using aHSCT. Several sequential strategies need to be evaluated to identify the best therapeutic strategy to cure these patients. Conflicts of Interest: M.L.: has received honoraria from Gilead, Novartis, and Celgene/BMS for projects out of the scope of this article. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results. P.F. and T.C.: have received honoraria from Aprea, Celgene/BMS, Novartis, and Pfizer for projects out of the scope of this article. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.
[fig] Author: Contributions: Writing-original draft preparation, M.L., P.F. and T.C. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. [/fig]
[table] Table 1: Outcome of TP53-mutated myeloid malignancies patients treated with intensive chemotherapy. CR: Complete Remission, EFS: Event Free Survival, OS: Overall Survival. [/table]
[table] Table 2: Outcome of TP53-mutated myeloid malignancy patients treated with hypomethymating agents. CR: Complete Remission, EFS: Event Free Survival, OS: Overall Survival. [/table]
[table] Table 3: Outcome of TP53-mutated myeloid malignancy patients treated with a combination of hypomethymating agents. CR: Complete Remission, CRi: Complete Remission with Incomplete count recovery, EFS: Event Free Survival, OS: Overall Survival. [/table]
[table] Table 4: Outcome of TP53-mutated myeloid malignancy patients after allogeneic stem cell transplantation. AML: Acute Myeloid Leukemia, CR: Complete Remission, EFS: Event Free Survival, MAC: Myeloablative Conditioning, MDS, Myelodysplastic Syndrome, OS: Overall Survival, PFS: Progression Free Survival, RIC: Reduced Intensity Conditioning. [/table]
|
Outcomes of 23-gauge pars plana vitrectomy in vitreoretinal diseases
Purpose: The aim of this study was to assess the efficiency and reliability of the 23-gauge (23G) transconjunctival vitrectomy system and examine possible complications of this surgical technique in a variety of vitreoretinal conditions along with early postoperative intraocular pressure (IOP) changes. Materials and methods: A total of 350 eyes of 324 patients having undergone 23G transconjunctival vitrectomy were included in this prospective study. A total of 150 (46.2%) were male and 174 (53.8%) female, with a mean age of 61.28 ± 15.67 years. Mean follow-up time was 8.3 months. Results: Mean BCVA logMARs were as follows: preoperatively 0.839 ± 0.59, postoperatively first day 2.07 ± 0.76, first week 1.14 ± 0,43, first month 0.63 ± 0.26 and last examination 0.359 ± 0.17. Mean BCVA decreased significantly (P , 0.001, P , 0.028, respectively) on postoperative first day and first week, mainly due to air or gas tamponade, and increased significantly in the first month and final control in almost all indications (P , 0.001). Postoperative mild hypotony (IOP # 10 mmHg) was detected in 112 (32%) eyes on day 1 and in 59 (16.8%) eyes in week 1. While postoperative serious hypotony (#5 mmHg) was detected in 34 (9.7%) eyes on day 1, it was not detected in any eyes at the end of the first week. None of the eyes required an additional gas tamponade or any other procedure in the early postoperative period due to hypotony. A total of 13 (3.7%) eyes were reoperated for recurrent vitreous hemorrhage; 23 (6.5%) eyes were reoperated on a second time, nine (2.5%) a third time, and 1 (0.2%) a fourth time for recurrent rhegmatogenous retinal detachment. Postoperative fibrinoid reaction was seen in 17 (4.8%) eyes on the first day and responded well to the medications. Cataract development was found in 61 (22.5%) of the 270 phakic eyes after a mean duration of 6.4 ± 3.5 months. Anatomical success was obtained in 86% of the patients and functional success in 72%. Conclusion: The 23G transconjunctival vitrectomy system is safe and effective in a wide field of vitreoretinal conditions. It is a good alternative to 20G and 25G techniques but needs some improvement mainly in regards to the instruments and related techniques; further larger controlled group studies are needed.
# Introduction
Technological advancements in vitreoretinal surgery have made it possible to surgically treat certain diseases which had been considered hard for a long time. Invented by Machemer in the early 1970s, the "pars plana vitrectomy" technique has led to an increasingly more advanced and promising treatment, which calls for limited intervention. Introduced for the first time by Claus Eckardt, the 23-gauge (23G) transconjunctival vitrectomy technique is commonly accepted by vitreoretinal surgeons in daily practice. In this study, the outcomes of 350 cases of 23G pars plana vitrectomies with different etiologies performed at the İstanbul Training and Research Hospital, Ophthalmology Clinic were assessed. The aim was to assess the efficiency and reliability of the 23G transconjunctival vitrectomy system and examine possible complications of this surgical technique along with early postoperative intraocular pressure (IOP) changes.
# Materials and methods
Between November 2005 and December 2007, 23G transconjunctival vitrectomy was performed in the clinic on 350 eyes of 324 patients. In this prospective study, all 350 eyes of these 150 male (46.2%) and 174 female (53.8%) patients were examined. Mean follow-up time was 8.3 months (4-48 months). Mean age was 61.28 ± 15.67 years. Informed consent according to the Declaration of Helsinki were taken prior to surgery. The vitrectomy indications in the study are indicated in [fig_ref] Table 1 23: Gauge vitrectomy indications [/fig_ref].
Best-corrected visual acuity (BCVA) was measured using the ETDRS (Early Treatment Diabetic Retinopathy Study) and Snellen charts. For ease of analysis, visual acuity was turned into logMAR (logarithm of the minimum angle of resolution) scores. IOP measurements were made with Goldmann applanation tonometry. IOP # 10 mmHg was considered as mild hypotony and #5 mmHg as serious hypotony. Biomicroscopic findings regarding the cornea, conjunctiva, and anterior chamber were noted. The funduscopic examination was carried out by contact lenses and binocular indirect ophthalmoscope. In cases where fundus imaging was not possible, mode A and B ultrasonographic examination was undertaken. In this series, B-mode ultrasonographic examination was undertaken on all eyes with vitreous hemorrhage (VH) in the preoperative period. All patients with vitreomacular surface disorders were examined with the Stratus OCT™ III (Carl Zeiss Meditec Inc, Dublin, CA) device. Intraoperative or postoperative complications such as leakage from sclerotomy, hemorrhage, vitreous or retinal incarceration, retinal detachment at transconjunctival entry, or dialysis were noted. Goldmann three-mirror and extreme wide field contact lenses were used to evaluate the vitreous base.
Retinal detachment, vitreous or suprachoroidal hemorrhage, choroidal detachment or folds, and endophthalmitis incidence was assessed. Anti-vascular endothelial growth factor was injected into some diabetic VH and diabetic tractional retinal detachment (DTRD) cases 2-7 days before surgery, and in one case at the beginning only of the operation.
The operations were performed under local or general anesthesia. Sclerotomies were performed from upper temporal, lower temporal, and upper nasal quadrants, 3.0, 3.5, and 4.0 mm from the corneoscleral limbus in aphakic, pseudophakic, and phakic cases, respectively, by using a 23G microvitreoretinal knife parallel to the limbus and at an angle. In order to prevent postoperative leakage, the scleral tunnel that formed in the sclerotomy with this technique was moved slightly to the conjunctiva with the help of a special fixation technique known as conjunctival fixation [fig_ref] Figure 1: Trocar and infusion canula positions in 23 gauge vitrectomy [/fig_ref]. The first microcannula was inserted in the lower temporal sclerotomy, followed by infusion. Balanced saline solution was used as the infusion fluid [fig_ref] Figure 2: Specially designed pressure plate for holding the conjunctiva [/fig_ref]. Air, expansile gas, or silicone oil (1000 centistokes, 5000 centistokes) was used as endotamponade when necessary; hexafluoride (SF6) was used in 60 (17.1%) eyes, perfluoropropane (C3F8) in 105 (30%) eyes, silicone oil in 65 (18.7%) eyes, and air in 74 (21.1%) eyes; 46 (13.1%) eyes were left with fluid. Silicone oil was injected after the fluid-air or fluid PFC -silicone exchange in some of the cases, without enlarging the sclerotomies.
Peripheral iridectomy at the 6 o'clock position was performed in aphakic eyes before silicone oil tamponade. Any sclerotomy leaking air, gas, liquid, or especially silicone oil was sutured with 7/0 vicryl at the end of surgery. Subconjunctival antibiotics and steroid injections were administered at the end of surgery. Statistical analyses were done by using SPSS Statistics (IBM Corporation, Somers, NY) software, version 11.5 for Windows. The difference between pre-and postoperative visual acuity and IOP levels were compared by using two-tailed paired t-test. Results with a P-value below 0.05 were considered statistically significant.
# Results
Mean BCVA logMARs were as follows: preoperatively 0.839 ± 0.59, postoperatively first day 2.07 ± 0.76, first week 1.14 ± 0,43, first month 0.63 ± 0.26 and last examination 0.359 ± 0.17. Mean BCVA decreased significantly (P , 0.001, P , 0.028, respectively) on postoperative first day and first week, mainly due to air or gas tamponade, and increased significantly in the first month and final control (P , 0.001).
Preoperative BCVA logMARs were 0.658 (3.10-0) in RRD, [bib_ref] Quantitative assessment of postsurgical breakdown of the blood-aqueous barrier, Sanders [/bib_ref] [fig_ref] Table 2: Mean preoperative and postoperative visual acuities [/fig_ref].
Mean intraocular pressure values for all patients prior to the operation was 14.75 ± 7.1, followed by the postoperative day 1 value of 11.62 ± 9.7 (P , 0.001), postoperative week 1 value of 15.86 ± 6.6 (P = 0.330), and postoperative month 1 value of 16.84 ± 5.6 (P = 0.556). Last postoperative mean intraocular pressure was 15.1 ± 3.9 (P = 0.306).
Postoperative mild hypotony (IOP # 10 mmHg) was detected in 112 (32%) eyes on day 1 and in 59 (16.8%) eyes in week 1. While postoperative serious hypotony (#5 mmHg) was detected in 34 (9.7%) eyes on day 1, it was not detected in any eyes at the end of the first week. None of the eyes required an additional gas tamponade or any other procedure in the early postoperative period due to hypotony. During the follow-up, no choroidal effusion, detachment, or folds developed, and no vitreous incarceration around the wound was noted. Single sclerotomies of four patients who had nucleus fragments in the vitreous cavity were changed to 20G for phacofragmentation, as a 23G phacofragmentation probe was not available at that time. In all macular hole (MH) cases (stages 2-4), internal limiting membrane peeling and fluid/air/gas exchange were performed. Postoperative optical coherence tomography (OCT) revealed opening of the MH in five (18.1%) eyes after a mean duration of 3.73 ± 2.92 months. Holes of four were closed after the second 23G pars plana vitrectomy. All grades of retinal detachments were submitted to 23G pars plana vitrectomy. Advanced proliferative vitreoretinopathy cases had combined lens surgery, radial retinotomies of up to 360°, retinectomies, and/or encircling scleral buckles. In some DTRD cases, bimanual surgery was performed. Among the rhegmatogenous retinal detachment (RRD) patients, 68 were given C3F8 gas and 44 were given silicone oil tamponade in the primary surgery. Of the 98 cases, 23 (6.5%) eyes were reoperated on a second time, nine (2.5%) a third time, and one (0.2%) a fourth time for recurrent RRD. Encircling scleral buckles were used in 13 (3.5%) of those cases. In some of the 11 cases (3.1%) who had silicone leakage either outside and/or under the conjunctiva after the removal of trocars, silicone oil was found inadequate in the vitreous cavity in the postoperative period. It was noted that recurrence was mostly observed in these patients. Anatomical and functional success rates were obtained from the patients' last visit findings. Anatomical success was defined as: reattachment of retinal detachments until the final visit, relieving of tractional detachments of DTRDs, cleaning of VHs, vitreous opacities, and nuclear fragments; closure of MHs; absence of epiretinal membranes (ERMs) and vitreomacular traction syndrome (VMT) confirmed by OCT. On the other hand, functional success was defined as BCVA of Snellen 1/20 (logMAR 1.3) or better on last follow-up and no reduction in visual acuity as compared with the preoperative period.
Functional success, on the other hand, was defined as BCVA of 1/20 or higher on last follow-up and no reduction in vision as compared with the preoperative period. Evaluated according to these success criteria, anatomical success was obtained in 86% of the patients and functional success in 72%. Revitrectomy was performed in 13 (3.7%) diabetic patients for intravitreous bleed. It was thought that the major reasons were because of mild or serious hypotony and inadequate peripheral vitreous dissection and inadequate peripheral laser photocoagulation, especially in the phakic eyes.
Of the 350 patients, 270 were phakic. On postoperative follow-up, cataract was found in 61 (22.5%) of the 270 phakic eyes after a mean duration of 6.4 ± 3.5 months. Of these patients, 52 underwent phacoemulsification and intraocular lens implantation, and nine were left aphakic. Sclerotomy had to be shifted to 20G for comfortable tool manipulation in four eyes that had lens fragmentation in the vitreous. Fibrin reaction was observed in the anterior chamber of 17 (4.8%) eyes on the first postoperative day. These all had resolution by pupillary dilatation, steroid treatment, and fortified antibiotic eyedrops during the follow-up.
# Discussion
The 20G vitrectomy technique has been a widely accepted technique by surgeons for more than 20 years. It is quite effective in cleaning thick hemorrhages, and fibrous bands and has large clinical application potential. However, transconjunctival techniques are starting to be preferred due to shorter operation time, less need of suturing, less postoperative inflammation, and less pain.
Developed recently by Fujii et al, [bib_ref] A new 25-gauge instrument system for transconjunctival sutureless vitrectomy surgery, Fujii [/bib_ref] the transconjunctival unsutured vitrectomy technique is one of the most important advances of vitreoretinal surgery. Owing to the thinness of 25G instruments, sclera recovery can be achieved without needing sutures. Therefore, suture-related astigmatism and discomfort can be eliminated. As is widely known, sutures used without covering the sclerotomy area cause irritation and pigmentation in the sclerotomy area. Thirty-two percent local inflammatory reaction is reported with Dacron and 5% with polyglycolic acid. With unsutured surgery, suture-related irritation and local inflammatory reaction can be prevented. Various previous studies have stated that unsutured smallincision cataract surgery reduces postoperative inflammatory reaction. 1-3 Karaçorlu et al compared 14 unsutured eyes and 20 sutured eyes after applying 20G pars plana vitrectomy. While the eight patients who were sutured displayed ocular surface irritation and one eye displayed scleral pigmentation, these complications were not seen in any of the unsutured eyes. [bib_ref] Sutureless pars plana vitrectomy, Karaçorlu [/bib_ref] It is believed that, when compared with conventional pars plana vitrectomy, 25G unsutured vitrectomy causes less postoperative inflammation and faster postoperative recovery. 5, [bib_ref] Initial experience using the transconjunctival sutureless vitrectomy system for vitreoretinal surgery, Fujii [/bib_ref] Rizzo et al compared 25G unsutured vitrectomy and 20G pars plana vitrectomy in ERM cases and concluded that 25G unsutured vitrectomy caused less postoperative inflammatory reaction and thus allowed faster rehabilitation. [bib_ref] 25-gauge, sutureless vitrectomy and standard 20-gauge pars plana vitrectomy in idiopathic epiretinal..., Rizzo [/bib_ref] In addition, discomfort levels in 25G unsutured vitrectomy patients were also lower. Romero et al compared 25G unsutured vitrectomy and standard 20G vitrectomy with respect to inflammation and surgery durations and found that 25G unsutured vitrectomy took less time and led to less inflammation after the surgery. [bib_ref] Experience with 25-gauge transconjunctival vitrectomy compared to a 20-gauge system. Analysis of..., Romero [/bib_ref] Likewise, Chen showed that 25G unsutured vitrectomy could be performed in a shorter time and led to less inflammation in the postoperative stage. [bib_ref] 25-Gauge transconjunctival sutureless vitrectomy, Chen [/bib_ref] Vitreous base cleansing takes less time when compared with 25G, and silicone oil injection is faster. In their 13 disease 25G silicone intake series, Kapran and Acar found the mean silicone oil injection time to be 7.27 ± 0.48 minutes, did not Dovepress Dovepress 1775 23-gauge pars plana vitrectomy in vitreoretinal diseases experience any complications during surgery, and did not need additional sutures. [bib_ref] Removal of silicone oil with 25-gauge transconjunctival sutureless vitrectomy system, Kapran [/bib_ref] Although 25G has thinner calibration, 23G causes less postoperative hypotony, as oblique entry to the sclera is used. The 23G vitrectomy technique also has some disadvantages, with postoperative temporary hypotony being the most serious one. However, its rate is lower than for 25G. While Kusuhara et al found a postoperative hypotony rate of 1.5% following 23G vitrectomy in 314 eyes, they found a rate of 18.4% after 25G vitrectomy. [bib_ref] Outcomes of 23-and 25-gauge transconjunctival sutureless vitrectomies for idiopathic macular holes, Kusuhara [/bib_ref] Also, shifting to 20G may be necessary in lens dislocations in the vitreous or in certain complicated cases such as intense fibrovascular membrane. In the present study, the transconjunctival vitrectomy outcomes were examined in 350 eyes of 324 patients with intraventricular hemorrhage, MH,VMT, ERM, DTRD, RRD, dislocation of lens fragments in the vitreous, and intravitreal opacity. Lakhanpal et al recorded the total surgery durations of their series of 140 eyes on which they performed 25G vitrectomy, along with entry, exit, and vitrectomy times. [bib_ref] Outcomes of 140 consecutive cases of 25-gauge transconjunctival surgery for posterior segment..., Lakhanpal [/bib_ref] They reported shorter total surgery times in 25G vitrectomy as it does not involve the stages of conjunctival peritomy and opening and suturing of the sclera. In their 23G vitrectomy series, Fine et al found a mean opening time of 1.7 minutes, mean closing time of 1.3 minutes, and mean procedural time of 24.1 minutes. [bib_ref] Outcomes of 77 consecutive cases of 23-gauge transconjunctival sutureless vitrectomy for posterior..., Fine [/bib_ref] Lakhanpal et al attributed longer surgery times to the fact that the 25G technique involves two stages. [bib_ref] Outcomes of 140 consecutive cases of 25-gauge transconjunctival surgery for posterior segment..., Lakhanpal [/bib_ref] Previous studies have reported shorter mean total surgery times with unsutured vitrectomy than in eyes on which 20G was performed. [bib_ref] A new 25-gauge instrument system for transconjunctival sutureless vitrectomy surgery, Fujii [/bib_ref] [bib_ref] Outcomes of 77 consecutive cases of 23-gauge transconjunctival sutureless vitrectomy for posterior..., Fine [/bib_ref] In the present study, a significant increase in postoperative visual acuity was identified in all patients in general. Considering patient groups, RRD, intraventricular hemorrhage, DTRP, ERM, and MH had a statistically significant increase in visual acuity. Patient groups who were admitted to the study due to endophthalmitis, lens dislocation in the vitreous, or vitreomacular traction were not included in the statistics.
Among patients with gas and air tamponade, mean postoperative visual acuity was lower on day 1 and in week 1, and BCVA was assessed after tamponade resorption. On postoperative follow-up, together with the resorption of internal gas endotamponade (2-8 weeks), mean visual acuity at month 1 was elevated as compared with the preoperative value, and the elevation continued on last follow-up. Other studies about 23G vitrectomy also reveal a statistically significant increase in postoperative follow-up visual acuity values when compared with those in the preoperative stage. [bib_ref] 25-gauge, sutureless vitrectomy and standard 20-gauge pars plana vitrectomy in idiopathic epiretinal..., Rizzo [/bib_ref] [bib_ref] Outcomes of 77 consecutive cases of 23-gauge transconjunctival sutureless vitrectomy for posterior..., Fine [/bib_ref] In the present series, low IOP did not cause choroidal detachment/folds or endophthalmitis. Ibarra et al [bib_ref] Longer-term outcomes of transconjunctival sutureless 25-gauge vitrectomy, Ibarra [/bib_ref] and Yanyali et al [bib_ref] 25-Gauge transconjunctival sutureless pars plana vitrectomy, Yanyali [/bib_ref] conducted studies using 25G, and both reported a significant decrease in IOP on postoperative day 1, similar to the findings of the present study. Fine et al 13 detected severe hypotony on postoperative day 1, one of their patients required suturing on the same day and choroidal effusion did not occur in any of their patients. After they started to use 23G vitrectomy, Eckardt stated that the angle of the incision is more important than its size in preventing postoperative hypotony. [bib_ref] Transconjunctival 23-gauge sutureless vitrectomy, Eckardt [/bib_ref] In general, the rate of postoperative hypotony is higher in 25G series than in 23G series. Shimada et al found hypotony in 20% of the eyes in which they used fluid as endotamponade. [bib_ref] Expanded indications for 25-gauge transconjunctival vitrectomy, Shimada [/bib_ref] Fujii et al reported an IOP value below 10 mmHg in 24% of 64 unsutured vitrectomy patients. [bib_ref] A new 25-gauge instrument system for transconjunctival sutureless vitrectomy surgery, Fujii [/bib_ref] [bib_ref] Initial experience using the transconjunctival sutureless vitrectomy system for vitreoretinal surgery, Fujii [/bib_ref] As a result of 63 operations, Shimada et al found postoperative low IOP in 9% of the eyes that they left with air or gas endotamponade and in 20% of the eyes that they left with fluid. [bib_ref] Expanded indications for 25-gauge transconjunctival vitrectomy, Shimada [/bib_ref] In the present study, SF6 was used as endotamponade in 60 (17.1%) eyes, C3F8 in 105 (30%) eyes, silicone oil in 65 (18.7%) eyes, air in 74 (21.1%) eyes, and fluid in 46 (13.1%) eyes. Those that were left with fluid were more hypotonic than gas or air, but a significant difference did not exist between the postoperative hypotony rates of eyes in which different endotamponades were applied. This has led to the conclusion that the scleral incision technique may be more important than the choice of endotamponade type in the prevention of leakage or hypotony. As stated by Fine et al, [bib_ref] Outcomes of 77 consecutive cases of 23-gauge transconjunctival sutureless vitrectomy for posterior..., Fine [/bib_ref] and Avcı and Alyamaç, [bib_ref] 23G vitrectomy surgery, Avcı [/bib_ref] the 23G vitrectomy technique is a more reliable one than 25G and carries fewer complication risks.
In conclusion, 23G transconjunctival vitrectomy system is safe and effective in a wide field of vitreoretinal conditions. It is a good alternative to 20G and 25G techniques but needs some improvement, mainly in regard to the instruments and related techniques; further larger controlled group studies are needed.
[fig] Figure 1: Trocar and infusion canula positions in 23 gauge vitrectomy. [/fig]
[fig] Figure 2: Specially designed pressure plate for holding the conjunctiva. [/fig]
[table] Table 1 23: Gauge vitrectomy indications [/table]
[table] Table 2: Mean preoperative and postoperative visual acuities [/table]
|
Development of Singlet Oxygen Luminescence Kinetics during the Photodynamic Inactivation of Green Algae
Recent studies show the feasibility of photodynamic inactivation of green algae as a vital step towards an effective photodynamic suppression of biofilms by using functionalized surfaces. The investigation of the intrinsic mechanisms of photodynamic inactivation in green algae represents the next step in order to determine optimization parameters. The observation of singlet oxygen luminescence kinetics proved to be a very effective approach towards understanding mechanisms on a cellular level. In this study, the first two-dimensional measurement of singlet oxygen kinetics in phototrophic microorganisms on surfaces during photodynamic inactivation is presented. We established a system of reproducible algae samples on surfaces, incubated with two different cationic, antimicrobial potent photosensitizers. Fluorescence microscopy images indicate that one photosensitizer localizes inside the green algae while the other accumulates along the outer algae cell wall. A newly developed setup allows for the measurement of singlet oxygen luminescence on the green algae sample surfaces over several days. The kinetics of the singlet oxygen luminescence of both photosensitizers show different developments and a distinct change over time, corresponding with the differences in their localization as well as their photosensitization potential. While the complexity of the signal reveals a challenge for the future, this study incontrovertibly marks a crucial, inevitable step in the investigation of photodynamic inactivation of biofilms: it shows the feasibility of using the singlet oxygen luminescence kinetics to investigate photodynamic effects on surfaces and thus opens a field for numerous investigations.
# Introduction
Biofilms play a major role in biofouling and the biodeterioration of construction materials. The biodeterioration of a building not only compromises its aesthetic appearance, but can also destroy cultural heritage when it comes to ancient buildings and even endanger people when the structural integrity is affected [bib_ref] Review of concrete biodeterioration in relation to nuclear waste, Turick [/bib_ref] [bib_ref] Aeroterrestrial microalgae growing in biofilms on facades-response to temperature and water stress, Häubner [/bib_ref].
Established methods for the suppression or removal of biofilms have major drawbacks: mechanical removal of biofilms comes along with removal of construction material, while chemical approaches like the use of biocides pose environmental risks. This has motivated numerous efforts to find alternate ways to remove/reduce biofilms or supress their formation.
The functionalized surfaces approach appears to be a highly suitable and promising general tactic [bib_ref] Evaluation of inhibitory effect of TiO 2 nanocoatings against microalgal growth on..., Graziani [/bib_ref] [bib_ref] Silver nanoparticulate enhanced aqueous silane/siloxane exterior facade emulsions and their efficacy against..., Macmullen [/bib_ref] [bib_ref] A Review of Heterogeneous Photocatalysis for Water and Surface Disinfection, Byrne [/bib_ref]. Most biofilms on outdoor surfaces contain phototrophic organisms, which need light to Molecules 2016, 21, 485 2 of 11 exist [bib_ref] Microbial diversity on a marble monument: A case study, Hallmann [/bib_ref] , and photocatalytic approaches have hence been adopted and displayed numerous interesting results up to now. The use of titanium dioxide in various forms stands out in that regard, though it was proven to be less effective for phototrophic organisms in field studies than under laboratory conditions. It was shown that titanium dioxide-functionalized surfaces can generate NOx species, due to electron transfer-based photocatalytic effects, and these species may act as fertilizers for the microorganisms [bib_ref] Evaluation of inhibitory effect of TiO 2 nanocoatings against microalgal growth on..., Graziani [/bib_ref] [bib_ref] Silver nanoparticulate enhanced aqueous silane/siloxane exterior facade emulsions and their efficacy against..., Macmullen [/bib_ref] [bib_ref] A Review of Heterogeneous Photocatalysis for Water and Surface Disinfection, Byrne [/bib_ref] [bib_ref] Photodegradation of aniline in aqueous suspensions of microalgae, Wang [/bib_ref] [bib_ref] Advantages and drawbacks in the field of construction and building materials, Pacheco-Torgal [/bib_ref] [bib_ref] Influence of material properties and photocatalysis on phototrophic growth in multi-year roof..., Gladis [/bib_ref].
Our aim is the development of functionalized surfaces, taking advantage of the photodynamic effect, as an attractive alternative to the photocatalytic effects that, e.g., titanium dioxide-functionalized surfaces, are based on. The idea of surface-immobilized photosensitizers for inhibiting the growth of biofilms is promising for several reasons. Efficient photosensitizers have been identified and developed with regard to their application in clinical photodynamic therapy (PDT) for many decades. Even though the requirements for photosensitizers differ between PDT and the photodynamic inactivation (PDI) of microorganisms, a variety of photosensitizers for PDI are available [bib_ref] Photoinactivation of Escherichia coli (SURE2) without intracellular uptake of the photosensitizer, Preuss [/bib_ref]. They all have an intense UV/vis absorption spectrum, with quite a dominant absorption in the visible part in common. In contrast to titanium dioxide, which absorbs in the UV, surfaces functionalized with photosensitizers would use a much larger part of the Sun's emission spectrum reaching our planet. Another advantage of photosensitizers is their low to non-existing dark toxicity, having been developed for medical use in most cases. In contrast to that, the concern regarding the toxicity of titanium dioxide-and nanoparticle-based photocatalysis applications to people and environment is raising, as the use of titanium dioxide in several fields of application is increasing [bib_ref] A Review of Heterogeneous Photocatalysis for Water and Surface Disinfection, Byrne [/bib_ref] [bib_ref] Chances and limitations of nanosized titanium dioxide practical application in view of..., Bogdan [/bib_ref] [bib_ref] Titanium dioxide nanoparticles: A review of current toxicological data, Shi [/bib_ref].
Even though PDT is in widespread use todayand photodynamic inactivation of bacteria (PIB) is becoming an acknowledged approach [bib_ref] Photoinactivation of Escherichia coli (SURE2) without intracellular uptake of the photosensitizer, Preuss [/bib_ref] [bib_ref] Photodynamic inactivation of mold fungi spores by newly developed charged corroles, Preuß [/bib_ref] [bib_ref] Fast and effective photodynamic inactivation of multiresistant bacteria by cationic riboflavin derivatives, Maisch [/bib_ref] [bib_ref] Fast and effective inactivation of Bacillus atrophaeus endospores using light-activated derivatives of..., Eichner [/bib_ref] [bib_ref] Resistance in antimicrobial photodynamic inactivation of bacteria, Maisch [/bib_ref] [bib_ref] The role of singlet oxygen and oxygen concentration in photodynamic inactivation of..., Maisch [/bib_ref] , only a few examples on PDI of phototrophic organisms have been reported [bib_ref] Photodynamic therapy against cyanobacteria, Drábková [/bib_ref] [bib_ref] Development of a biocidal treatment regime to inhibit biological growths on cultural..., Young [/bib_ref] [bib_ref] Photosensitized Destruction of Chlorella vulgaris by Methylene Blue or Nuclear Fast Red..., Mccullagh [/bib_ref] [bib_ref] Inhibition of green algae growth by corrole-based photosensitizers, Pohl [/bib_ref]. A first preliminary step towards developing photoactive surfaces using the photodynamic effect, was reported by Pohl et al., who revealed the feasibility of green algae inhibition using the photodynamic effect [bib_ref] Inhibition of green algae growth by corrole-based photosensitizers, Pohl [/bib_ref].
For advancing the PDI of phototrophic microorganisms beyond a mere proof of concept, we considered that the next step should be to evaluate the utilization of direct singlet oxygen ( 1 O 2 ) luminescence and photosensitizer fluorescence measurements to characterize and optimize the conditions for photodynamic inactivation of green algae. Fluorescence spectroscopy in general is a powerful tool; however, it was proven in the past that the combination of fluorescence spectroscopy and the time-resolved detection of the phosphorescence of generated 1 O 2 is the most effective approach to investigate photodynamic inactivation processes. While the photosensitizer 1 s fluorescence may be analysed for deducing on its location and possible aggregation, the analysis of the 1 O 2 luminescence kinetics provides an insight on the microenvironment of the photosensitizer. This may be used for revealing the biochemical mechanism of the inactivation, thus allowing for optimization of the photosensitizer in terms of localisation and effective singlet oxygen generation [bib_ref] Singlet oxygen luminescence kinetics in a heterogeneous environment-identification of the photosensitizer localization..., Hackbarth [/bib_ref] [bib_ref] Highly sensitive time resolved singlet oxygen luminescence detection using LEDs as the..., Hackbarth [/bib_ref]. [fig_ref] Figure 1: Figure 1 [/fig_ref] shows photographs, spatial fluorescence and 1 O 2 phosphorescence plots of the algae reference, the PCor + -algae, and the TMPyP-algae samples after one day of incubation (on day 2). Photographs of the samples are presented in the first column, the spatial distribution of the fluorescence in the second and the spatial distribution of the 1 O 2 phosphorescence in the third column. The photographs illustrate the sample geometry and the inhomogeneity due to the sample preparation technique. The fluorescence intensity in the plot equates the integral over the spectral range of 550-900 nm. The 1 O 2 phosphorescence intensity shown in [fig_ref] Figure 1: Figure 1 [/fig_ref] equates the fitted amplitude.
# Results
As expected, the algae reference is fluorescent, but devoid of any 1 O 2 phosphorescence. The TMPyP-algae sample shows spatial correlation of fluorescence and 1 O 2 phosphorescence even though the sharp edges of the sample that can be seen in the photograph and in the fluorescence plot appear smooth in the 1 O 2 phosphorescence plot. In contrast to this high spatial correlation, there is hardly any spatial correlation between the photograph and the fluorescence plot on one hand and the 1 O 2 phosphorescence plot on the other hand in the PCor + -algae sample. Comparison of the photographs, fluorescence intensity plots and singlet oxygen intensity plots of all samples. As expected, the algae reference does not display any 1 O2 phosphorescence. The TMPyP-algae-sample shows high spatial correlation of the photograph, and both the fluorescence and 1 O2 phosphorescence intensity plots. Interestingly, the 1 O2 phosphorescence intensity plot of the PCor + -algae-sample appears to contradict the fluorescence intensity plot and the photograph.
The abovementioned puzzling mismatch was resolved by recording the full fluorescence spectra of all samples on day 2. [fig_ref] Figure 2: Comparison of fluorescence spectra of PCor + and TMPyP with algae to... [/fig_ref] shows the fluorescence spectra summed up over five pixels of highest signal intensity, which reveals that: (a) the fluorescence intensity due to the algae′s photosynthesis system is much stronger than that of the photosensitizers; (b) hardly any information can be drawn from the integrated fluorescence measurements of the algae/TMPyP combination because of the almost complete overlap of the emission spectra of the components; (c) the PCor + ′s fluorescence spectrum is located partly outside the algae emission, which may be used for its separate evaluation. The abovementioned puzzling mismatch was resolved by recording the full fluorescence spectra of all samples on day 2. [fig_ref] Figure 2: Comparison of fluorescence spectra of PCor + and TMPyP with algae to... [/fig_ref] shows the fluorescence spectra summed up over five pixels of highest signal intensity, which reveals that: (a) the fluorescence intensity due to the algae 1 s photosynthesis system is much stronger than that of the photosensitizers; (b) hardly any information can be drawn from the integrated fluorescence measurements of the algae/TMPyP combination because of the almost complete overlap of the emission spectra of the components; (c) the PCor +1 s fluorescence spectrum is located partly outside the algae emission, which may be used for its separate evaluation. TMPyP-algae-sample shows high spatial correlation of the photograph, and both the fluorescence and 1 O2 phosphorescence intensity plots. Interestingly, the 1 O2 phosphorescence intensity plot of the PCor + -algae-sample appears to contradict the fluorescence intensity plot and the photograph.
The abovementioned puzzling mismatch was resolved by recording the full fluorescence spectra of all samples on day 2. [fig_ref] Figure 2: Comparison of fluorescence spectra of PCor + and TMPyP with algae to... [/fig_ref] shows the fluorescence spectra summed up over five pixels of highest signal intensity, which reveals that: (a) the fluorescence intensity due to the algae′s photosynthesis system is much stronger than that of the photosensitizers; (b) hardly any information can be drawn from the integrated fluorescence measurements of the algae/TMPyP combination because of the almost complete overlap of the emission spectra of the components; (c) the PCor + ′s fluorescence spectrum is located partly outside the algae emission, which may be used for its separate evaluation. The latter hypothesis was confirmed by the results depicted in [fig_ref] Figure 3: Comparison of the fluorescence intensity plots integrated over the whole recorded spectrum... [/fig_ref]. The left plot shows the fluorescence intensity as above, i.e., integrated from 550 to 900 nm. The plot in the middle shows the fluorescence intensity integrated only over the range of 550-650 nm, so as to isolate the PCor +1 s fluorescence. This fluorescence plot is in correlation with the 1 O 2 phosphorescence plot, indicating a diffusion of the photosensitizer onto the filter paper. The latter hypothesis was confirmed by the results depicted in [fig_ref] Figure 3: Comparison of the fluorescence intensity plots integrated over the whole recorded spectrum... [/fig_ref]. The left plot shows the fluorescence intensity as above, i.e., integrated from 550 to 900 nm. The plot in the middle shows the fluorescence intensity integrated only over the range of 550-650 nm, so as to isolate the PCor + ′s fluorescence. This fluorescence plot is in correlation with the 1 O2 phosphorescence plot, indicating a diffusion of the photosensitizer onto the filter paper.
This evaluation of the spatial distribution of the fluorescence signal illustrates the importance of combining it with direct 1 O2 phosphorescence measurements. As mentioned in the Introduction, the 1 O2 kinetics provides information about the microenvironment of the photosensitizer. The most vital requirement for a detailed analysis of 1 O2 kinetics is a sufficient signal strength. To check this requirement and to explain how the later shown data was obtained, an excursion to the performed data analysis is necessary. [fig_ref] Figure 3: Comparison of the fluorescence intensity plots integrated over the whole recorded spectrum... [/fig_ref]. Comparison of the fluorescence intensity plots integrated over the whole recorded spectrum and the for the PCor + fluorescence relevant part of the recorded spectrum and the 1 O2 phosphorescence intensity plot for a PCor + -algae sample. In contrast to TMPyP, PCor + appears to diffuse in the surrounding filter paper.
During this study, no changes were observed between the first and second scans of each sample on each day. Therefore, the signals of both scans per day were summed up pixel by pixel for each sample. The signal for the analysis and comparison of the samples and their temporal evolution were obtained by estimating the signal intensity of each pixel, evaluating the five pixels with highest intensity and calculating the sum of the signal of these five pixels.
To facilitate the comparison of the different signals, the standard biexponential model as described in [bib_ref] In vivo detection of time-resolved singlet oxygen luminescence under PDT relevant conditions, Schlothauer [/bib_ref] was extended by additional exponential terms in order to fit the data. The standard biexponential model for 1 O2 kinetics is the solution of a rate equation contemplating the generation and deactivation processes of 1 O2. The necessity of extending the biexponential model depends on the sample and is discussed along with the respective results. In general, the extension of the standard biexponential model was required because it describes 1 O2 kinetics in a homogeneous microenvironment, a condition that is apparently not experienced by the photosensitizers in the samples described in this study.
This approach was chosen to provide facile comparison of the signals and to visualize their development. As it is purely phenomenological for this study, no fitted parameter will be discussed here. The investigation of the applicability of this approach and the detailed analysis of the kinetics is an object of future work. [fig_ref] Figure 4: Raw data and fitted model functions [/fig_ref] shows the raw data and the fitted model functions for the first two days, obtained in the aforementioned fashion. Even though the detailed analysis of the kinetics parameters is not the subject of this article, which presents a feasibility study, changes of the 1 O2 kinetics during the first two days of measurement and differences in the 1 O2 kinetics between the two photosensitizers as well as references and algae samples are clearly observed. [fig_ref] Figure 5: Temporal development of the singlet oxygen signals, measured on the surfaces of... [/fig_ref] shows the 1 O2 kinetics of the PCor + reference sample and the PCor + -algae sample. The 1 O2 kinetics of the reference sample could be described by the unmodified biexponential model, i.e., without any extension. For the kinetics of the 1 O2 kinetics of the PCor + -algae sample, the biexponential model had to be extended by one additional simple exponential term. This evaluation of the spatial distribution of the fluorescence signal illustrates the importance of combining it with direct 1 O 2 phosphorescence measurements. As mentioned in the Introduction, the 1 O 2 kinetics provides information about the microenvironment of the photosensitizer. The most vital requirement for a detailed analysis of 1 O 2 kinetics is a sufficient signal strength. To check this requirement and to explain how the later shown data was obtained, an excursion to the performed data analysis is necessary.
During this study, no changes were observed between the first and second scans of each sample on each day. Therefore, the signals of both scans per day were summed up pixel by pixel for each sample. The signal for the analysis and comparison of the samples and their temporal evolution were obtained by estimating the signal intensity of each pixel, evaluating the five pixels with highest intensity and calculating the sum of the signal of these five pixels.
To facilitate the comparison of the different signals, the standard biexponential model as described in [bib_ref] In vivo detection of time-resolved singlet oxygen luminescence under PDT relevant conditions, Schlothauer [/bib_ref] was extended by additional exponential terms in order to fit the data. The standard biexponential model for 1 O 2 kinetics is the solution of a rate equation contemplating the generation and deactivation processes of 1 O 2 . The necessity of extending the biexponential model depends on the sample and is discussed along with the respective results. In general, the extension of the standard biexponential model was required because it describes 1 O 2 kinetics in a homogeneous microenvironment, a condition that is apparently not experienced by the photosensitizers in the samples described in this study.
This approach was chosen to provide facile comparison of the signals and to visualize their development. As it is purely phenomenological for this study, no fitted parameter will be discussed here. The investigation of the applicability of this approach and the detailed analysis of the kinetics is an object of future work. [fig_ref] Figure 4: Raw data and fitted model functions [/fig_ref] shows the raw data and the fitted model functions for the first two days, obtained in the aforementioned fashion. Even though the detailed analysis of the kinetics parameters is not the subject of this article, which presents a feasibility study, changes of the 1 O 2 kinetics during the first two days of measurement and differences in the 1 O 2 kinetics between the two photosensitizers as well as references and algae samples are clearly observed. [fig_ref] Figure 5: Temporal development of the singlet oxygen signals, measured on the surfaces of... [/fig_ref] shows the 1 O 2 kinetics of the PCor + reference sample and the PCor + -algae sample. The 1 O 2 kinetics of the reference sample could be described by the unmodified biexponential model, i.e., without any extension. For the kinetics of the 1 O 2 kinetics of the PCor + -algae sample, the biexponential model had to be extended by one additional simple exponential term. [fig_ref] Figure 6: Temporal development of the singlet oxygen signals, measured on the surfaces of... [/fig_ref] shows the 1 O2 kinetics of the TMPyP reference sample and the TMPyP-algae sample. An additional simple exponential term has to be added to the biexponential model to describe the singlet oxygen luminescence kinetics, even for the TMPyP reference. The most complex kinetics were found for the TMPyP-algae sample, where two additional exponential terms had to be added to the biexponential model to fit the measured kinetic data. It must be pointed out that, as mentioned before, these fits were performed phenomenologically in order to facilitate the comparison of the data. [fig_ref] Figure 6: Temporal development of the singlet oxygen signals, measured on the surfaces of... [/fig_ref] shows the 1 O2 kinetics of the TMPyP reference sample and the TMPyP-algae sample. An additional simple exponential term has to be added to the biexponential model to describe the singlet oxygen luminescence kinetics, even for the TMPyP reference. The most complex kinetics were found for the TMPyP-algae sample, where two additional exponential terms had to be added to the biexponential model to fit the measured kinetic data. It must be pointed out that, as mentioned before, these fits were performed phenomenologically in order to facilitate the comparison of the data. [fig_ref] Figure 6: Temporal development of the singlet oxygen signals, measured on the surfaces of... [/fig_ref] shows the 1 O 2 kinetics of the TMPyP reference sample and the TMPyP-algae sample. An additional simple exponential term has to be added to the biexponential model to describe the singlet oxygen luminescence kinetics, even for the TMPyP reference. The most complex kinetics were found for the TMPyP-algae sample, where two additional exponential terms had to be added to the biexponential model to fit the measured kinetic data. It must be pointed out that, as mentioned before, these fits were performed phenomenologically in order to facilitate the comparison of the data. [fig_ref] Figure 6: Temporal development of the singlet oxygen signals, measured on the surfaces of... [/fig_ref] shows the 1 O2 kinetics of the TMPyP reference sample and the TMPyP-algae sample. An additional simple exponential term has to be added to the biexponential model to describe the singlet oxygen luminescence kinetics, even for the TMPyP reference. The most complex kinetics were found for the TMPyP-algae sample, where two additional exponential terms had to be added to the biexponential model to fit the measured kinetic data. It must be pointed out that, as mentioned before, these fits were performed phenomenologically in order to facilitate the comparison of the data. Moreover, measurements at 1210 nm show a weak NIR signal for all TMPyP samples, in contrast to the algae references and all PCor + samples where no signal was observed. The signal decreases exponentially over several µs, corresponding to one of the additional exponential terms of the 1 O 2 kinetics. Due to the kinetics and the design of both, the laser and the 1 O 2 detection system, a measurement artefact is very unlikely. Since the signal is not significant in comparison to the here presented signals, it is not shown here.
Intra-vs. extra-cellular localization of the photosensitizers in the green algae after incubation was evaluated via CLSM and compared to the intrinsic fluorescence of chromophores (chlorophylls) inside the algae. In [fig_ref] Figure 7: Localisation of the photosensitizer in Chlorella fusca var [/fig_ref] , the resulting images of an untreated reference as well as samples incubated with either one of the two PS are shown alongside scattering images of the monitored cells. In contrast to scanning the fluorescence on macroscopic areas, CLSM imaging allows for separation of the fluorescence signals of PCor + and intrinsic chlorophylls by eliminating cross talk between channels using untreated reference samples. The (red coded) fluorescence of PCor + is apparent only at the periphery of the live cells (seen as circular lines of varying thickness), there is absolutely no overlap of signals attributed to intracellular chlorophyll and PCor + , and the few cases where PCor +1 s fluorescence appears round may be safely attributed to dead cells. The latter conclusion may also be deduced from the scattering image where no apparent cellular structures can be found in those regions. Moreover, measurements at 1210 nm show a weak NIR signal for all TMPyP samples, in contrast to the algae references and all PCor + samples where no signal was observed. The signal decreases exponentially over several μs, corresponding to one of the additional exponential terms of the 1 O2 kinetics. Due to the kinetics and the design of both, the laser and the 1 O2 detection system, a measurement artefact is very unlikely. Since the signal is not significant in comparison to the here presented signals, it is not shown here.
Intra-vs. extra-cellular localization of the photosensitizers in the green algae after incubation was evaluated via CLSM and compared to the intrinsic fluorescence of chromophores (chlorophylls) inside the algae. In [fig_ref] Figure 7: Localisation of the photosensitizer in Chlorella fusca var [/fig_ref] , the resulting images of an untreated reference as well as samples incubated with either one of the two PS are shown alongside scattering images of the monitored cells. In contrast to scanning the fluorescence on macroscopic areas, CLSM imaging allows for separation of the fluorescence signals of PCor + and intrinsic chlorophylls by eliminating cross talk between channels using untreated reference samples. The (red coded) fluorescence of PCor + is apparent only at the periphery of the live cells (seen as circular lines of varying thickness), there is absolutely no overlap of signals attributed to intracellular chlorophyll and PCor + , and the few cases where PCor + ′s fluorescence appears round may be safely attributed to dead cells. The latter conclusion may also be deduced from the scattering image where no apparent cellular structures can be found in those regions. In contrast, fluorescence of TMPyP originates almost entirely from areas that also emit fluorescence of algae-chromophores. The distribution of the TMPyP fluorescence varies from fully filled circular areas (see [fig_ref] Figure 7: Localisation of the photosensitizer in Chlorella fusca var [/fig_ref] , arrow 1) to detailed structures (see [fig_ref] Figure 7: Localisation of the photosensitizer in Chlorella fusca var [/fig_ref] , arrow 2). The scattering image of the same sample shows cellular structures at all areas of fluorescence origin, with varying apparent viability of the cells.
# Discussion
The obvious mismatch in the spatial correlation of fluorescence and 1 O2 phosphorescence for the PCor + -algae sample in [fig_ref] Figure 3: Comparison of the fluorescence intensity plots integrated over the whole recorded spectrum... [/fig_ref] emphasizes once more the importance of direct 1 O2 phosphorescence detection for the investigation of photodynamic inactivation processes. The spatial distribution of the [fig_ref] Figure 7: Localisation of the photosensitizer in Chlorella fusca var [/fig_ref]. Localisation of the photosensitizer in Chlorella fusca var. vacuolata after three days of incubation. PCor + appears to locate outside the algae cells while TMPyP appears to accumulate inside the algae cells.
In contrast, fluorescence of TMPyP originates almost entirely from areas that also emit fluorescence of algae-chromophores. The distribution of the TMPyP fluorescence varies from fully filled circular areas (see [fig_ref] Figure 7: Localisation of the photosensitizer in Chlorella fusca var [/fig_ref] , arrow 1) to detailed structures (see [fig_ref] Figure 7: Localisation of the photosensitizer in Chlorella fusca var [/fig_ref] , arrow 2). The scattering image of the same sample shows cellular structures at all areas of fluorescence origin, with varying apparent viability of the cells.
# Discussion
The obvious mismatch in the spatial correlation of fluorescence and 1 O 2 phosphorescence for the PCor + -algae sample in [fig_ref] Figure 3: Comparison of the fluorescence intensity plots integrated over the whole recorded spectrum... [/fig_ref] emphasizes once more the importance of direct 1 O 2 phosphorescence detection for the investigation of photodynamic inactivation processes. The spatial distribution of the 1 O 2 phosphorescence further indicates a diffusion of this photosensitizer into the surrounding filter paper, which is much more significant than for TMPyP. The 1 O 2 phosphorescence examinations reveal a nearly homogeneous allocation for the PCor + -algae sample, while only the smoothing of the edges indicates a diffusion of TMPyP. This kind of effect would probably go unnoticed by looking only at the spatial correlation of the fluorescence, especially for TMPyP, where it is impossible to separate the TMPyP fluorescence and algae autofluorescence.
CLSM images clearly indicate a localisation of PCor + mainly onto the cell wall of the algae while TMPyP appears to accumulate inside the algae cell. With respect to future applications regarding the suppression of biofilms, photosensitizers accumulating and acting from the outside are clearly preferable.
The development of the 1 O 2 phosphorescence kinetics of the PCor + reference shows a decrease of the signal with hardly any change in the kinetics. This indicates either a photobleaching of the photosensitizer or a change of the photosensitizer concentration on the investigated surface. Since PCor + is highly water soluble and did not show any signs of photobleaching in prior studies, the possibility of a diffusion of the photosensitizer in the BBM-agar film appears much more likely.
The development of the 1 O 2 phosphorescence kinetics of the TMPyP reference raises more issues than that of the PCor + reference. The decrease of the signal indicates a similar diffusion in the BBM-agar film. In prior experiments, TMPyP showed such diffusion into the agar substrates.
The percentage of exponential increase in the signal, developing after 4 days, is due to unobvious effects that will have to be investigated in the future. The very weak luminescence at 1210 nm, whose kinetics corresponds to one of the additional exponential terms of the 1 O 2 kinetics, which was measured for all TMPyP samples, indicates an autophosphorescence of TMPyP. However, this auto-phosphorescence alone obviously cannot explain the signal development.
The development of the 1 O 2 phosphorescence kinetics of all the algae-photosensitizer samples have two effects in common: after one day of incubation the 1 O 2 phosphorescence kinetics changes drastically and an exponential increase in the signal is observed over time. Since the first measurement was performed right after the sample preparation, it may safely be assumed that nearly no interaction of algae and photosensitizer took place. That explains the similarity of the measurements of the photosensitizer reference measurements and the photosensitizer-algae samples on day one.
After one day of incubation, the 1 O 2 phosphorescence kinetics of the PCor + -algae sample reaches a nearly stable condition for the rest of the measurement period. Only the amplitude decreases slightly. A correlation of this decrease with cell viability is a reasonable assumption, which however must be further validated by viability assays. This behaviour is in line with the indications for localisation of the PCor + on the outside of the cell wall, since this kind of process should take place in a time period of less than a day. In this context, the development of the 1 O 2 phosphorescence kinetics of the TMPyP-algae samples also correlates with the assumed localisation of the TMPyP inside the algae, as intracellular uptake of this and related porphyrins has been shown to require several days. In this study, no 1 O 2 phosphorescence signal was measured on green algae reference samples, despite being formed during photosynthesis. This is due to the abundance of naturally occurring molecules like carotenoids inside phototrophic organisms, which act as quenchers for naturally generated amounts of 1 O 2 . As shown in [bib_ref] Inhibition of green algae growth by corrole-based photosensitizers, Pohl [/bib_ref] , this does not hinder the photodynamic inactivation of green algae. It was shown herein for the first time that the measurement of 1 O 2 phosphorescence kinetics on surfaces containing phototrophic organisms is possible. Even though the kinetics are complex, a distinct development of the kinetics over the four days of measurement that correlates with the localization of the photosensitizer can be observed.
# Materials and methods
An experimental setup was constructed for obtaining reproducible scans of luminescence on sample surfaces [fig_ref] Figure 8: Scheme of 1 O2 scanning setup [/fig_ref]. It consists of an LDM-405D excitation laser (Omikron-Laserage, Rodgau-Dudenhofen, Germany) a cross table allowing one to scan a sample in the X and Y directions, a detection optics system coupling the excited luminescence into a fibre, movable in the Z direction, a TCMP-1270 Singlet Oxygen Luminescence Detection System by SHB Analytics (Berlin, Germany) and a C10083CAH Fluorescence Spectrometer by Hamamatsu (Hamamatsu, Shizuoka, Japan). The TCMPC-1270 Singlet Oxygen Detection System allows for the time-resolved measurement of NIR-luminescence with highest sensitivity. Using a Hamamtsu H10330 photomultiplier tube, TCMPC electronics with time frames of 80 μs, 160 μs and 320 μs, and optical band path filters with centre wavelengths of 1270 nm and 1210 nm (both ± 15 nm), it is optimized for the detection of very weak singlet oxygen luminescence signals. The Omikron LDM-405D laser is a 2 W diode laser of the TA Deepstar series with a wavelength of 405 nm. With a modulation speed of up to 150 MHz it is highly suitable as excitation laser in combination with the TCMPC-1270. The laser is PC controlled and modulated by the TCMPC-1270. The laser beam is focused to a spot of ca. 150 μm. The size of the laser spot is a good approximation for a lower boundary of the spatial resolution. The scanning motion is realized with custom build mechanics and software-controlled stepping motors.
The fluorescence detection can be realized by either coupling the detection fibre directly in the Hamamatsu C10083CAH fluorescence spectrometer or by using a dichroic mirror with a longpass cutoff wavelength of 1000 nm to detect fluorescence and 1 O2 phosphorescence simultaneously. To avoid the signal losses due to the dichroic mirror, in the later presented data, fluorescence and 1 O2 phosphorescence were measured sequentially.
A strain of green algae Chlorella fusca var. vacuolata (SAG 211-8b) was used as model organism for measurement of 1 O2 phosphorescence in phototrophic organisms. Suspension cultures of the algae were inoculated 3 days prior to experiments and grown at room temperature with shaking on a rotary shaker at 250 rpm in Bold's Basal Medium (according to. Illumination of the algal cultures during cultivation was realized with a daylight bulb (Photographic Lamp, 5400 K, Realm Industrial GmbH, Berlin, Germany) in a day-night cycle of 12 h:12 h. All suspension cultures were provided with fresh medium directly before the start of any of the experiments. 5,10,15,20-Tetrakis(1-methylpyridinium-4-yl)porphyrin tetra(p-toluenesulfonate) (TMPyP) and 5,10,15-tris-(1-methylpyridinium-2-yl)corrolato-(trans-dihydroxo)phosphorus(V) (PCor + ), the corrolebased photosensitizer reported in [bib_ref] Photodynamic inactivation of mold fungi spores by newly developed charged corroles, Preuß [/bib_ref] [bib_ref] Inhibition of green algae growth by corrole-based photosensitizers, Pohl [/bib_ref] , were used as photosensitizers [fig_ref] Figure 9: Chemical structures of the photosensitizers TMPyP and PCor + [/fig_ref]. TMPyP (CAS 36951-72-1) was purchased from Sigma-Aldrich Chemie GmbH (Munich, Germany), while the tricationic metallocorrole PCor + was synthesized according to [bib_ref] Photodynamic inactivation of mold fungi spores by newly developed charged corroles, Preuß [/bib_ref].
For obtaining reproducible sample surfaces, algal cultures for the experiments were grown on filter paper (∅90 mm) on Bold′s Basal Medium (BBM) agar plates. To this end, 200 μL of algal suspension cultures with an initial cell density of 7 × 10 6 mL −1 were placed on a circular area with a diameter of 5-7 mm. Photosensitizers were added to the cultures prior to placement on the filter paper in a concentration of 5 μmol/L. Samples of photosensitizer without green algae and green algae without photosensitizer were prepared as positive and negative controls for the singlet oxygen luminescence measurements, respectively.
The samples were covered with a fused silica window. This allows a high reproducibility of the singlet oxygen luminescence scan and prevents the samples from drying out. The samples were exposed to the 12 h day-night cycle using the 5400 K 20 W daylight bulbs for 4 days. The TCMPC-1270 Singlet Oxygen Detection System allows for the time-resolved measurement of NIR-luminescence with highest sensitivity. Using a Hamamtsu H10330 photomultiplier tube, TCMPC electronics with time frames of 80 µs, 160 µs and 320 µs, and optical band path filters with centre wavelengths of 1270 nm and 1210 nm (both˘15 nm), it is optimized for the detection of very weak singlet oxygen luminescence signals. The Omikron LDM-405D laser is a 2 W diode laser of the TA Deepstar series with a wavelength of 405 nm. With a modulation speed of up to 150 MHz it is highly suitable as excitation laser in combination with the TCMPC-1270. The laser is PC controlled and modulated by the TCMPC-1270. The laser beam is focused to a spot of ca. 150 µm. The size of the laser spot is a good approximation for a lower boundary of the spatial resolution. The scanning motion is realized with custom build mechanics and software-controlled stepping motors.
The fluorescence detection can be realized by either coupling the detection fibre directly in the Hamamatsu C10083CAH fluorescence spectrometer or by using a dichroic mirror with a longpass cutoff wavelength of 1000 nm to detect fluorescence and 1 O 2 phosphorescence simultaneously. To avoid the signal losses due to the dichroic mirror, in the later presented data, fluorescence and 1 O 2 phosphorescence were measured sequentially.
A strain of green algae Chlorella fusca var. vacuolata (SAG 211-8b) was used as model organism for measurement of 1 O 2 phosphorescence in phototrophic organisms. Suspension cultures of the algae were inoculated 3 days prior to experiments and grown at room temperature with shaking on a rotary shaker at 250 rpm in Bold's Basal Medium (according to. Illumination of the algal cultures during cultivation was realized with a daylight bulb (Photographic Lamp, 5400 K, Realm Industrial GmbH, Berlin, Germany) in a day-night cycle of 12 h:12 h. All suspension cultures were provided with fresh medium directly before the start of any of the experiments. 5,10,15,20-Tetrakis(1-methylpyridinium-4-yl)porphyrin tetra(p-toluenesulfonate) (TMPyP) and 5,10,15-tris-(1-methylpyridinium-2-yl)corrolato-(trans-dihydroxo)phosphorus(V) (PCor + ), the corrole-based photosensitizer reported in [bib_ref] Photodynamic inactivation of mold fungi spores by newly developed charged corroles, Preuß [/bib_ref] [bib_ref] Inhibition of green algae growth by corrole-based photosensitizers, Pohl [/bib_ref] , were used as photosensitizers [fig_ref] Figure 9: Chemical structures of the photosensitizers TMPyP and PCor + [/fig_ref]. TMPyP (CAS 36951-72-1) was purchased from Sigma-Aldrich Chemie GmbH (Munich, Germany), while the tricationic metallocorrole PCor + was synthesized according to [bib_ref] Photodynamic inactivation of mold fungi spores by newly developed charged corroles, Preuß [/bib_ref].
For obtaining reproducible sample surfaces, algal cultures for the experiments were grown on filter paper (H90 mm) on Bold 1 s Basal Medium (BBM) agar plates. To this end, 200 µL of algal suspension cultures with an initial cell density of 7ˆ10 6 mL´1 were placed on a circular area with a diameter of 5-7 mm. Photosensitizers were added to the cultures prior to placement on the filter paper in a concentration of 5 µmol/L. Samples of photosensitizer without green algae and green algae without photosensitizer were prepared as positive and negative controls for the singlet oxygen luminescence measurements, respectively.
The samples were covered with a fused silica window. This allows a high reproducibility of the singlet oxygen luminescence scan and prevents the samples from drying out. The samples were exposed to the 12 h day-night cycle using the 5400 K 20 W daylight bulbs for 4 days. The singlet oxygen luminescence was scanned twice a day on an area of 8 mm by 6 mm with 1 mm pixel width in each direction. At each pixel the NIR luminescence signal was measured for 20 s using the 1270 nm bandwidth filter and an excitation energy of 0.5 μJ. On days 1 and 4, additional scans of the NIR luminescence at 1210 nm were performed under the same conditions.
To get information about the localisation of the photosensitizer in the green algae, incubated samples were examined by confocal laser scanning microscopy (CLSM) using a Fluo-ViewTM FV1000 (Olympus, Hamburg, Germany) after 3 days of incubation. Samples of Chlorella fusca var. vacuolata were grown on 2 cm × 2 cm plastic slides for microscopy (Carl Roth GmbH, Karlsruhe, Germany) incubated in suspension cultures with initial cell density of 7 × 10 6 mL −1 . The photosensitizers were applied to the cultures at a concentration of 5 μmol/L. After 3 days of incubation, the plastic slides were rinsed with BBM and the remaining biofilms and precipitate was examined using a 405 nm laser for excitation. The fluorescence from the samples was detected on two channels: between 570 and 640 nm and above 640 nm using a bandpass filter. To avoid fluorescence crosstalk between those channels, excitation intensity and detector amplification were adjusted using untreated samples of the algae, until the intrinsic fluorescence signal is detected only in the second channel.
## Conflicts of interest:
The authors declare no conflict of interest.
## Abbreviations
The following abbreviations are used in this manuscript:
## Bbm
Bold′s Basal Medium PDT Photodynamic Therapy PDI Photodynamic Inactivation PIB Photodynamic Inactivation of Bacteria 1 O2
Singlet Oxygen TMPyP 5,10,15,20-Tetrakis(1-methylpyridinium-4-yl)porphyrin tetra(p-toluenesulfonate) PCor + 5,10,15-tris-(1-methylpyridinium-2-yl)corrolato-(trans-dihydroxo)phosphorus(V) CLSM Confocal Laser Scanning Spectroscopy The singlet oxygen luminescence was scanned twice a day on an area of 8 mm by 6 mm with 1 mm pixel width in each direction. At each pixel the NIR luminescence signal was measured for 20 s using the 1270 nm bandwidth filter and an excitation energy of 0.5 µJ. On days 1 and 4, additional scans of the NIR luminescence at 1210 nm were performed under the same conditions.
To get information about the localisation of the photosensitizer in the green algae, incubated samples were examined by confocal laser scanning microscopy (CLSM) using a Fluo-ViewTM FV1000 (Olympus, Hamburg, Germany) after 3 days of incubation. Samples of Chlorella fusca var. vacuolata were grown on 2 cmˆ2 cm plastic slides for microscopy (Carl Roth GmbH, Karlsruhe, Germany) incubated in suspension cultures with initial cell density of 7ˆ10 6 mL´1. The photosensitizers were applied to the cultures at a concentration of 5 µmol/L. After 3 days of incubation, the plastic slides were rinsed with BBM and the remaining biofilms and precipitate was examined using a 405 nm laser for excitation. The fluorescence from the samples was detected on two channels: between 570 and 640 nm and above 640 nm using a bandpass filter. To avoid fluorescence crosstalk between those channels, excitation intensity and detector amplification were adjusted using untreated samples of the algae, until the intrinsic fluorescence signal is detected only in the second channel.
## Abbreviations
The following abbreviations are used in this manuscript:
[fig] Figure 1: Figure 1. Comparison of the photographs, fluorescence intensity plots and singlet oxygen intensity plots of all samples. As expected, the algae reference does not display any 1 O2 phosphorescence. The TMPyP-algae-sample shows high spatial correlation of the photograph, and both the fluorescence and 1 O2 phosphorescence intensity plots. Interestingly, the 1 O2 phosphorescence intensity plot of the PCor + -algae-sample appears to contradict the fluorescence intensity plot and the photograph. [/fig]
[fig] Figure 2: Comparison of fluorescence spectra of PCor + and TMPyP with algae to photosensitizer and algae reference (fluorescence intensity not corrected for different sample geometry). [/fig]
[fig] Figure 3: Comparison of the fluorescence intensity plots integrated over the whole recorded spectrum and the for the PCor + fluorescence relevant part of the recorded spectrum and the 1 O 2 phosphorescence intensity plot for a PCor + -algae sample. In contrast to TMPyP, PCor + appears to diffuse in the surrounding filter paper. [/fig]
[fig] Figure 4: Raw data and fitted model functions. [/fig]
[fig] Figure 5: Temporal development of the singlet oxygen signals, measured on the surfaces of the PCor + reference samples and the samples of PCor + incubated in green algae over four days. [/fig]
[fig] Figure 6: Temporal development of the singlet oxygen signals, measured on the surfaces of the TMPyP reference sample and the samples of TMPyP incubated in green algae over four days. [/fig]
[fig] Figure 7: Localisation of the photosensitizer in Chlorella fusca var. vacuolata after three days of incubation. PCor + appears to locate outside the algae cells while TMPyP appears to accumulate inside the algae cells. [/fig]
[fig] Figure 8: Scheme of 1 O2 scanning setup. [/fig]
[fig] Figure 9: Chemical structures of the photosensitizers TMPyP and PCor + . [/fig]
|
Repeatability of echocardiographic parameters to evaluate the hemodynamic relevance of patent ductus arteriosus in preterm infants: a prospective observational study
Background: The hemodynamically relevant patent ductus arteriosus in preterm infants is not well defined. Different clinical and echocardiographic parameters are used and the diagnostic accuracy is unknown because of the lack of a gold standard definition. Our study evaluates the inter-observer repeatability of echocardiographic and Doppler-ultrasound parameters.Methods: This prospective observational study included 19 very low birth weight preterm infants (median [interquartile range]: gestational age 28.0 (28.0-29.0) weeks, birth weight 1130 (905-1321) g, postnatal age at measurement 8.7 (4.8-23.5) d) with a clinical suspicion of ductal patency in whom 27 repeated echocardiographic and Doppler-ultrasound examinations were performed within 30 min by 2 of 3 independent observers (54 measurements overall). The repeatability index (=2 times the standard deviation of the differences/mean of all measurements) according to Bland and Altman was used to assess repeatability of different parameters. Results: The repeatability indices of the echocardiographic parameters (left Atrium-to-Aortic root-ratio, diameter of the patent ductus arteriosus at its narrowest part, the left-ventricular-preejection-period-to-ejection-time-ratio and the ratio of the velocity time integrals in the large vessels were 16, 21, 23 and 26 % respectively. The repeatability indices of Doppler-ultrasound measurements (resistance index in celiac artery and anterior cerebral artery) were 11 and 14 %, respectively. Conclusions: The inter-observer repeatability of all echocardiographic parameters was poor compared to that of resistance indices in peripheral vessels. Therefore, interventions for ductal patency should be indicated based on averaged repeated rather than single measurements, especially when measured values are close to their cut-off value -both in clinical routine and for study purposes.
# Background
The patent ductus arteriosus (PDA) in preterm infants is associated with increased mortality and morbidity [bib_ref] A randomized, controlled trial of very early prophylactic ligation of the ductus..., Cassady [/bib_ref] [bib_ref] Patent ductus arteriosus and cerebral circulation in preterm infants, Shortland [/bib_ref] [bib_ref] Changing trends in the epidemiology and pathogenesis of neonatal chronic lung disease, Rojas [/bib_ref] [bib_ref] Early ductal shunting and intraventricular haemorrhage in ventilated preterm infants, Evans [/bib_ref] [bib_ref] Ductal shunting, high pulmonary blood flow, and pulmonary hemorrhage, Kluckow [/bib_ref] [bib_ref] Association between fluid intake and weight loss during the first ten days..., Oh [/bib_ref] [bib_ref] Failure of ductus arteriosus closure is associated with increased mortality in preterm..., Noori [/bib_ref]. However, there is little evidence as to which parameters define a PDA that requires treatment. Zonnenberg and de Waal showed that, besides clinical parameters, echocardiographic and Doppler-ultrasound measurements are used to evaluate the magnitude and clinical relevance of the left-to-right shunt through a PDA, and hence the need for treatment: In a systematic review of 67 randomised controlled trials (RCTs) they described the following most frequently used parameters and applied cut-off values: Left-atrium-to-aortic-root-ratio (LA/Ao-ratio) used in 34 trials, median cut-off >1.3 (range: 1.15-1.70); diastolic reverse flow in peripheral vessels (21 trials); and PDA-diameter (8 trials), cut-off >1.5 (1.5-2.0) mm [bib_ref] The definition of a haemodynamic significant duct in randomized controlled trials: a..., Zonnenberg [/bib_ref].
McNamara and Sehgal suggested a scoring system including clinical and echocardiographic criteria to define hrPDA [bib_ref] Towards rational management of the patent ductus arteriosus: the need for disease..., Mcnamara [/bib_ref]. The echocardiographic part of this staging seems to be predictive for neonatal morbidity and can serve as a guide to clinical decisions [bib_ref] Association between hemodynamically significant patent ductus arteriosus and bronchopulmonary dysplasia, Schena [/bib_ref] , whereas the clinical criteria comprise unspecific respiratory signs. Prospective data suggesting that application of the echocardiographic parameters summarized by Zonnenberg and de Waal or the score by McNamara and Sehgal results in improved outcome is lacking. However, recent retrospective data suggest that echocardiographic screening for PDA within the first 3 postnatal days may reduce mortality in infants born at <29 weeks gestation [bib_ref] Association between early screening for patent ductus arteriosus and in-hospital mortality among..., Rozé [/bib_ref].
To inform future studies and clinical guidelines on PDA treatment, this study aimed to evaluate the inter-observer repeatability of echocardiographic and Doppler-ultrasound parameters, which are frequently determined to assess the need for PDA treatment.
# Methods
This prospective observational cohort study was approved by the research ethics committee at the facutly of medicine and the university hospital of the Eberhard Karls University Tuebingen and written informed parental consent obtained. To assess the repeatability of echocardiographic parameters commonly used to determine the magnitude of the left-to-right shunt, a convenience sample of preterm infants with suspected PDA was analysed. Inclusion criteria were: birth weight ≤1500 g and clinical suspicion of PDA such as cardiac murmur, bounding pulses, ventilator dependency and increased oxygen demand. Syndromal anomalies and congenital heart defects except persisting foramen ovale or atrial septal defect were exclusion criteria. The period of recruitment was between June 2012 and May 2013 at the Department of Neonatology, University Children's Hospital of Tuebingen, University of Tuebingen, Germany.
Within 30 min (to minimize fluctuations in hemodynamic status), 2 of 3 investigators (with more than 20, 10, or 3 years, respectively, of experience in neonatal echocardiography, everyday skills or every week, respectively, 2 were board certified paediatric cardiologists, one investigator is attending physician at the NICU) prospectively and independently performed repeated echocardiographic and Doppler-sonographic measurements including the following parameters:
LA/Ao-ratio [bib_ref] Furosemide promotes patent ductus arteriosus in premature infants with the respiratory-distress syndrome, Green [/bib_ref] ; resistance index (RI) in celiac artery (CA) [bib_ref] Effects of patent ductus arteriosus on left ventricular output and organ blood..., Shimada [/bib_ref] and anterior cerebral artery (ACA) [bib_ref] Abnormal cerebral blood flow patterns in preterm infants with a large patent..., Martin [/bib_ref] ; diameter of the PDA at its narrowest part [bib_ref] Early echocardiographic prediction of symptomatic patent ductus arteriosus in preterm infants undergoing..., Kluckow [/bib_ref] ; the left-ventricular-preejection-period-to-ejection-time-ratio, calculated by including 3-4 cardiac cycles (LVPEP/ LVET) [bib_ref] Dopplersonographic findings in neonates with significant persistent ductus arteriosus, Robel-Tillig [/bib_ref] ; and the ratio of the velocity time integrals in the large vessels (VTI_Ao/VTI_Pa). The VTI_Ao was measured from an apical-5-chamber-view, the VTI_Pa was measured in a parasternal short axis calculated automatically with built-in software. We assumed that, in the absence of congenital heart defects, this ratio correlates with the ductal left-to-right shunt. The PDA-diameter was measured at its narrowest part (identified via colour Doppler) and measured in B-Mode to avoid the influence of gain-settings on the PDA-width if assessed on colour Doppler images.
All measurements were done with a Toshiba "Aplio" using a 6.5 MHz phased array transducer.
Statistical analyses involved repeatability coefficient (RepC = 2 times the standard deviation of the differences) and repeatability index (RepI = RepC/mean of all measurements) according to Bland and Altman [bib_ref] Statistical methods for assessing agreement between two methods of clinical measurement, Bland [/bib_ref] and a confidence-step-analysis (CSA) [bib_ref] Estimation of pulmonary arterial pressure in the newborn: study of the repeatability..., Skinner [/bib_ref]. The RepC represents the upper limit of the 95 % confidence interval of the absolute differences between two measurements performed by two independent observers. The RepI describes the relation between RepC and the mean value of the measurements. This allows comparison of repeatability between different measures. A high CSA value (CSA = difference between lowest and highest value / RepC) indicates that differences between low and high values in this parameter observed in a given population are unlikely related to inter-/intra-observer variability, whereas a CSA of ≤1 indicates that observed differences between low and high values in this population are likely due to inter-/intra-observer variability.
For unexpected RepI differences, 95 % confidence intervals (CI) were exploratorily calculated post-hoc and differences between echo-parameters classified as 'significant' if these 95 %-CI did not overlap.
Data are shown as median (interquartile range).
# Results
Twenty-seven repeated measurements were performed in 19 preterm infants. One infant with a birth weight of 1550 g was included inadvertently due to a weight at the time of measurement of 1465 g. Gestational age at birth was 28.0 (28.0-29.0) weeks, birth weight 1130 (905-1321) g, postnatal age at measurement 8.7 (4.8-23.5) d and weight at the time of measurements 1243 (1024-1528) g. The mean difference in time between the first measurements of the repeated echocardiographic examinations was 12 min with a standard deviation (SD) of 4 min. The mean heart rate while recording left ventricular time intervals during all examinations was 167/min, and mean heart rate difference between repeated examinations was 2/min with a SD of 10/min. Arterial oxygen saturation was targeted at 90-95 % if on supplemental oxygen. Supplemental oxygen was necessary in 8 patients (respiratory support: 4 intubated and ventilated, 4 on binasal CPAP). Of 19 infants in room air, 5 were without respiratory support and 14 on binasal CPAP. No infant required catecholamines, and 3 had indomethacin within 24 h prior to measurements (0.1/0.2/0.4 mg/kg bodyweight/day, respectively).
A left-to-right shunt was identified by colour-Dopplerultrasound in 15/27 measurements. PDA-diameter at the narrowest part could rarely be measured by both investigators (n = 6) because of difficulties visualizing the PDA in its complete course in B-mode (not colour Doppler-mode). The results are presented in [fig_ref] Table 1: Repeatability Index [/fig_ref].
# Discussion
In general, a good diagnostic parameter can easily and quickly be determined and has high repeatability, sensitivity and specificity. Neonatal echocardiography can be performed easily and quickly to determine the need for treatment in preterm infants with PDA, however, the diagnostic accuracy is unknown because of the lack of a gold standard definition of a hrPDA.
This work on the largest cohort reported to date shows that repeatability of neonatal echocardiographic and Doppler-ultrasound parameters in preterm infants with suspected PDA is far from optimal. This is not due to a lack of expertise because our results are in the range of those few reports that previously addressed the issue of the repeatability of echocardiographic parameters in smaller cohorts [fig_ref] Table 2: Repeatability Index [/fig_ref] [bib_ref] Estimation of pulmonary arterial pressure in the newborn: study of the repeatability..., Skinner [/bib_ref] [bib_ref] Echocardiographic assessment of blood flow volume in the superior vena cava and..., Groves [/bib_ref] [bib_ref] Reproducibility of measurements of cardiac output in newborn infants by Doppler ultrasound, Hudson [/bib_ref] [bib_ref] Reproducibility of cerebral artery Doppler measurements, Moorthy [/bib_ref] [bib_ref] Is noninvasive determination of pulmonary artery pressure feasible using deceleration phase Doppler..., Van Dijk [/bib_ref]. However, as summarised in [fig_ref] Table 2: Repeatability Index [/fig_ref] , of the parameters elected here, only the RI_ACA has previously been addressed in a repeatability study.
In fact, our protocol simulated a "best case scenario", as it evaluated repeated measurements by experienced investigators using the same ultrasound device within a short time interval on the same patient. Our study adds that the concerns regarding repeatability raised in the 1990s [bib_ref] Estimation of pulmonary arterial pressure in the newborn: study of the repeatability..., Skinner [/bib_ref] [bib_ref] Echocardiographic assessment of blood flow volume in the superior vena cava and..., Groves [/bib_ref] [bib_ref] Reproducibility of measurements of cardiac output in newborn infants by Doppler ultrasound, Hudson [/bib_ref] [bib_ref] Reproducibility of cerebral artery Doppler measurements, Moorthy [/bib_ref] [bib_ref] Is noninvasive determination of pulmonary artery pressure feasible using deceleration phase Doppler..., Van Dijk [/bib_ref] are still relevant today despite improved ultrasound technology. Nevertheless, knowledge about the poor repeatability has not yet been taken into account in clinical treatment guidelines or current study protocols. The comparability and generalizability of results of data on echocardiography-guided PDA treatment are limited because of differences in the parameters applied and the poor reproducibility of all these parameters.
A large number of echocardiographic and Dopplerultrasound parameters are used to quantify left-to-right shunt through, and hemodynamic relevance of, a PDA (summarised in [bib_ref] The definition of a haemodynamic significant duct in randomized controlled trials: a..., Zonnenberg [/bib_ref]. These may include ductal flow pattern and velocity, absent or reverse diastolic flow in superior mesenteric artery, diastolic flow velocity in left pulmonary artery, reverse flow in descending aorta, and LVO/SVC-flow ratio (left ventricular output/superior vena cava-flow ratio). Some of these parameters may include redundant information [bib_ref] Duration of indomethacin treatment of the preterm patent ductus arteriosus as directed..., Carmo [/bib_ref] , others, such as LVO/ SVC-flow ratio, may not be trivial to measure, because of the complex cross sectional area of the SVC. Our selection of parameters reflects local preferences and was limited to reduce examination time and hence studydriven burden on the infants and subsequent fluctuations of their hemodynamic status in time.
In general, precision of a measurement with poor repeatability can be increased by averaging results of repeated measurements. In the context of this study, the effect of averaging measurements was cut-off dependent: Choosing, for example, cut-offs of >1.15, 1.3, 1.5, and 1.7 for the LA/Ao-ratio as the most frequently used parameter (i.e., cut-offs previously reported [bib_ref] The definition of a haemodynamic significant duct in randomized controlled trials: a..., Zonnenberg [/bib_ref] would result in n = 22, 16, 10 and 5, respectively, of the 27 episodes with at least one single-observermeasurement above the cut-off. In contrast, if only the mean of 2 measurements were considered, LA/Ao would RI (resistance index) in CA (celiac artery) and ACA (anterior cerebral artery), LA/ Ao-ratio (Left-atrium-to-aortic-root-ratio), LVPEP/LVET (left-ventricularpreejection-period-to-ejection-time-ratio), VTI_Ao/VTI_Pa (ratio of the velocity time integrals in the large vessels) and PDA diameter (patent ductus arteriosus); 95 % CI (95 % confidence interval) significantly smaller than RepI of LVPEP/LVET and VTI_Ao/VTI_Pa marked with " * ", "significantly" smaller than RepI of VTI_Ao/VTI_Pa marked with " ** "
have been above the cut-off in 20, 15, 8 and 2 episodes, indicating that in 4-11 % of cases a treatment decision based on LA/Ao-ratio would have been changed by averaging results of only 2 repeated measurements. Before embarking on this study, we assumed that the VTI_Ao/VTI_Pa-ratio might be another easily determined parameter suitable for quantifying ductal left-to-right shunt. Unfortunately, repeatability was similarly poor, presumably because this parameter required measurements in two different views (parasternal short axis and apical 5-chamber view) and VTI_PA was corrupted by the ductal jet . It is also important to note that VTI_Ao/VTI_Pa-ratio may not accurately reflect the degree of shunt through a PDA because of inter-atrial shunting which is commonly observed in VLBW infants just like in our cohort (only 1 out of 19 infants had no inter-atrial shunting, no infant had a ventricular septal defect). This latter limitation also applies to more commonly used parameters such as the LA/Ao ratio. Furthermore, the assumption underlying the determination of VTI_Ao/ VTI_Pa-ratio that the cross-sectional areas of P-and Ao-valve are similar may not applicable to all infants. However, despite poor repeatability, VTI_Ao/VTI_Pa had a high CSA-value, indicating a high potential in identifying inter-individual differences and consequently permitting accurate classification [fig_ref] Table 1: Repeatability Index [/fig_ref]. Similarly, determination of the PDA-diameter was challenging, Measurement of VTI_Pa in a parasternal short axis view. The pulsed-wave Doppler-sonographic measurement of VTI_Pa in a parasternal short axis view is corrupted by ductal jet extending to the pulmonary valve because it requires visualisation of the PDA from the aorta to the pulmonary artery. A limitation of our study is that extremely immature preterm infants with the highest risk of PDA are underrepresented because we were hesitating to subject these most vulnerable infants during their first postnatal days to repeated measurements. Future studies need to assess intra-observer repeatability.
# Conclusions
The repeatability of echocardiographic parameters to evaluate ductal left-to-right shunt is poor. The highest repeatability was achieved by RIs in ACA and CA. This has implications for clinical practice as well as the design of future studies on PDA treatment. In both settings, repeated measurements and averaging of results should be implemented, especially when measured values are close to their cut-off value.
Abbreviations ACA: anterior cerebral artery; CA: celiac artery; CI: confidence interval; CPAP: continuous positive airway pressure; CSA: confidence-step-analysis; LA/Ao-ratio: left-atrium-to-aortic-root-ratio; LVO/SVC-flow: left -ventricular -output-to-superior -vena -cava-flow -ratio; LVPEP/LVET: left-ventricularpreejection-period-to-ejection-time-ratio; PDA: patent ductus arteriosus; RepC: repeatability coefficient; RepI: repeatability index; RI: resistance index; SD: standard deviation; VLBW: very low birth weight; VTI_Ao: velocity time integral ascending Aorta; VTI_Pa: velocity time integral pulmonary artery.
[table] Table 1: Repeatability Index (RepI), Repeatability Coefficient (RepC) and Confidence-Step-Analysis (CSA) values for Echocardiographic Parameters in Preterm Infants with Suspected Patent Ductus Arteriosus [/table]
[table] Table 2: Repeatability Index (RepI) of Echocardiographic Parameters in Paediatric Patients According to the Literature,[18][19][20][21][22] Abbreviations: VTI velocity time integral, PDA patent ductus arteriosus, Ao Diameter Aorta, RI_ACA resistance index in anterior cerebral artery, RVPEP right ventricular preejection period, RVET right ventricular ejection time, w weeks, y years, NR not reported [/table]
|
Correction of Psychoemotional Disorders and Short-Term Prognosis in Patients with COVID-19
This review discusses the importance of the main psychoemotional risk factors for the development of chronic noncommunicable diseases. Current data on the prevalence of anxiety and depressive disorders in patients with cardiovascular disease (CVD) are presented. Data on the relationship between the development of psychoemotional disorders and CVD are summarized and the prospects for managing such patients within the framework of interdisciplinary cooperation are discussed. The main pathogenetic mechanisms for the development of complications, including CNS damage in COVID-19, are considered. The significance of the selection of pathogenetic therapy for patients with comorbid somatic and mental diseases in the context of the COVID-19 pandemic is discussed. Results from multicenter placebo-controlled studies of the use of fl uvoxamine in patients with COVID-19 of varying severity are addressed.
veloping CVD resulting in disability and death [bib_ref] Psychosocial perspectives in cardiovascular disease, Pedersen [/bib_ref] [bib_ref] Socioeconomic status and cardiovascular outcomes: challenges and interventions, Schultz [/bib_ref]. PSRF are known to produce signifi cant reductions in motivation to utilize any therapy; they minimize adherence to a healthy lifestyle, i.e., are associated with malnutrition, low physical activity, alcohol abuse, smoking, etc., which affect patients' health status and quality of life [bib_ref] Psychosocial perspectives in cardiovascular disease, Pedersen [/bib_ref]. Not all PSRF have an equal impact on health status and prognoses; depression and anxiety make the greatest contributions to the development of CNCD. These are the PSRF most commonly encountered in the clinical practice of primary care physicians -neurologists, internists, and cardiologists [bib_ref] Increased 12-month prevalence rates of mental disorders in patients with chronic somatic..., Harter [/bib_ref]. Signifi cant increases in the prevalences of depressive and anxiety disorders, which are among the top three causes of loss of work capacity, has been demonstrated.
Depression and anxiety increase the risk of both CVD and adverse cardiovascular outcomes, including rehospitalization and death, independently of traditional risk factors [bib_ref] Effects of stress on the development and progression of cardiovascular disease, Kivimaki [/bib_ref] [bib_ref] Psychosocial stress and cardiovascular disease, Dar [/bib_ref] [bib_ref] Depression and coronary heart disease: 2018 position paper of the ESC working..., Vaccarino [/bib_ref]. A prospective study (involving about 2 million initially healthy individuals) showed that the appearance of symptoms of depression is associated with an increase in the risk of developing chronic heart failure (CHF) by almost 20%, while the effect of depression persisted even after adjusting for all traditional CVD risk factors [bib_ref] Depression as a risk factor for the initial presentation of twelve cardiac,..., Daskalopoulou [/bib_ref]. Another prospective study of more than 80,000 respondents who are also independent risk factors for the development of coronary artery disease, cardiovascular events, and cardiac death. Although the specifi c mechanisms by which depression and anxiety lead to the development of CVD or worsening of its course have not been established, they have potentially negative impacts on behavioral factors (smoking, excessive alcohol consumption, physical inactivity, malnutrition) and treatment compliance and operate in combination with the adverse effects of stress on the central nervous system (CNS) [bib_ref] Effects of stress on the development and progression of cardiovascular disease, Kivimaki [/bib_ref] [bib_ref] Psychosocial stress and cardiovascular disease, Dar [/bib_ref] [bib_ref] Depression and coronary heart disease: 2018 position paper of the ESC working..., Vaccarino [/bib_ref] [bib_ref] Psychosocial aspects in cardiac rehabilitation: from theory to practice. A position paper..., Pogosova [/bib_ref]. Such relationships are often underestimated by primary care physicians in clinical practice [bib_ref] Psychosocial aspects in cardiac rehabilitation: from theory to practice. A position paper..., Pogosova [/bib_ref]. In the context of the COVID-19 pandemic, which also leads to affective and psychoemotional disorders, the association of depressive and anxiety disorders with somatic diseases becomes more signifi cant [bib_ref] Long-term psychiatric sequelae of acute SARS-COV-2 coronavirus infection, Mosolov [/bib_ref] [bib_ref] Guidelines for the diagnosis and treatment of circulatory diseases in the context..., Shlyakhto [/bib_ref] [bib_ref] Pathogenesis of cognitive and psychoemotional disorders in patients with cardiovascular and metabolic..., Shishkova [/bib_ref].
It should be noted that the frequency and severity of various long-term effects of COVID-19 do not always depend on its initial severity, existing comorbid diseases, or the patient's age, so even mild or asymptomatic COVID-19 may cause a long-term decline in quality of life. So-called post-COVID syndrome (PCS), which is a period of at least 12 weeks of various of the sequelae of COVID-19, determines the further strategy for managing patients. The clinical manifestations of PCS cannot be ignored, given their impact on the quality of life and work capacity of those who have been ill [bib_ref] COVID-19 and stress-related disorders, Kotova [/bib_ref].
The pathogenesis of the main PCS disorders is currently being studied. The virus itself is assumed to have significant damaging effects in the development of complications affecting the respiratory, excretory, cardiovascular, and central nervous systems. SARS-CoV-2 virus has been shown to be able to damage the cells that make up neurovascular units in the CNS, leading to the development of cognitive and psychoemotional disorders [bib_ref] Neurologic features in severe SARS-CoV-2 infection, Helms [/bib_ref]. The processes of systemic infl ammation, as well as the triggering of cytokine storms, are of key importance in the development of most of the complications of infection caused by SARS-CoV-2, including CVD and depressive and anxiety disorders [bib_ref] Cytokine storm induced new onset depression in patients with COVID-19. A new..., Alpert [/bib_ref]. The occurrence of microcirculatory and thrombotic disorders is an important element in the pathogenesis of many initially did not have CVD found that the appearance of depressive symptoms increased the risk of developing CHF by 21% over the next six years [bib_ref] Depression and human immunodefi ciency virus infection are risk factors for incident..., White [/bib_ref]. An increase in depressive symptoms or the establishment of a diagnosis of depression in patients with atrial fi brillation and CHF was associated with increases in the numbers of hospitalizations and recurrent cardiovascular events, as well as mortality [bib_ref] Elevated depression symptoms predict long-term cardiovascular mortality in patients with atrial fi..., Frasure-Smith [/bib_ref]. A meta-analysis of a series of studies investigating the association between depression and CHF outcomes showed that the presence of depressive symptoms or the diagnosis of depression led to a two-fold increase in the risk of death or a non-fatal cardiovascular event [bib_ref] Depression in heart failure a meta-analytic review of prevalence, intervention and associations..., Rutledge [/bib_ref]. It should be emphasized that CVD is most commonly associated with depression both in clinical studies and in real medical practice [bib_ref] Depression as a risk factor for the initial presentation of twelve cardiac,..., Daskalopoulou [/bib_ref]. One patient in fi ve with ischemic heart disease (IHD) or CHF suffers from depression; patients with CVD and concomitant depression have increased risks of recurrent cardiovascular events and death [bib_ref] Psychosocial aspects in cardiac rehabilitation: from theory to practice. A position paper..., Pogosova [/bib_ref]. Clinical observations indicate that depression is more likely to develop in men who have had a stroke and in women with more than three chronic diseases [bib_ref] Distribution of anxiety and depression in different regions of the Russian Federation..., Shal'nova [/bib_ref].
The development of depression has been shown to produce a near doubling of the risk of IHD, a 1.4-fold increase in the risk of stroke, and a 1.2-fold increase in overall mortality [bib_ref] Distribution of anxiety and depression in different regions of the Russian Federation..., Shal'nova [/bib_ref] [bib_ref] Depression as an aetiologic and prognostic factor in coronary heart disease: a..., Nicholson [/bib_ref]. The ESSE-RF study shows that women with subclinical and clinical depression have an almost 2.4-fold increase in mortality, while men with subclinical and clinical depression have a 1.5-fold increase as compared with people without symptoms of depression. Depression increases the risk of death from CVD by a factor of 1.6 and the risk of death from all causes by a factor of 1.8, especially in elderly patients [bib_ref] Observational study of the differential impact of time-varying depressive symptoms on all-cause..., Moise [/bib_ref].
Taking data from the ESSE-RF study on the prevalence of anxiety states into account, it is clear that these are common in the population, at a rate of more than 18% [bib_ref] Distribution of anxiety and depression in different regions of the Russian Federation..., Shal'nova [/bib_ref]. It should be noted that anxiety disorders develop during life in more than a quarter of the population and that symptoms of severe anxiety are detected in 40% of patients seen by primary care physicians, i.e., general practitioners, neurologists, and cardiologists [bib_ref] Psychosocial aspects in cardiac rehabilitation: from theory to practice. A position paper..., Pogosova [/bib_ref]. Anxiety states, like depression, We will consider the anti-infl ammatory effects of fl uvoxamine separately; these are mediated by activation of sigma-1 receptors, which are chaperone proteins able to migrate within cells. Sigma-1 receptors are present in various tissues and are located mainly on the endoplasmic reticulum membrane, which is involved in lipid synthesis, calcium homeostasis, energy metabolism, autophagy, and apoptosis [bib_ref] Sigmar1's molecular, cellular, and biological functions in regulating cellular pathophysiology, Aishwarya [/bib_ref]. Experimental studies have shown that mice with inactivation of sigma-1 receptors had higher mortality from septic shock and higher levels of proinfl ammatory cytokines (interleukin-6 and tumor necrosis factor α) [fig_ref] Figure 3: Effects of fl uvoxamine on survival in mice with septic shock [/fig_ref]. The levels of these cytokines correlate with poor survival prognosis in patients with sepsis. Decreases in cytokine contents increased survival in animals, while administration of fl uvoxamine promoted survival in mice with preserved sigma-1 receptors, though administration to mice with inactivated sigma-1 receptors did not affect survival. The authors concluded that the anti-infl ammatory effect of fl uvoxamine is mediated by activation of sigma-1 receptors. It should be noted that the survival rate of mice with experimental septic shock and given fl uvoxamine was comparable to the survival rate on the background of treatment with ceftriaxone; when the two drugs were combined, their positive effects were summed. The data obtained in these experiments were reproduced in human blood samples in vitro, when addition of fl uvoxamine contributed to a manifold decrease in the level of proinfl ammatory cytokines, which had been increased by prior addition of bacterial lipopolysaccharide [bib_ref] Modulation of the sigma-1 receptor-IRE1 pathway is benefi cial in preclinical models..., Rosen [/bib_ref]. Thus, the greater affi nity of fl uvoxamine for sigma-1 receptors than other antidepressants may provide an indirect indication of its possible potential in the treatment of patients with COVID-19 [bib_ref] Mechanisms of action of fl uvoxamine for COVID-19: a historical review, Hashimoto [/bib_ref] [bib_ref] Fluvoxamine for the treatment of COVID-19, Boulware [/bib_ref].
Effects of Fluvoxamine on Patient Survival during the COVID-19 Pandemic. Results of studies using fl uvoxamine in individuals infected with the SARS-CoV-2 virus are very interesting and important from the practical standpoint. Several randomized, double-blind, placebo-controlled studies have demonstrated the ability of fl uvoxamine complications, including cardiovascular and cerebrovascular complications [bib_ref] Endothelial cell infection and endotheliitis in COVID-19, Varga [/bib_ref] [bib_ref] Understanding the neurotropic characteristics of SARSCoV-2: From neurological manifestations of COVID-19 to..., Zhou [/bib_ref]. Also important are correct assessment of the impact of side effects of treatments provided for COVID-19 and the consequences of prolonged immobilization, the stress factor of hospitalization and, of course, social isolation [bib_ref] Central nervous system damage in COVID-19, Kurushina [/bib_ref]. Thus, the COVID-19 pandemic is characterized by additional impact on both the occurrence and exacerbation of pre-existing PSRF in many patients.
The Sigma-1 Receptor-Associated Anti-Infl ammatory Effect of Fluvoxamine. Establishment of the mechanisms of action of the SARS-CoV-2 virus on the human body has led to studies of new properties of drugs that have not previously been used in viral infections, i.e., drugs of the antimalarial, antipsychotic, antihistamine, and antidepressant classes. Among the latter, fl uvoxamine (Rokona) stands out; this has a wide range of positive effects, not only improving patient's psychoemotional state, but also resisting SARS-CoV-2 virus infection [fig_ref] Figure 2: Positive effects of fl uvoxamine in COVID-19 [/fig_ref]. Fluvoxamine has been shown on the one hand to have an antiviral effect, disrupting the intracellular transport of viruses and preventing their exit from cells, while on the other hand it has significant anti-infl ammatory effects mediated by sigma-1 receptor activation and a decrease in histamine release by mast cells; it weakens the negative consequences of systemic infl ammation and reduces the severity of cytokine storm. Fluvoxamine also produces signifi cant increases in melatonin levels which, in addition to the regulatory effect on the sleep-wake cycle, has anti-infl ammatory effects. Given that serotonin is directly involved in the pathogenesis of microcirculatory disorders, the direct effect of fl uvoxamine -it decreases serotonin reuptake by platelets -is associated with an antithrombocyte action, and this is an additional advantage of its use during the COVID-19 pandemic [bib_ref] Fluvoxamine: a review of its mechanism of action and its role in..., Sukhatme [/bib_ref] [bib_ref] Do the serotonin reuptake inhibitor antidepressants fl uoxetine and fl uvoxamine reduce..., Hoertel [/bib_ref] [bib_ref] The acid sphingomyelinase/ceramide system in COVID-19, Kornhuber [/bib_ref]. These effects of fl uvoxamine appear not to apply to the entire class of antidepressants, as evidenced by results from a meta-analysis demonstrating an increased risk of death in COVID-19 patients associated with taking antipsychotics and antidepressants [bib_ref] Mental disorders and risk of COVID-19-related mortality, hospitalisation, and intensive care unit..., Vai [/bib_ref]. addition to basic therapy led to a signifi cant reduction in mortality in hospitalized patients with severe COVID-19. These results led to the inclusion of fl uvoxamine in clinical guidelines and protocols for the treatment of patients with COVID-19 in many countries, including Russia [bib_ref] Prevention, Diagnosis, and Treatment of the New Coronavirus Infection (COVID-19). Version 15..., Avdeev [/bib_ref].
Conclusions. The impact of the COVID-19 pandemic in terms of an increase in the incidence of anxiety and depressive disorders in the population, including in the longterm period after the disease, affecting both prognosis and patients' quality of life, should be emphasized. Thus, the whole set of unfavorable aspects associated with the mutually potentiating effects of somatic diseases, particularly CVD, and mental disorders, in combination with the impact of the COVID-19 pandemic, is of particular importance today. We can be confi dent that further studies of the problem of managing comorbid patients on the backdrop of the COVID-19 pandemic will, in the future, reduce the number of complications and improve patients' quality of life. This will largely depend on the selection of therapies effectively able to infl uence the main manifestations of COVID-19 and provide good amelioration of mental or somatic pathologies. Fluvoxamine can be regarded as a promising drug in this direction, as it has not only high effi cacy against anxiety and depressive symptoms and a good safety profi le, including in patients with comorbid diseases, but also immunomodulatory and antiviral activity against the SARS-CoV-2 virus, as well as having positive effects on the prognosis of the progression of coronavirus infection -reducing the risk of hospitalization and death.
The authors declare no confl ict of interest.
to reduce the risks of hospitalization and mortality in patients with COVID-19 [bib_ref] Fluvoxamine vs placebo and clinical deterioration in outpatients with symptomatic COVID-19: a..., Lenze [/bib_ref] [bib_ref] Prospective cohort of fl uvoxamine for early treatment of coronavirus disease 19...., Seftel [/bib_ref] [bib_ref] Effect of early treatment with fl uvoxamine on risk of emergency care..., Reis [/bib_ref] [bib_ref] Safety and effi cacy of fl uvoxamine in COVID-19 ICU patients: an..., Calusic [/bib_ref]. Of all the studies conducted to date addressing the effects of fl uvoxamine on the course of COVID-19, the multicenter, randomized, double-blind, placebo-controlled TOGETHER study, the largest in terms of the number of participants, should be singled out [bib_ref] Effect of early treatment with fl uvoxamine on risk of emergency care..., Reis [/bib_ref]. This trial was conducted on unvaccinated outpatients with confi rmed COVID-19 who had additional risk factors for the development of severe complications (CVD, CerVD, DM2, obesity, respiratory system diseases, immunodeficiency, cancer, chronic kidney disease, etc.). Disease duration at randomization was no greater than seven days; all patients received symptomatic therapy, and, depending on randomization, fl uvoxamine 100 mg b.i.d. for 10 days or placebo. All side effects and unwanted effects, including those possibly associated with taking placebo or fl uvoxamine, were thoroughly documented. The results exceeded expectations: for example, taking fl uvoxamine at a dose of 200 mg/day for at least eight days prevented worsening of patients' condition and the progression of the disease with subsequent hospitalization in 99.8% of cases. The safety analysis demonstrated equal overall numbers of adverse events for placebo and fl uvoxamine, though the number of serious adverse events was signifi cantly higher in the placebo group. As regards the safety of fl uvoxamine in patients, including the elderly, with COVID-19 and various comorbidities, other authors have also reported similar conclusions, emphasizing the lack of any effect of fl uvoxamine on the QT interval, whereby it differs from many other antidepressants [bib_ref] Fluvoxamine vs placebo and clinical deterioration in outpatients with symptomatic COVID-19: a..., Lenze [/bib_ref] [bib_ref] Prospective cohort of fl uvoxamine for early treatment of coronavirus disease 19...., Seftel [/bib_ref]. Another important step in studying the effi cacy and safety of fl uvoxamine treatment was a case-control study in patients with severe COVID-19 treated in intensive care units [bib_ref] Safety and effi cacy of fl uvoxamine in COVID-19 ICU patients: an..., Calusic [/bib_ref]. This showed that fl uvoxamine at a dose of 300 mg/day (100 mg t.i.d.) for 15 days in
[fig] Fig 1: Main PSRF for chronic noncommunicable diseases. [/fig]
[fig] Figure 2: Positive effects of fl uvoxamine in COVID-19. [/fig]
[fig] Figure 3: Effects of fl uvoxamine on survival in mice with septic shock. [/fig]
|
International estimates of intended uptake and refusal of COVID-19 vaccines: A rapid systematic review and meta-analysis of large nationally representative samples
a b s t r a c tBackground: Widespread uptake of COVID-19 vaccines will be essential to controlling the COVID-19 pandemic. Vaccines have been developed in unprecedented time and quantifying levels of hesitancy towards vaccination among the general population is of importance. Methods: Systematic review and meta-analysis of studies using large nationally representative samples (n ! 1000) to examine the percentage of the population intending to vaccinate, unsure, or intending to refuse a COVID-19 vaccine when available. Generic inverse meta-analysis and meta-regression were used to pool estimates and examine time trends. PubMed, Scopus and pre-printer servers were searched from January-November 2020. Registered on PROSPERO (CRD42020223132). Findings: Twenty-eight nationally representative samples (n = 58,656) from 13 countries indicate that as the pandemic has progressed, the percentage of people intending to vaccinate decreased and the percentage of people intending to refuse vaccination increased. Pooled data from surveys conducted during June-October suggest that 60% (95% CI: 49% to 69%) intend to vaccinate and 20% (95% CI: 13% to 29%) intend to refuse vaccination, although intentions vary substantially between samples and countries (I 2 > 90%). Being female, younger, of lower income or education level and belonging to an ethnic minority group were consistently associated with being less likely to intend to vaccinate. Findings were consistent across higher vs. lower quality studies. Interpretation: Intentions to be vaccinated when a COVID-19 vaccine becomes available have been declining across countries and there is an urgent need to address social inequalities in vaccine hesitancy and promote widespread uptake of vaccines as they become available. Funding: N/A.
# Introduction
The COVID-19 pandemic has resulted in more than 1 million deaths worldwide from March to October 2020 and is likely to continue to have far reaching impacts on healthcare systems. The development of vaccines against COVID-19 has been occurring at unprecedented speed, and as of November 2020, there are multiple candidate vaccines in the final stages of testing . The success of any vaccination programme is dependent on the proportion of the population willing to be vaccinated and based on recent estimates it is likely that up to three quarters of the population may require vaccination to bring an end to the pandemic.
Early in the pandemic a small number of studies surveyed adults to gauge public willingness to be vaccinated against COVID-19 and although a number of studies were reliant on nonrepresentative convenience samples, the majority of the populations sampled intended to vaccinate. For example, a cross country survey found a relatively high average level of intended vaccination (72%), although sample sizes of individual countries were low (N =~600-800) and may not provide accurate nationally representative coverage.
However, as the pandemic has evolved there have been reports of widespread misinformation about COVID-19, distrust in governmentand public concerns about the safety of COVID-19 vaccines given their rapid development, all of which may have affected intended vaccine uptake. It is also unclear whether vaccine acceptability will be socio-demographically patterned. Disadvantaged minority groups have previously been shown to be less likely to intend to be vaccinated for influenza. However, a systematic review concluded that patterning for other demographic groups during previous influenza vaccination programmes is inconsistent. Given that current evidence on socio-demographic patterning of COVID-19 vaccination intentions is lacking, it will be important to understand how vaccination intentions differ within and across countries to inform measures to improve public acceptability and uptake of vaccination programmes.
We conducted a rapid systematic review and meta-analysis of large nationally representative study samples to address the current lack of consensus on; i) the proportion of the population willing to be vaccinated against COVID-19 when a vaccine become available and how this differs across countries, ii) whether vaccination intentions have declined as the pandemic has progressed, iii) socio-demographic inequalities in intended vaccine uptake.
# Method
To inform mass COVID-19 vaccination programmes, we used rapid systematic review methodology. Rapid reviews provide timely evidence synthesises whilst maintaining the rigour of traditional systematic reviewsby using expedited review processes, such as limiting databases searched or number of reviewers (e.g. cross-checking of a proportion of extraction as opposed to independent extraction by a second author).
Eligibility criteria. We included studies that measured intentions to be vaccinated against coronavirus in nationally representative samples of the general public. Published journal articles and pre-prints were eligible for inclusion. News articles reporting on opinion polls (with no corresponding scientific report including methodology used) were not eligible. The review is registered on PROSPERO (CRD42020223132).
Populations and study design: To be eligible studies were required to have used a sampling approach designed to be nationally representative on key population demographics of the country (e.g. gender, education level), such as quota sampling or random probability sampling. Sampling error is problematic when examining prevalence estimates in small sample sizes. Because sampling error tends to be minimal at sample sizes of ! 1000, as is considered common practice for nationally representative surveys, we limited eligibility to studies with a sample of n ! 1000 from the same country. Studies that collected non-general public samples (e.g. healthcare professionals, parents, students) were not eligible. Studies that used non-representative sampling (e.g. convenience sampling, snowball sampling) were not eligible and studies that lacked sufficient methodological information to determine sampling used were ineligible.
Interest (measure): To be eligible studies were required to include a question that measured intentions/willingness to use a vaccine for COVID-19 when one becomes available (e.g. 'I would use a vaccine for COVID-19 when it becomes available'). Studies that exposed participants to information designed to alter vaccine intentions (e.g. experiments comparing how public health messages may improve vaccine intentions) were not eligible. Studies that compared willingness to take different types of vaccine (e.g. varying hypothetical effectiveness/price) were ineligible, unless they also included a questionnaire item measuring general willingness to take a COVID-19 vaccine.
Outcome: Studies were required to report the proportion of participants responding to the different response options for the vaccine intention question (e.g. Yes vs. No, Willing vs. Unsure Vs. Unwilling, Likely vs. Undecided vs. Unlikely).
Article identification strategy. During November 2020, we searched PUBMED and Scopus (2020 onwards) for published articles in peer reviewed journals. We used the following search terms: (COVID OR coronavirus OR SARS-COV-2) AND (Vaccine OR Vaccination) AND (Inten* OR willing* OR attitud* OR hypothetical). We also searched three pre-print servers; Open Science Framework (which hosts 30 other preprint archives, including PsychArxiv), MedrXiv and the Social Science Research Network (SSRN). One author conducted the initial title and abstract screening to exclude unrelated articles and a second author checked 25% of this (there were no discrepancies). A single author conducted full-text screening to determine eligibility and a second author cross-checked all eligibility decisions. For all eligible articles identified through searches, we used forward citation tracking (Google Scholar) and our knowledge of existing research to identify any further articles.
Data extraction. For each study one author extracted information, all extraction was cross-checked by a second author and disagreements were resolved through discussion. We extracted the following information; bibliographic information, country, sampling procedure (e.g. quota vs. probability), sample size, month of survey, measure of vaccination intentions and results. We extracted results based on response options, resulting in % choosing response options indicative of definitely or probably yes, % definitive or probable no, % unsure (if measure allowed for the latter). If studies had multiple waves of data collection, we extracted results from the most recent wave. We also extracted information on whether studies reported results of analyses examining demographic predictors of vaccination intentions that were commonly examined (i.e. at least 5 studies) in studies. To be eligible for extraction, demographic predictors were required to be examined adjusting for other demographics (i.e. zero-order correlations were not eligible) in order to be confident of independent effects. We prioritised extraction of results of analyses that examined vaccine intentions (i.e. reference of category of yes vs. other (unsure/no combined). If this analysis was not available, in order of priority we favoured extracting demographic results for analyses examining yes vs. no, then yes vs. unsure.
Risk of bias indicators. We considered methodological factors of studies that may impact on results by increasing risk of bias.
Quota-based samples were rated as being higher in risk of bias than probability-based samples, as the latter tends to be a more accurate/representative sampling approach. We also reasoned that studies not having yet undergone formal peer review (unpublished pre-prints) may increase risk of bias and coded for this. For studies examining demographic predictors, we considered relatively smaller sample sizes (n < 2500) as higher in risk of bias due to concerns over small numbers of cases in analyses when examining sub-groups. For example, with n ! 2500, for minority sub-groups (e.g.~2% of the population, such as Black people in the UK) there would be expected to be a minimum of 50 cases in analyses, as opposed to only 20 cases with n = 1000. Studies of demographic predictors that adjusted for attitudinal predictors (e.g. attitudes towards vaccination) of vaccine intentions were considered higher in risk of bias, as inclusion may mask associations between demographics and vaccination intentions. For example, if demographic patterning of vaccination intentions is caused in part by attitudinal differences, including attitudinal measures as predictor variables of intentions (alongside demographic factors) would reduce and 'mask' an association between a demographic factor and vaccination intentions.
Synthesis of evidence: We meta-analysed proportions of the samples ((Total sample N / 100) * % reporting) reporting: i) intending to vaccinate, ii) unsure, and iii) not intending to vaccinate. Analysis was performed using the 'metafor' package in R. We used a logit transformation on raw proportions in random effects, generic inverse variance meta-analyses with a restricted maximumlikelihood estimator. Transformations were conducted using the 'escalc' function in the metafor package. Back-transformed (inverse) logit values are presented in the text, whilst raw proportion data is presented in forest and funnel plots to aid interpretation. Heterogeneity was assessed with the I 2 statistic. We conducted meta-regressions to examine the relationship between outcomes and month of study data collection (treating month as a continuous variable, e.g. March = 1, April = 2). We conducted leave-oneout analysis to examine the stability of the pooled estimates and identify any influential samples in the main analyses and metaregressions. We reasoned that studies which did not use an 'unsure' or 'undecided' response option (e.g. responses grouped into yes vs. no, as opposed to yes vs. unsure vs. no) may affect vaccine acceptance and rejection estimates, so we examined if pooled estimates of intended vaccination and intended refusal differed between these two types of study. Furthermore, in sensitivity analyses we examined whether the inclusion of an 'unsure/undecided' response option moderated any effects of month of study data collection on intention estimates. We also examined the effect of month of study data collection separately in studies with vs. without an 'unsure/undecided' response option.
To address risk of bias, we conducted sub-group analyses to examine whether results differed between studies using quota vs. probability sampling and pre-prints vs. journal articles. We limited these analyses to studies which collected data early in the pandemic (March-May) to account for the majority of probability samples (3/4) and journal article sample (12/12) studies being conducted in this period. Sub-group analysis compared studies that did vs. did not allow for an 'unsure/undecided' response option. We also examined potential publication bias and small study effects (see online supplementary materials). Demographic predictors were measured and/or analysed differently across studies, so for each demographic we reported the proportion of studies finding evidence vs. no evidence of significant (p < .05) relationships with vaccination intentions, and whether results were similar when limited to studies that had larger sample sizes and did not adjust for attitudinal predictors in analyses (i.e. higher quality studies).
# Results
Study selection. A total of 792 unique articles were found through database searches and other sources. Of the 145 articles full-text screened, 20 articles reporting on 28 samples were eligible for inclusion. See. For bibliographic information of all included articles, see online supplementary material.
Overview. Of the 28 samples included, the majority were from the UK and North America (7 = US, 2 = Canada, 7 = UK) as well as samples from France (n = 3), Australia, China, Denmark, Germany, Italy, Ireland Netherlands, Poland and Portugal. Full study information is reported in. Sample sizes ranged from 1000 to 7547 (median = 1198). Samples were collected in the early phases of the pandemic (March-May, n = 18) and later (June onwards, n = 10). Twenty-four samples used quota-based sampling and four used probability sampling. Twelve articles were in peer-reviewed journals and 16 were pre-prints (at the time of identification). Sixteen studies included an 'unsure/undecided' response option for vaccination intentions and 12 studies did not include this response option.
Intentions Presence vs. absence of 'unsure' response option in survey. There was a significant difference in the proportion of participants intending to vaccinate when an 'unsure/undecided' response option was used vs. when there was no 'unsure/undecided' response option (X 2 (1) = 16.82, p < .001). When there was no unsure response option the proportion was 0.828 [95% CI: 0.759 to 0.880: I 2 = 99.4%], and when unsure was a response option the proportion was 0.635 [95% CI: 0.569 to 0.670: I 2 = 99.2%]. There was not a significant difference in the proportion intending to refuse a vaccine between samples with [0.124: 95% CI: 0.092 to 0.164: I2 = 99.0%] vs. without [0.172: 95% CI: 0.120 to 0.240: I 2 = 99.4%] an 'unsure/undecided' response option (X 2 (1) = 2.12, p = .146).
Time trends. There was a significant association between proportion of individuals intending to vaccinate and month study was conducted (coefficient = -0.25 [95% CI: -0.37 to -0.12], z = 3.83, p < .001), see. There was also a significant association between intentions not to vaccinate and month of study (coefficient = 0.16 [95% CI: 0.03 to 0.28], z = 2.48, p = .013). Over time intentions to vaccinate decreased, and intentions not to vaccinate increased. For example, in studies (n = 18) that collected data during the early phase of the pandemic (March-May), the proportion intending to vaccinate was 79% and not intending to was 12%, compared with 60% and 20% June-October studies (n = 10). There was no association between proportion of individuals being unsure about vaccination and when the study was conducted (coefficient = 0.10 [95% CI: À0.03 to 0.24], z = 1.54, p = .124]. Because samples varied by country, to confirm time trend findings were not explained by different countries being sampled earlier vs. later in the pandemic, we replicated the meta-regressions for the two countries with multiple samples collected during different months. Among UK samples (n = 7, coefficient = À0.39 [95% CI: À0.57 to À0.22], z = 4.34, p < .001) and US samples (n = 7, coefficient = À0 .22 [95% CI: À0.38 to À0.05], z = 2.53, p = .011)] the same negative associations were observed as in the main analyses. There was no evidence that the relationship between month of survey data col-lection and intending to be vaccinated was moderated by whether a study did vs. did not have an 'unsure/undecided' response options. See online supplementary materials.
Publication and risk of bias analyses. There was minimal evidence of publication bias and leave out one analyses indicated limited variation in estimates (online supplementary materials document). We found no evidence that results differed between samples reported in journal articles vs. pre-prints. There was minimal evidence of differences in findings between studies using probability vs. quota sampling, with the exception of 'unsure' responses being lower in quota samples (see online supplementary materials).
Demographic predictors. Fourteen studies examined demographic predictors. See. In 12/14 studies older adults were significantly more likely to report intending to vaccinate than younger, one study found no effect of age and in one study young adults (<25 years) were more likely than middle aged adults, but not older adults. In 9/14 studies males were more likely to intend to vaccinate than females (no significant association in five studies). Higher education level was associated with intending to vaccinate in 7/14 studies (no association in seven). White ethnic groups were more likely to vaccinate in 7/11 studies (no association in four). Higher income was associated with being more likely to intend to vaccinate in 8/9 studies and in one study there was no association. Presence of a health condition was examined in five studies. Presence of a health condition was a non-significant predictor in n = 4 and not having a health condition was associated with being more likely to intend to be vaccinated in one study. When analyses were limited to the five higher quality studies (large sample size, no inclusion of attitudinal predictors) the role of demographic factors was more consistent; 5/5 studies found that older adults, males and higher education levels were associated with increased likelihood of intending to vaccinate. Similarly, 3/4 studies found that being white and 4/4 found that those on higher income were more likely to intend to vaccinate. Presence of a health condition was non-significant in n = 2.
# Discussion
Results of this systematic review and meta-analysis of 58,656 participants drawn from 28 large nationally representative study samples across 13 countries indicates that the percentage of the population intending to be vaccinated when a COVID-19 vaccine becomes available has declined markedly across countries as the pandemic has progressed. Numbers reporting that they will refuse a vaccine have increased over time and a substantial proportion of adults now intend to refuse a vaccine, when available (June-October estimate = 20%). There is also consistent sociodemographic patterning of vaccination intentions; being female, younger, of lower income or education level and belonging to an ethnic minority group are associated with a reduced likelihood of intending to be vaccinated when a vaccine become available. Some of the present findings are in line with a recently published narrative review of studies examining vaccine intentions by Lin et al. The authors concluded that COVID-19 vaccination intentions have declined over time and there was socioeconomic and demographic patterning of vaccine intentions. Lin et al.included a number of small non-representative samples and results from news reports in their review. These methodological differences may explain why in the present research of large nationally representative studies we found more consistent evidence of gender differences.
Emerging evidence suggests that both exposure to misinformation about COVID-19and public concerns over the safety of vaccinesmay be contributing to the observed declines in intentions to be vaccinated, and this highlights the need for measures to address public acceptability, trust and concern over the safety and benefit of approved vaccines. As well as observing declines in intentions to vaccinate over time in our main analyses, we also note intentions differed markedly by country. When sampled during a similar period early on in the pandemic (March-April), 91% of adults in China reported intending to be vaccinated, compared to 76% of adults in France. However, the most recent estimate (September-October) in France is now 52% and similar to the US (54%).
Across studies males were more likely to report intending to vaccinate than females and this is consistent with previous research examining influenza vaccine hesitancy and some studies identifying higher uptake among males. However, there is also recent evidence of greater influenza vaccine uptake among females in both the US and Hong Kong. A range of social and contextual factors may explain gender patterning of vaccination intentions (e.g. current or planned pregnancy). It will therefore be important to understand why females show lower intentions to vaccinate against COVID-19 and whether uptake shows a similar gender patterning. There was also very consistent socioeconomic patterning of vaccination intentions: lower income or education and ethnic minorities were less likely to intend to vaccinate. There was no evidence in any studies reviewed that presence of a chronic health condition was associated with increased vaccination intentions, even though these groups are at increased risk of dying from COVID-19. Measures are required to maximise vaccine uptake in vulnerable and disadvantaged groups who have already been disproportionately affected by the pandemic, such as those from lower income and ethnic minority groups.
# Strengths and limitations
Strengths of the present research are that we limited evidence synthesis to study designs that allow for accurate estimates of population level intentions with minimal sampling error, as small studies of non-representative samples are likely to provide biased estimates. A large proportion of the included studies used quota (as opposed to probability-based sampling) and were pre-prints yet to be peer reviewed (as opposed to published journal articles). However, the type of sampling method used (quota vs. probability) had minimal impact on intentions estimates and that studies reported in pre-prints produced similar effect estimates as peer-reviewed journals. Analyses also accounted for studies using different response formats and findings were similar. However, studies that included an 'unsure/undecided' response option had lower estimates of the proportion of the sampled population intending to vaccinate compared to studies with only two response options (e.g. 'yes' vs. no'), but there was no difference for the proportion of sampled populations indicating they would not vaccinate. These findings indicate that studies which do not include an 'unsure' response option overestimate vaccination intentions and that 'unsure' response options may be primarily driven by participants Older adults more likely to vaccinate (55yrs + ) who are unsure but more likely to intend to be vaccinated than not get vaccinated. We used a rapid systematic review approach and this resulted in us surveying only two online database for published articles and rather than two authors independently assessing article eligibility and extraction, one author was responsible with crosschecking by a second author. However, we retained a number of features of best practice for systematic review and meta-analysis (e.g. assessing risk of bias, publication bias and thorough searching of grey literature). There were fewer studies later (as opposed to earlier) in the pandemic and this may affect the reliability of the time trends analyses. However, we found evidence of declining vaccination intentions when data was analysed using metaregression (continuous month-of-year variable), when comparing study estimates from early in the pandemic (March-May) to later (June-October) and when examining time trends within countries (UK, US). Furthermore, results of the time trends analyses were consistent irrespective of whether studies used an 'unsure/undeci ded' response option or not. Two included individual studies (US, France) also reported vaccination intentions over time in the same population and results were consistent. As included studies examined nationally representative samples we do not know whether a similar pattern of results would be expected among other population groups (e.g. healthcare workers) and research is required to address these questions. Findings of the present research are limited to studies that met eligibility criteria for having used large and nationally representative sampling and this tended to be developed western countries.
# Conclusions
Intentions to vaccinate when a COVID-19 vaccine becomes available have been declining across countries and there is an urgent need to address social inequalities in vaccine hesitancy and promote widespread uptake of vaccines as they become available.
## Data sharing
Study data files and analysis code are openly available on the Open Science Framework; https://osf.io/hj4ds/.
## Contributors
The study was designed by ER. Data was collected by ER, AJ, IL and MD. AJ and ER analysed the data. The first draft was written by ER. All authors critically revised the manuscript and agree to be accountable for all aspects of the work.
## Role of the funding source
There was no funding source for this study.
## 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.
|
Do health checks for adults with intellectual disability reduce emergency hospital admissions? Evaluation of a natural experiment
Background Annual health checks for adults with intellectual disability (ID) have been incentivised by National Health Service (NHS) England since 2009, but it is unclear what impact they have had on important health outcomes such as emergency hospitalisation. Methods An evaluation of a 'natural experiment', incorporating practice and individual-level designs, to assess the effectiveness of health checks for adults with ID in reducing emergency hospital admissions using a large English primary care database. For practices, changes in admission rates for adults with ID between 2009-2010 and 2011-2012 were compared in 126 fully participating versus 68 non-participating practices. For individuals, changes in admission rates before and after the first health check for 7487 adults with ID were compared with 46 408 age-sex-practice matched controls. Incident rate ratios (IRRs) comparing changes in admission rates are presented for: all emergency, preventable emergency (for ambulatory care sensitive conditions (ACSCs)) and elective emergency. Results Practices with high health check participation showed no change in emergency admission rate among patients with ID over time compared with nonparticipating practices (IRR=0.97, 95% CI 0.78 to 1.19), but emergency admissions for ACSCs did fall (IRR=0.74, 0.58 to 0.95). Among individuals with ID, health checks had no effect on overall emergency admissions compared with controls (IRR=0.96, 0.87 to 1.07), although there was a relative reduction in emergency admissions for ACSCs (IRR=0.82, 0.69 to 0.99). Elective admissions showed no change with health checks in either analysis. Conclusions Annual health checks in primary care for adults with ID did not alter overall emergency admissions, but they appeared influential in reducing preventable emergency admissions.
# Introduction
Adults with intellectual disability (ID) experience high levels of morbidity, hospitalisation and premature mortality. [bib_ref] Mortality in adults with moderate to profound intellectual disability: a population-based study, Tyrer [/bib_ref] In response to recommendations from the Disability Rights Commission, 2 in 2009 the English National Health Service (NHS) introduced an annual health check scheme as a Directed Enhanced Service (DES) in primary care for adults with ID.This was intended to identify undetected health problems and improve prescribing and coordination with secondary care. Systematic reviews on the effectiveness of health checks in people with ID have confirmed that they are effective in identifying new health problems, improving uptake of preventive interventions and improving indicators of process of care. [bib_ref] The impact of health checks for people with intellectual disabilities: an updated..., Robertson [/bib_ref] However, there is little evidence on their effectiveness in modifying outcomes such as hospitalisation, [bib_ref] Systematic reviews of the health or health care of people with intellectual..., Robertson [/bib_ref] which is important for patients, carers and the health services. With only half of the eligible adults receiving health checks by 2011-2012,this provided an opportunity to evaluate the scheme by viewing it as a 'natural experiment'.
In this paper, we use a robust observational methodology, using practice-level and individuallevel designs, to assess whether the introduction of health checks in 2009 reduced emergency hospitalisation for adults with ID. We first compare high with low uptake practices, evaluating change in admission rates for all adults with ID, controlling for underlying differences between practices. However, the possibility remains that participating practices improved the care of their patients with ID independent of introducing health checks. Therefore, we also present a matched cohort study comparing change in admission rates of individuals with ID who had health checks to that seen for a matched group of patients without ID, controlling for secular trends in practice care or hospital admissions. Finally, a second matched cohort study for individuals with ID not receiving health checks is then used to confirm the specificity of findings to those having a health check only.
# Methods
## Data source
The Clinical Practice Research Datalink (CPRD) is a large primary care database representative of the UK population. [bib_ref] Data resource profile: Clinical Practice Research Datalink (CPRD), Herrett [/bib_ref] We included 343 practices in England recording data on 1 January 2009, anonymously linked to Hospital Episodes Statistics (HES) data. HES records clinical and administrative information on all NHS-funded inpatient episodes, and allows for identification on method of admission (eg, emergency), in addition to the primary reason for the admission.
## Identification of patients with id and their health checks
We have previously detailed our methodology for identifying adults (aged 18-84) with ID in CPRD in England. [bib_ref] Health characteristics and consultation patterns of people with intellectual disability: a cross-sectional..., Carey [/bib_ref] Briefly, we included all codes used by the Quality and Outcome Framework (QOF) for learning disability, 9 plus additional codes for conditions usually associated with ID such as chromosomal and metabolic disorders (see online supplementary e-table 1). Health checks were identified by specific Read codes used by practices to facilitate future payment. We only included health checks from 1 April 2009, the point from when practices received remuneration for carrying them out.
We classified patients with ID with high levels of support needs based on either a record of severe or profound ID or, where no record of severity was available (59%), at least two of the following: cerebral palsy/significant mobility problem, severe visual impairment, severe hearing impairment, epilepsy (excluding absence seizures), a continence problem, or use of percutaneous endoscopic gastrostomy feeding [fig_ref] Table 2: Hospital admissions in 2011-2012 vs 2009-2010 by practice level of participation in... [/fig_ref]. Patients with ID were estimated to be living in a communal setting by specific Read codes (see online supplementary e-table 3), or the presence of three or more people with ID with the same address flag.
## Hospital admission outcomes
Our main outcome was a count of emergency hospital admissions, defined as distinct periods of care on the HES record. We were also interested in emergency admissions for ambulatory care sensitive conditions (ACSCs), [bib_ref] Is secondary preventive care improving? Observational study of 10-year trends in emergency..., Bardsley [/bib_ref] which are thought to be potentially preventable with better clinical management. We included 20 widely used ACSCs, adding three further conditions (constipation, aspiration, gastro-oesophageal reflux disease) which are more relevant reasons for admission among adults with ID. [bib_ref] Hospitalisation rates for ambulatory care sensitive conditions for persons with and without..., Balogh [/bib_ref] We identified these using the primary International Classification of Diseases (ICD)-10 diagnosis for the first episode of the hospitalisation. We also analysed elective admissions as an outcome, to test whether health checks had an impact on this aspect of care.
## Practice-level assessment of health checks
We classified practice participation in the DES by calculating the percentage of patients registered on 1 January 2009 on the QOF learning disability register that subsequently received a health check. For this analysis, we restricted to 289 practices with complete data from 1 January 2009 to 31 December 2012, including all adults with IDs irrespective of whether they received a health check (figure 1). We defined full practice participation (n=126) as ≥50% of their adults with ID having a health check by 31 December 2010. Practices (n=68) with <25% adults having a health check by 31 December 2012 were classed as non-participating, with the remainder (n=95) having participation rates of 25-50%. We then compared practice
## Individual analysis of first health check
For our analysis of individuals, we carried out a matched cohort analysis that compared within participant, the rate of admission after the first recorded health check from 1 January 2009 to 31 December 2013, with that seen before the health check [fig_ref] Figure 1: Summary of number of practices, adults with intellectual disability [/fig_ref]. Up to seven controls (with no record of ID) were matched on age, sex and practice to control for any temporal trends in admissions during the study. In total, 7487 adults with ID aged 18-84 with a first health check were identified and matched to 46 408 controls. We excluded the period 30 days on either side of the health check to avoid it directly leading to an admission, or being the result of a recent discharge from hospital. All patients were required to be registered for at least 90 days prior to the health check, and be alive for 90 days after it. All patients were followed to 31 December 2013, or their death if it was earlier. Those who de-registered from their practice were still included in the follow-up as linkage to hospital admissions continues as long they remain resident in England.
Finally, we carried out a complementary analysis using 6922 adults with ID without health checks (figure 1). We allocated a random index date based on the known dates of the health checks, and similarly matched them to 47 662 population controls. We then repeated the above analysis using the non-health check adults with ID and their controls to check whether any observed changes in admissions for adults with ID were specific to those receiving health checks only.
# Statistical analysis
The analyses used a conditional Poisson model (xtpoisson, Stata V.13) to compare the rate of change over time at a practice or individual level. At the practice level, these were conditioned on practice, and all admissions from patients with ID were counted, using an offset term to account for total time registered. The effect of practice participation on hospital admissions was estimated by the interaction between practice participation (fully vs none) and period . At the individual level, we conditioned on individual as opposed to matchset, as accounting for the matching variables is not paramount in matched cohort analyses. [bib_ref] Ignoring the matching variables in cohort studies -when is it valid and..., Sjölander [/bib_ref] This model was fitted to adults with ID and controls separately, estimating the individual change in hospital admission rate after as compared with before the health check, with an offset accounting for time registered. A combined model of adults with ID and controls with a caseperiod interaction provides an estimate of the effect of health checks on admission rates among adults with ID, adjusted for temporal trends in admissions. All models used a sandwich estimator to obtain robust standard errors.
# Results
## Practice-level analyses of health checks and hospital admissions
Practices fully participating in health checks compared with those not participating (table 1) were more likely to have larger numbers of adults with ID in their practice, as well as higher percentages recorded of those living in communal establishments (median 15.8% vs 5.9%) and having high levels of support need (median 22.2% vs 15.2%).
A [fig_ref] Table 5: Interaction IRRs comparing the change in emergency hospital admission rates post-health check... [/fig_ref] summarises the estimate of the impact of health checks on emergency hospital admissions stratified by individual characteristics, both for adults with ID with and without health checks. A significant rise in admissions among Down's syndrome adults with health checks compared with their controls (IRR=1.55, 95% CI 1.15 to 2.08), was replicated among Down's adults without health checks (IRR=1.55), suggesting a
# Discussion
In this study, we found little evidence to suggest that the introduction of incentivised health checks by NHS England for adults with ID in 2009 had any discernible impact on subsequent overall emergency or elective admissions. However, when we only considered potentially preventable emergency admissions (ACSCs) we found that practices which were fully participating in health checks experienced a greater fall in admissions than those not participating. This beneficial association with preventable admissions was replicated when we looked directly at individuals with ID who had a recorded health check. This analysis also suggested a wider benefit of health checks on all emergency admissions among those with more complex health needs. We believe our study is the first to report benefits of health checks for adults with ID on a health outcome as opposed to process measures. [bib_ref] Assessment of an incentivised scheme to provide annual health checks in primary..., Buszewicz [/bib_ref] While a systematic review has shown the effectiveness of health checks in detecting unrecognised health needs in people with ID, [bib_ref] The impact of health checks for people with intellectual disabilities: an updated..., Robertson [/bib_ref] it highlighted the lack of evidence regarding whether their provision translated into important longer term benefits, such as a reduction in avoidable hospitalisations or mortality. The evidence for effectiveness of health checks in general adult populations is similarly uncertain, with no evidence that they reduce mortality, hospitalisation or disability. [bib_ref] General health checks in adults for reducing morbidity and mortality from disease:..., Krogsboll [/bib_ref] In the UK, NHS health checks for 40-74 years old have been shown to increase the identification of cardiovascular risk factors in a large untreated population, [bib_ref] Estimating the yield of NHS Health Checks in England: a population-based cohort..., Forster [/bib_ref] but their impact of longer term outcomes is still unclear.
Reducing emergency hospital admissions is a major international concern to contain healthcare costs, but evidence for successful community interventions is limited. [bib_ref] Reducing emergency admissions through community based interventions, Wallace [/bib_ref] While our primary outcome of emergency hospital admission showed no change after introduction of health checks for participants with ID, evidence for a reduction in potentially preventable admissions was consistent in all our analyses and plausible. Given that admissions for ACSCs represent about 1 in 5 emergency admissions in the UK, [bib_ref] Is secondary preventive care improving? Observational study of 10-year trends in emergency..., Bardsley [/bib_ref] it is unsurprising that we did not detect a change for the broader group. Among adults with ID in our study, admissions for epilepsy contributed about 45% of emergency admissions for ACSCs, so one possible explanation is that health checks are facilitating better overall management of epilepsy and seizures among patients with ID. This would be an important benefit, as improved service provision of patients with ID with epilepsy has been identified as a mechanism for reducing excess mortality among all people with ID. 5 *Has been classed as having severe or profound ID by the GP or has two or more of the following in addition to an ID diagnosis: epilepsy, cerebral palsy or significant mobility problem (wheelchair use or greater problem), severe visual impairment, severe hearing impairment, a continence problem, or use of PEG feeding. ID, intellectual disability; GP, general practitioner; PEG, percutaneous endoscopic gastrostomy. Our study reached a similar conclusion from two different analytic strategies, one based on practice comparisons and the other on individuals. As these used slightly different patient groups and definitions of time, this outcome would not necessarily be expected. For example, individual analyses suggested emergency hospital admissions were rising among patients with ID post-health check, while practice-level analyses showed a fall during 2011-2012. The rise in admissions in the same individuals is partly explained by their ageing over time, plus the requirement to be alive at the health check, resulting in deaths only post-health check (and associated admissions). By contrast, practice trends were based on a fluid group of all patients with ID aged 18-84 years in each year, keeping average age effectively constant and allowing deaths within the year.
Our study has some limitations. We were not able to comment on the quality or overall content of the health checks that have taken place. Although there is published guidance on what the general practitioner (GP) should cover during a health check, 3 a general observation from our data extract is that there is substantial variation in what is recorded, which is likely to mirror what is taking place in the health checks. We have not attempted to make any economic costing of the effectiveness of the health check scheme. A small Scottish trial of nurse delivered health checks for adults with ID demonstrated costeffectiveness compared with standard care. [bib_ref] Practice nurse health checks for adults with intellectual disabilities: a cluster-design, randomised..., Cooper [/bib_ref] However, they did not include hospitalisation costs, except accident and emergency attendances, which may have led to them underestimating potential economic savings.
The analysis at practice level was unmatched, and likely subject to residual confounding from unmeasured factors at both practice and individual level, as we would expect practices that participate in the DES to be different than those that do not, and possibly have differing characteristics of patients with ID. For example, practices that went on to regularly carry out health checks in our study already had lower emergency hospital admissions rates among their patients with ID at the outset in 2009. These practices might have further reduced admissions anyway, and subsequent adoption of health checks may simply be a marker of other improvements in their care over the study period.
In order to control for any practice level-changes over time, we matched individuals with ID receiving health checks to population controls in the same practice. This analysis now adjusts for any change (artefact or real) across practice or hospitals during the study. However, it fails to account for changes specific to people with ID that may have happened in the UK in light of several high profile reports during this period. Therefore, we similarly analysed patients with ID without health checks, assigning them a random date instead of a health check date. Since this group showed no fall in ACS admissions compared with their controls, it provided additional evidence for the effectiveness of health checks. This contrasts with our finding that adults with Down's syndrome increased emergency admissions by 55% post-health check, but since a similar increase was seen in Down's adults without health checks, we concluded the trend was specific to Down's and not health checks. This increase may reflect the premature ageing associated with Down's such as early onset Alzheimer's disease, [bib_ref] Down's syndrome, Roizen [/bib_ref] combined with better survival into middle age in part due to advances in childhood cardiac surgery. [bib_ref] Mortality associated with Down's syndrome in the USA from 1983 to 1997:..., Yang [/bib_ref] In England, an increasing number of adults with ID now live in the community, and as a result, the GP's role in managing their health has increased. Preliminary work around the time of the introduction of health checks in 2009 in England suggested there were no associated higher costs in terms of service use, [bib_ref] Cost estimation of a health-check intervention for adults with intellectual disabilities in..., Romeo [/bib_ref] however costs implications since are not clear and should be evaluated. It has been argued that regular health checks for adults with ID are an efficient way of closing the health inequality gap that this group may experience; however, this may also be widened if more easily managed patients are more likely to get health checks. [bib_ref] Narrowing the health inequality gap by annual health checks for patients with..., Martin [/bib_ref] In our study, the decrease in emergency admission rates for ACSCs was more marked (26%) when we directly compared participating with non-participating practices, which suggests that there may be a 'practice-level benefit' of health checks, where changes in care have benefited all patients with ID within the practice irrespective of whether they have the health check. However, this may be an oversimplification, as a recent serious case review in the UK into the deaths of two adults with ID found that they had been invited to a health check but had failed to attend.Interestingly, our analysis of individuals suggested that health checks produced the greatest benefit in reducing emergency admission to hospital in those with more severe and complex needs.
In summary, to continue to successfully address issues of health inequality and discrimination for adults with ID, the policy implications from our results are: (1) to increase the practice uptake of the health check DES from current levels (<60%) towards a suggested and necessary target of 90%; [bib_ref] Narrowing the health inequality gap by annual health checks for patients with..., Martin [/bib_ref] to ensure that all eligible adults, especially those with the most severe or complex needs, receive an annual health check within practices who participate in the DES. †Has been classed as having severe or profound ID by the GP or has two or more of the following in addition to an ID diagnosis: epilepsy, cerebral palsy or significant mobility problem (wheelchair use or greater problem), severe visual impairment, severe hearing impairment, a continence problem, or use of PEG feeding. ID, intellectual disability; IRR, incidence rate ratio; GP, general practitioner; PEG, percutaneous endoscopic gastrostomy.
## What is already known on this subject
A systematic review of the impact of health checks for people with intellectual disabilities (IDs) in 2014 concluded that while health checks were 'effective in identifying previously unrecognised health needs, including life-threatening conditions', very few studies had 'evaluated the extent to which providing health checks for people with IDs leads to health benefits either in the short or long term'. We were not aware of any study that used emergency hospitalisations as an outcome when evaluating health checks.
## What this study adds
While there was no evidence to suggest that health checks had an impact on overall emergency hospital admissions for adults with ID, the study did reveal a reduction in preventable emergency hospitalisations after the introduction of health checks. These findings should encourage further implementation of health checks to include all general practices in England, in addition to wider participation within practices already carrying them out.
[fig] Figure 1: Summary of number of practices, adults with intellectual disability (ID) and matched controls used in analyses. [/fig]
[fig] Figure 2: Hospital admissions (all emergency, emergency for ACSCs (ambulatory care sensitive conditions), elective) in each quarter during 2009-2012 by practice level of participation in health checks. [/fig]
[table] Table 1: Summary of adults with ID in each practice by practice-level participation in health checks Fully participating practices had ≥50% of their adults with ID with a health check by end of 2010. Non-participating practices had <25% of their adults with ID with a health check by end of 2012. 95 (partial participating) practices did not meet either criterion. In all 72 of the 289 practices had zero participation by 2010, compared to 35 by 2012. *Medians are calculated among all adults with ID registered on 1 January 2009, except for the 'number registered during 2009-2012'. First, a mean is calculated at practice level, and then a median of the practice means is calculated. †Patients who spent at least 1 day registered during 2009-2012. ID, intellectual disabillity.Characteristics of adults with ID with and without health checksAmong the 7487 adults with ID with a first health check between 1 April 2009 and 31 March 2013, the average age was 42.6 years (SD=15.4), with 57.5% being male(table 3). Almost 3 in 10 were classified as having high levels of support needs, with a similar proportion identified as being resident in a communal establishment. By contrast, the 6922 ID adults without a health check were younger (mean=39.0) and less likely to have high levels of support needs or communal living recorded on their record.Individual analyses of health checks and hospital admissionsHospital admission rates before and after the health check are summarised intable 4, and also for adults without health checks using their random index date. For adults with a health check, all emergency admissions rose by 22% from 145.7 to 173.2 annually per 1000 patients. By contrast, in their matched Emergency admissions for ACSCs among adults with health checks showed an association with change in admission rate posthealth check compared with controls (IRR=0.82, 95% CI 0.69 to 0.99). This trend was not replicated in adults with ID without a health check (IRR=1.11, 95% CI 0.92 to 1.36). The change in elective admission rate was similar between ID adults with health checks and controls (IRR=0.96, 95% CI 0.87 to 1.06). [/table]
[table] Table 2: Hospital admissions in 2011-2012 vs 2009-2010 by practice level of participation in health checks Fully participating practices had >50% of their patients with ID with a health check by end of 2010. Non-participating practices had <25% of their patients with ID with a health check by end of 2012. Ninety-five practices did not meet either criteria and were excluded from the comparison. *This represents the within-practice change in admission post-health check compared with prehealth check estimated from the conditional Poisson model. †This represents the within-practice post-health check change in admissions between the fully participating practices versus the non-participating practices estimated from the conditional Poisson model. [/table]
[table] Table 3: Characteristics of registered adult patients with ID by whether they had a health check between April 2009 and March 2013 [/table]
[table] Table 5: Interaction IRRs comparing the change in emergency hospital admission rates post-health check between adults with ID and matched controls, stratified by individual characteristics Patients with ID with a health check Patients with ID without a health check Status at time of health check Change in IRR (95% CI) vs age-sex-practice matched controls* Change in IRR (95% CI) vs age-sex-practice matched controls* *This represents the within-person post-health check change in admissions between the patients with ID and their respective controls (n=46 408 for patients with ID with health check, n=47 622 for patients with ID without health check) estimated from the conditional Poisson model. [/table]
|
Investigating the economic case of a service to support carers of people with dementia: A cross‐sectional survey‐based feasibility study in England
Carers contribute essential support to enable people with dementia to continue living within the community. Admiral Nurses provide specialist dementia support for carers of people with dementia, including offering expert emotional support and guidance, and work to join up different parts of the health and social care system to address needs in a co-ordinated way. The cost-effectiveness of this service is not clear. We undertook a feasibility study to explore related outcomes and costs for these carers. A cross-sectional, clustered survey was undertaken in England in 2017, in areas with and without Admiral Nursing (AN). The survey questionnaire included questions on the characteristics of the carers and the person with dementia, outcomes (care-related quality of life [CRQoL], self-efficacy and subjective well-being), use of health and social care services, out-of-pocket costs and time spent on informal care. We used different econometric techniques to compare the outcomes and the costs of the carers with and without AN services: linear regression, propensity score matching and instrumental variables analysis. These techniques allowed us to control for differences in observed and unobserved characteristics between the two groups of carers which determined outcomes and costs. We concluded that AN services might have a positive effect on carers' CRQoL, self-efficacy and subjective wellbeing. Furthermore, we found little difference in costs between carers using AN and those using usual care, or in the costs of the people with dementia they care for. Our findings provided an initial indication as to whether AN services could be good value for money. The key limitation of the study was the difficulty in controlling for unobserved characteristics because of the cross-sectional nature of our observational data. To diminish this limitation, our survey could be used in future studies following carers with and without AN services over time.
## | introduc ti on
Carers contribute essential support to enable people with dementia to continue living within the community. Admiral Nursing (AN), supported by the charity Dementia UK, is the only specialist nursing service with a specific focus on supporting carers of people with dementia. This service provides carers with expert emotional support and guidance, and aims to join up different parts of the health and social care system to address needs in a co-ordinated way.
Provision of AN services is diverse across England and Wales, and sometimes depends on financial support from charitable grants.
If AN services are to be commissioned across the country and paid for by the public sector, information is needed on their outcomes and costs. To date, there is little quantitative evidence on the outcomes and costs of the AN service. One of the first evaluations of the service compared the mental health of carers using the service (n = 43) with those without the AN service (n = 61), based on the general health questionnaire [bib_ref] Support in the community for people with dementia and their carers: A..., Woods [/bib_ref]. It found that carers with AN support had better outcomes on anxiety and insomnia but similar levels of general health and survival. A 2013 systematic review concluded that the literature is sparse in terms of evaluating the outcomes and costs associated with AN service, but that carers were satisfied with the AN service and they valued its support.
To address this evidence gap, we investigated the outcomes and costs of AN services on carers compared to usual care for carers not using the AN service. The primary objective of the main study, of which this economic study is a part, was to test the acceptability and feasibility of surveying carers of people with dementia through a selfadministered questionnaire including questions on their characteristics, outcomes and costs. The findings of our study could inform the design and implementation of a full-scale evaluative study and could provide early-stage information to commissioners on whether AN services might be effective and cost-effective. The findings reported here contribute to the expanding evidence base, to inform practitioners and policy makers about assessment of the effectiveness and costs of services that support carers of people with dementia.
## | me thods
## | selection of local authorities and recruitment of carers
To recruit AN carers, we selected 16 AN services that were not in- proportion of people over 65 receiving income support, pension credit, or job seekers allowance; proportion of homeless people; population density; proportion of households in social rented accommodation;
proportion of males over 65, and area cost adjustment.
We took a pragmatic approach to sample size calculation using the effect sizes from a randomised controlled trial of community occupational therapy in the Netherlands [bib_ref] Effects of community occupational therapy on quality of life, mood, and health..., Graff [/bib_ref] , given that we did not have commensurate data relating to AN. To run a multivariate analysis controlling for approximately 20 variables, we calculated that a sample of 320 participants would be enough to detect differences of the size observed by [bib_ref] Effects of community occupational therapy on quality of life, mood, and health..., Graff [/bib_ref].
In local authorities with AN, the survey was sent out to those carers for whom the AN providers held postal or email addresses.
In local authorities without AN, national and local 'voluntary sector' groups for carers and people with dementia were contacted to disseminate the survey. Carers residing in local authorities with the AN service and without it (for simplicity, respectively, AN and non-AN carers from now on) received the request to complete the survey at one point in time between January and March 2017.
## | development of the self-administered survey
The survey was developed for self-completion by the carers through an online or postal questionnaire. The questionnaires collected
What is known about the topic - The Admiral Nursing (AN) service provides carers with expert emotional support and guidance.
- One of the first evaluations of AN found that carers using AN had similar general health and survival compared to carers not using AN.
- A 2013 systematic review concluded that quantitative evaluations of AN on outcomes and costs are sparse.
## What this paper adds
- This is the first time that outcomes and costs have been compared between carers with and without AN.
- The outcomes of carers using AN were similar if not better than their counterparts without access to AN services.
- The costs of health and social care services were similar between the two groups.
information on the characteristics of the carer and the person with dementia, carer quality of life, use of health services by the carer and the person with dementia, and use of social care services by the person with dementia. Additionally, we asked about the carer's use of any carer-specific services such as carers' groups or advice services.
Selection of the outcome measures was informed by focus groups and interviews with carers. Through these, the outcomes that the carers thought were most influenced by the quality and level of support they received or might receive from specialist support services were CRQoL, self-efficacy, and mental and physical health.
Additionally, we included measures of well-being because they were likely to be relevant to the policymaker. After cognitive interviewing and piloting with carers, the questionnaire was revised. The survey questionnaire is available in the full report of the research.
Prior to sending out the questionnaire, we tested it through cognitive interviews with nine carers to explore understanding and acceptability of the questions. In addition, we assessed how best to administer the questionnaire, and whether the questionnaire was sufficiently comprehensible, having sought advice from our virtual advisory group members and steering group members.
## | instruments to measure carer outcomes
Having identified the outcomes above, we explored the international literature and, chose, guided by the work of INTERDEM [bib_ref] A European consensus on outcome measures for psychosocial intervention research in dementia..., Moniz-Cook [/bib_ref] , what appeared to be the best measure of each outcome.
For CRQoL we chose ASCOT-Carer. This is a validated instrument for measuring CRQoL of informal unpaid carers who care for adults with a variety of long-term conditions, disability or problems related to old age [bib_ref] Factor structure and construct validity of the adult social care outcomes toolkit..., Rand [/bib_ref]. ASCOT-Carer covers seven domains: spending time on valued or enjoyable activities, having control over daily life, looking after oneself, feeling safe, having social contact, having space and time to be oneself and feeling encouraged and supported in the caring role [bib_ref] Factor structure and construct validity of the adult social care outcomes toolkit..., Rand [/bib_ref].
Each domain has four response categories from 'no needs' to 'high levels of needs'. An Adult Social Care Outcomes Toolkit (ASCOT)-Carer score involves summing the answers to the seven domains, giving a range from nought (lowest CRQoL) to 21 (highest CRQoL).
We measured self-efficacy using the caregiver self-efficacy for managing dementia (SEMD) tool [bib_ref] Measurement and correlates of family caregiver self-efficacy for managing dementia, Fortinsky [/bib_ref]. It includes two domains comprising the carers' confidence in managing the dementia symptoms and their confidence and experiences in using support services. The former comprises five questions with answers on a 10-point scale, where one represents 'not at all certain' and 10 represents 'very certain'. For this domain, a summed score can be derived by summing the question scores with a possible range from five (least self-efficacy) to 50 (greatest self-efficacy).
The domain on support services use is based on four questions with answers on the same 10-point scale. The summed score for this domain has a range from four (least self-efficacy) to 40 (greatest self-efficacy).
We captured mental and physical health through EQ-5D-5L.
EQ-5D-5L is recommended by the National Institute for Health and Care Excellence (NICE) for use in economic evaluations of health and social care interventions in the UK. It comprises five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression [bib_ref] EQ-SD: A measure of health status from the EuroQol Group, Rabin [/bib_ref]. Each dimension is described on five levels: no problems, slight problems, moderate problems, severe problems and unable to/extreme problems. EQ-5D-5L thus describes 3,125 possible health states, which can be converted into a preference-based score anchored at nought for death to one for full health using a national tariff [bib_ref] Valuing health-related quality of life: An EQ-5D-5L value set for England, Devlin [/bib_ref]. The preference-based score reflects the preference for one health state over another. It ranges from −0.281 (for extreme problems on all dimensions) to 1 (no problems on any dimensions).
Although the improvement in health-related quality of life (HRQoL) appeared to be one of the carers' expected outcomes, AN support aims to help carers to 'cope' rather than to increase their HRQoL per se. We therefore excluded HRQoL from the set of outcomes of interest and instead used it to capture carer health as an additional measure of needs within the econometric analysis.
To measure subjective well-being, carers were asked how satis-
## | resource use and costs
We took a broad perspective to investigate the economic case for AN. We costed resource use falling on the NHS, social care services, voluntary services, out-of-pocket costs and time spent caring (informal unpaid care). We measured service use by carers and the person with dementia (reported by the carer) such as specialist support services for carers (including AN), healthcare, social care and voluntary sector services, as well as any out-of-pocket costs incurred in accessing or using associated services. In order to reduce recall bias, questions on resource use referred to the past 4 weeks.
We costed resource use using nationally available unit costs [bib_ref] Home care re-ablement services: investigating the longer-term impacts (prospective longitudinal study). York/Canterburry, Glendinning [/bib_ref] to aid transferability of results. Costs relate to the financial year 2015/16. Unit costs are presented in Tables S1 and S2. The cost dependent variable of key interest is a measure of overall costs calculated as the sum of health and social care costs for both carers and care recipients, including the cost of AN. This assumes that the cost of AN falls on the health and social care budget, although this may not always be the case and may vary across local authorities. Health and social care costs are calculated by multiplying the amount of resources used in the past 4 weeks by the relevant unit cost , and they do not include any out-of-pocket or informal care costs.
In addition, we quantify out-of-pocket costs and informal care costs. The out-of-pocket costs were self-reported by the respondent and they referred to the use of services other than the AN service, such as voluntary sector or social care services. Time spent in caring for the person with dementia was costed using the proxy good method (Van den [bib_ref] Measurement of informal care: An empirical study into the valid measurement of..., Van Den Berg [/bib_ref] [bib_ref] Valuing informal care for economic evaluation, Weatherly [/bib_ref].
This method values informal care time with the market price of a close substitute for a specific care task [fig_ref] Table 2: reports descriptive statistics on the costs before controlling for the observed confounders [/fig_ref]. To obtain information on care time, respondents were asked to indicate which care tasks they carried out from a list of 10 different tasks (obtained from the 2009 'survey of carers in households'; NHS Information Centre, 2010), and the amount of time spent on each of the indicated tasks.
Where people indicated that they were involved in three or more tasks, we asked them to provide the information about hours of care for the three tasks that had taken up the most time. Finally, we asked carers to record how much time they had spent caring overall, in the previous 24 hr.
## | observed confounders
Our qualitative work suggested that AN services tend to target carers with greater needs for support. Carers with greater needs were more likely to self-refer themselves to the AN service. Given the observational nature of our study, AN carers were likely to have different characteristics compared to non-AN carers. A direct or unadjusted comparison between AN and non-AN carers may highlight differences in outcomes and costs driven by the different carers' characteristics, rather than the use of AN. To account for potential sources of confounding, the survey included questions on the characteristics of the carer and the characteristics of the person with dementia. Carer characteristics comprised: gender, age, education, work situation, household financial difficulties, whether the carer was a sole carer, relationship with the care recipient, type and amount of time of care provided, number of years caring, and availability of a replacement for a break. HRQoL measured using EQ-5D-5L was used as a control variable in the analysis, rather than a dependent variable. Additionally, we collected information on care recipient characteristics including age, duration of symptoms of dementia, existence of a formal diagnosis, type of dementia such as Alzheimer, vascular dementia, or other type of dementia, and perceived severity of dementia (categorised from moderate to high severity). As these characteristics were observed, they are called observed confounders.
## | econometric analysis of outcomes and costs
The dependent variables in our econometric analyses are outcomes including ASCOT-Carer score, SEMD score on both management of dementia symptoms and service use domains, overall life satisfaction and happiness yesterday, and costs including overall costs (see Section 2.4), carer and care recipient healthcare costs, and social care costs. We controlled for observed characteristics in the carers and care recipients (the observed confounders) using linear regression analysis and propensity score matching (PSM).
Linear regression analysis provides unbiased estimates of the effect of the AN service if the regression includes all the characteristics which affect outcomes, costs, and the decision to use AN services, where the relationship between these characteristics and the outcomes or costs is constant.
Propensity score matching matches AN and non-AN carers given their propensity score [bib_ref] Constructing a control group using multivariate matched sampling methods that incorporate the..., Rosenbaum [/bib_ref]. The propensity score is the conditional probability of receiving the AN service given the observed confounders. It was obtained by regressing whether or not carers were AN carers on the observed confounders (see Section 2.5) using a logit regression. We then matched AN and non-AN carers using the kernel technique, which compares each AN carer with a counterfactual constructed as the kernel weighted average of multiple individuals in the control group. We used the Epanechnikov kernel function although the choice of the kernel function tends, in practice, not to make a difference [bib_ref] Some practical guidance for the implementation of propensity score matching, Caliendo [/bib_ref] [bib_ref] Nonparametric density and regression estimation, Dinardo [/bib_ref]. The counterfactual mostly depends on the distance between propensity scores of the treated and untreated individual within a specific bandwidth. Following Heckman, Ichimura, and
Todd (1997) Instrumental variable (IV) regression accounts for any unobserved characteristics that determine outcomes, costs, and the use of AN services, such as resilience and ability to care. This econometric approach can deal with these unobserved characteristics through a variable, the instrument, that is correlated with having AN services but has no direct effect on outcomes and costs, and is not correlated with unobserved characteristics that affect costs and outcomes.
The difference in outcomes and costs obtained with the IV analysis represents the effect of the AN service in those carers who were induced to take up the AN service due to the instrument.
We explored two IVs: travel time by car between the carer and the nearest AN service and type of local authority classified as County, London, Metropolitan and Unitary local authority. Travel time is likely to be correlated with accessing AN services (the further away the service the less likely its use) but it is unlikely to directly affect outcomes and health and social costs. We selected type of local authority following [bib_ref] Using cost-effectiveness estimates from survey data to guide commissioning: An application to..., Forder [/bib_ref] , who argued that it determines the local authority's culture and, in turn, the local authority's propensity to invest in services for carers. Some local authorities will therefore be more willing to fund AN than others, but the culture will not have a direct effect on carer's outcomes.
We tested the strength of each instrument with the Cragg-Donald F statistic [bib_ref] Testing identifiability and specification in instrumental variable models, Cragg [/bib_ref].
To deal with missing data we used complete case analysis, thus removing observations with missing data. This assumes that data were missing not at random. In both the regression analysis and the IV approach we estimated robust standard errors. Since carers from the same local authority may exhibit correlations between each other, as a sensitivity analysis, we clustered standard errors within local authorities. In addition, we re-estimated linear regressions by including local authority random effects. As a further sensitivity analysis, we estimated alternative econometric models including generalised linear model (GLM) with log link and normal or gamma distribution for outcomes and two-part model for costs. Regressions including local authority random effects and GLM models used likelihood-based estimators, which assumed that data were missing at random conditional on the variables included. Finally, to investigate whether our models were over fitted, we estimated linear regressions, PSM, and IV regressions using a more parsimonious specification. We carried out all analyses in Stata 15.
## | re sults
## | response and characteristics of study sample
Calculating an overall response rate for our survey is impossible.
While we know how many paper questionnaires we sent to control area third sector organisations, we do not know how many they distributed. Furthermore, while we know which organisations we sent the electronic survey, we do not know how many people received the link, nor how many people chose to open it. Also, we do not know how many carers of people with dementia potentially lived in the local authorities with and without AN services as these data do not currently exist. Twenty-six per cent of the paper questionnaires we distributed to the AN services and third sector organisations were returned to us and were in scope. For the two organisations where we knew how many links were sent to carers, 25% and 43% of carers provided in-scope responses. In total, we received 346 completed questionnaires which were in scope-158 (46%) were from AN service users in our selected areas and 188 (54%) were from carers in non-AN areas. reports the descriptive statistics for carer outcomes before controlling for the observed confounders. CRQoL using ASCOT-Carer was 10.1, on average, and was statistically significantly lower (worse) for AN carers versus non-AN carers (9.6 vs. 10.6) at the 5% level. This is a small difference (i.e. 1% of the average). AN carers reported significantly lower life satisfaction (4.3 vs. 5) which is a small difference on a scale from zero to 10. Self-efficacy on symptoms management was on average 27.4 and self-efficacy on service use 22.3. AN and non-AN carers were statistically similar on both measures of self-efficacy. AN carers were typically as happy as non-AN carers. Note: In the regression analysis, we controlled for carer characteristics (including gender, age, education, work situation, household financial difficulties, whether the carer was a sole carer, relationship with the care recipient, type and amount of time of care provided, number of years caring, availability of a replacement for a break, and HRQoL) and care recipient characteristics (including age, duration of symptoms of dementia, existence of a formal diagnosis, type of dementia such as Alzheimer, vascular dementia, or other type of dementia, and perceived severity of dementia). We used the same independent variables in the logit regression for the calculation of the propensity score to be used in the Propensity Score Matching. Similarly, we controlled for the same independent variables in the IV regression, for which the instrument was the travel time to the closest AN provider.
## | outcomes of an and non-an carers
## | costs of an and non-an carers
No adjustment for multiple testing was implemented because of the feasibility nature of this study. Abbreviations: AN, Admiral Nursing; CI, confidence intervals; Coeff, estimated coefficient on the Admiral Nursing dummy; HRQoL, health-related quality of life; IV, instrumental variable; Std Err, robust standard errors. *p-value < 0.1, **p-value < 0.05, ***p-value < 0.01. dementia (£40) and day sitting (£37). Other costs include home care (£29), other day care service (£15), meals (£10), memory café (£7) and support group (£6). The out-of-pocket costs for services are similar between AN and non-AN carers apart from the cost of day care centre (£34 for AN carer vs. £47 for non-AN carer, p = 0.029) and the cost of home care (£13 vs. £38, p = 0.002).
On average, carers had spent 12 hr providing informal care over the previous 24 hr. Using the top three informal care tasks carried out in the previous 24 hr to cost the care at its closest market replacement value, the cost was £459 on average. There was no statistically significant difference in these costs between AN and non-AN carers.
# | regression analysis
We used linear regressions to estimate the effect of AN on outcomes and costs after controlling for the observed confounders ; . . Similar results were obtained when standard errors were clustered within local authorities or local authority random effects were added to the model . This was due to a generally low intraclass correlation varying between 0.029 (for overall satisfaction) and 0.186 (for the care recipient's health care costs). Results were robust to alternative econometric models, including GLM with log link and normal or gamma distribution for outcomes, and twopart model for costs . Similar estimates obtained through the model including local authority random effects and GLM suggested that our results were robust to the assumption that data were missing at random . Finally, results were robust also to a specification including fewer control variables .
## | propensity score matching
We constructed the propensity score by regressing whether or not carers were AN carers on the observed confounders using a logit regression. Carers taking care of a person with vascular dementia had twice the odds of being in the AN group compared to carers of people with Alzheimer's disease. Carers with Master's or higher degrees had 15%-23% lower odds of being in the AN group compared to carers with no university education. The longer the time since dementia diagnosis, the less likely carers were to be in the AN group .
The kernel PSM outperformed nearest neighbour and calliper (with a radius of 0.2) PSM in terms of average standardised difference of the covariates . Kernel PSM performed well in terms of standardised difference in covariates and matching of the propensity score across AN and non-AN carers . shows that using PSM to analyse carer outcomes produced results mostly in line with the linear regression results. The self-efficacy measure related to service use was the only exception. AN carers were associated with greater self-efficacy on service use by almost three points compared to non-AN carers, although this association was statistically significant at the 5% level. The PSM produced a statistically insignificant estimate of the association between being an AN carer and costs, similar to that of the regression analysis. These results were robust to a more parsimonious specification of the logit regression for the estimation of the propensity score .
# | iv analysis
Travel time to the closest AN provider is on average 13 min and its distribution is right-skewed . Non-AN carers were 17 min (0.286 hr) on average away from AN services whilst AN carers were 9 min away (0.151 hr). This difference is statistically significant at the 1% level. Travel time was a strong instrument as the Cragg-Donald F statistic is between 41 and 56 [fig_ref] Table 2: reports descriptive statistics on the costs before controlling for the observed confounders [/fig_ref]. We could not reject the hypothesis of no effect of travel time on outcomes when additional instruments (i.e. type of local authority dummies) are employed. This suggested that travel time has no relationship with the outcomes and was therefore a suitable variable to use in this part of our analysis.
Finally, shows the results of the IV approach for outcomes and costs when travel time was used as an instrument. IV results were in line with those from the regression and PSM analysis showing a statistically insignificant effect of AN on outcomes and costs, except for ASCOT-Carer which is weakly significant (at the 10% level). When standard error in the IV regressions were clustered within local authorities the effect of AN on happiness yesterday became statistically significant at the 5% level . Results remain robust to a specification including fewer covariates .
## | d iscuss i on and con clus i on s
Carers of people with dementia receiving AN services reported having slightly lower CRQoL and subjective well-being than carers of people with dementia not receiving AN services. After controlling for differences in their observed and unobserved characteristics, however, we found that the CRQoL, self-efficacy and subjective well-being of AN carers was similar if not better than carers without access to AN services. The costs of health and social care services were similar across the two groups.
To our knowledge this is the first time that outcomes and costs have been compared in AN versus non-AN carers. Our analysis used data from a recent feasibility survey (2017). We received 346 completed questionnaires which were in scope and most questionnaires were answered fully. The size and extent of these data allowed us to use different econometric techniques in an attempt to control for any systematic differences in the characteristics of AN and non-AN carers, thereby minimising the risk of bias due to confounding.
Our study, however, has several limitations. We cannot be sure that our non-probability sample reflects a fully representative sample of the target population of carers of people with dementia and this impacts on the transferability of our results beyond our sample of carers.
Our analysis is based on cross-sectional, observational data.
Although we controlled for several observed confounders, there exists a risk of confounding if AN and non-AN carers are systematically different in characteristics that determine their outcomes and costs, and that we are not able to control for. The IV analysis helps us to address the selection bias due to unobserved characteristics and it relies on our choice of instrument and refers to the subgroup of carers who were induced to take up the AN service given their proximity to the service. Admiral Nursing is a complex intervention and it is not possible to fully disentangle the effect on carers who received AN support in the past from other support services which they may also have utilised. Diversity in the referral process, such as referral to AN after a triage assessment or via self-referral across AN providers, may generate high heterogeneity within the group of AN carers, hampering us from identifying an effect.
A feature of the survey design was that questions on service use related to the past 4 weeks for those services we thought would be used on a regular basis. In this way, we aimed to reduce recall bias.
Moreover, the use of a postal, self-administered survey meant that we could not comprehensively measure level of dementia severity, which is likely to affect the carer's needs. We could not find an appropriate measure in the published literature, hence using our own, low carer burden questionnaire which provides only partial information. shows that, in the regression analysis, higher levels of perceived severity are always statistically significantly associated with worse outcomes.
We were unable to include an ASCOT-Carer score which included preference weights for the UK population as this is not yet available [bib_ref] Carer social care-related quality of life outcomes: Establishing preference weights for the..., Batchelder [/bib_ref]. As an interim approach, we summed the domains and this assumes that all the domains are equally important. Finally, we put a monetary value or cost on informal care time using the proxy good method. This might involve an overestimation and it is only one out of many approaches to value informal care (Van den [bib_ref] Economic valuation of informal care: Lessons from the application of the opportunity..., Van Den Berg [/bib_ref] [bib_ref] Valuing informal care for economic evaluation, Weatherly [/bib_ref].
Considering these limitations, our findings provide an initial indication as to whether AN services could offer value for money.
Full-scale evaluation is required to make more definitive recommendations. Future research could build on our survey and collect data over multiple time-points to better estimate the causal effect of AN services on carer outcomes and costs. Our survey could also be adapted to explore the outcomes and costs of other services for carers of people with dementia.
## Ack n owled g em ents
[table] Table 2: reports descriptive statistics on the costs before controlling for the observed confounders. Over 4 weeks, the average cost of health and social care services of the carer-care recipient dyad was approximately £1,000. This includes £36 for the AN service, £239 for healthcare services for the carer, £324 for healthcare services for the care recipient, and £627 for social care services. The costs varied widely. There were some differences in the cost of AN carer-care recipient dyads versus non-AN, although these differences were not statistically significant.The largest out-of-pocket cost was for short respite or break services (£240), followed by day care centre for the person with TA B L E 1 Descriptive statistics on outcomes [/table]
|
Treating asthma with omega-3 fatty acids: where is the evidence? A systematic review
Background: Considerable interest exists in the potential therapeutic value of dietary supplementation with the omega-3 fatty acids. Given the interplay between pro-inflammatory omega-6 fatty acids, and the less pro-inflammatory omega-3 fatty acids, it has been thought that the latter could play a key role in treating or preventing asthma. The purpose was to systematically review the scientific-medical literature in order to identify, appraise, and synthesize the evidence for possible treatment effects of omega-3 fatty acids in asthma.Methods: Medline, Premedline, Embase, Cochrane Central Register of Controlled Trials, CAB Health, and, Dissertation Abstracts were searched to April 2003. We included randomized controlled trials (RCT's) of subjects of any age that used any foods or extracts containing omega-3 fatty acids as treatment or prevention for asthma. Data included all asthma related outcomes, potential covariates, characteristics of the study, design, population, intervention/exposure, comparators, and co interventions.Results: Ten RCT's were found pertinent to the present report.Conclusion:Given the largely inconsistent picture within and across respiratory outcomes, it is impossible to determine whether or not omega-3 fatty acids are an efficacious adjuvant or monotherapy for children or adults. Based on this systematic review we recommend a large randomized controlled study of the effects of high-dose encapsulated omega-3 fatty acids on ventilatory and inflammatory measures of asthma controlling diet and other asthma risk factors. This review was limited because Meta-analysis was considered inappropriate due to missing data; poorly or heterogeneously defined populations, interventions, intervention-comparator combinations, and outcomes. In addition, small sample sizes made it impossible to meaningfully assess the impact on clinical outcomes of co-variables. Last, few significant effects were found.
# Background
Asthma is one of the most common chronic conditions, and affects both adults and children. Its prevalence has been noted to have increased significantly in Western nations with overall prevalence rates ranging from 7-15% [bib_ref] Diet and asthma, Mckeever [/bib_ref]. The National Heart, Lung, and Blood Institute (NHLBI) has defined asthma as a chronic inflammatory disease of the airways [bib_ref] Bethesda, National Institutes of Health 3. Simopoulos AP: The importance of the..., Heart [/bib_ref].
The inflammatory response is complex, and involves a variety of inflammatory cell types including mast cells, alveolar macrophages, neutrophils, eosinophils, lymphocytes, platelets, and a variety of inflammatory mediators [bib_ref] Expert Panel Report: Guidelines for the Diagnosis and Management of Asthma Update..., Busse [/bib_ref]. Over time, various therapeutic strategies have been developed to manage asthma, including the use of short acting beta-2 agonist bronchodilator medications as symptom relievers and anti-inflammatory preventer medications such as inhaled corticosteroids and oral leukotriene antagonists [bib_ref] Bethesda, National Institutes of Health 3. Simopoulos AP: The importance of the..., Heart [/bib_ref]. Given that the cause of airway inflammation can be multifactorial and involves a multitude of cell types and inflammatory mediators, the medications used to control inflammation may act at several different points in the inflammatory pathway [bib_ref] Bethesda, National Institutes of Health 3. Simopoulos AP: The importance of the..., Heart [/bib_ref]. Despite many therapeutic advances over the last thirty years, asthma continues to result in significant childhood and adult morbidity .
Given the rapidity of the increase in asthma's prevalence, environmental factors rather than increased genetic susceptibility are more likely to be responsible for this trend. Thus, there has been increased interest in investigating various environmental factors that may contribute to this illness, including diet. McKeever and Britton recently concluded that the trial evidence is less conclusive about the protective effects of fish oil intake on asthma and/or allergy than is the observational evidence [bib_ref] Diet and asthma, Mckeever [/bib_ref]. There has been considerable interest in the potential therapeutic and protective value of dietary supplementation with the omega-3 fatty acids. Horrobin hypothesized that the low incidence of asthma in the northern aboriginal population stems from their consumption of large quantities of oily fish rich in omega-3 fatty acids [bib_ref] Low prevalences of coronary heart disease (CHD), psoriasis, asthma and rheumatoid arthritis..., Horrobin [/bib_ref]. Moreover, given the competitive interplay (e.g., for enzymes in the metabolic pathway and for positions in cell membranes) between pro-inflammatory omega-6 fatty acids, which may contribute to the cascade of events marking asthma, and the less pro-inflammatory omega-3 fatty acids, it has been thought that the latter could play a key role in treating or preventing asthma [bib_ref] Polyunsaturated fatty acids and inflammatory diseases, Gil [/bib_ref]. From this perspective, respiratory benefits might be attributable to changes in the status of mediators of inflammation such as leukotrienes and prostaglandins.
A systematic review was conducted to review the scientificmedical literature to identify, appraise and synthesize the evidence regarding the possible treatment and preventive value of omega-3 fatty acids in asthma [bib_ref] Health Effects of Omega-3 Fatty Acids on Asthma, Schachter [/bib_ref]. This report describes the randomized controlled study (RCT) evidence concerning the efficacy of the treatment of asthma with omega-3 fatty acid supplementation in adult or pediatric populations. In addition to determining whether this intervention improves respiratory outcomes, its impact on the mediators of inflammation thought to be relevant to the pathogenesis of asthma is evaluated with this RCT evidence. Safety data are also presented. Searches were not restricted by language of publication, publication type, or study design except with the MeSH term "dietary fats," which was limited by study design to increase its specificity. Search elements included: scientific terms, with acronyms, as well as generic and trade names relating to the exposure and its sources (e.g., eicosapentaenoic acid [EPA]; MaxEPA ® ; fish oil); and relevant population terms (e.g., asthma; inflammation). Additional published or unpublished literature was sought through manual searches of reference lists of included studies and key review articles, and from the files of content experts. A final set of 1,010 unique references was identified and posted to an Internet-based software system for review.
# Methods
Studies were considered to be relevant if they described human populations of any age, involved any type of study design, and investigated the use of any foods or extracts known to contain omega-3 fatty acids from any source (e.g., marine; plant) or of any type (e.g., EPA; alpha linolenic acid acid [ALA]) or dose, and used as a treatment or prevention. These eligibility criteria distinguish our work from that of a recent Cochrane review, which exclusively focused on the role of fish oil [bib_ref] Dietary marine fatty acids (fish oil) for asthma in adults and children, Woods [/bib_ref]. For the purposes of this report, however, only treatment RCT's are described because this research design is considered the gold standard for investigating questions of treatment efficacy or effectiveness [bib_ref] Improving the quality of reports of meta-analyses of randomised controlled trials: the..., Moher [/bib_ref]. Populations in treatment studies had to have received a formal diagnosis of asthma. Studies where an asthmatic response was experimentally induced in non-asthmatic populations were excluded. Methods had to have been employed in primary studies to identify the omega-3 fatty acid intervention/exposure. Studies investigating "polyunsaturated fatty acids" or specific diets were included if explicit reference was made to their omega-3 fatty acid content. A treatment study could assess respiratory outcomes, data concerning mediators of inflammation, or adverse effects. Forced expiratory volume in 1 second (FEV 1 ) was selected as the primary outcome, given its status as a gold standard index of pulmonary function for asthma.
Two levels of screening for relevance, and two reviewers per level, were employed (bibliographic records, then full articles) following calibration exercises. Disagreements were resolved by forced consensus and if needed, third party intervention. Following a calibration exercise, three reviewers independently abstracted the contents of each included study using an electronic form. A second abstractor verified all data, which included outcomes (i.e., respiratory, mediators of inflammation, safety; discontinuations), potential covariates (e.g., fatty acid content of blood lipid biomarkers, tissue ratios of fatty acid during the investigative period), in addition to the characteristics of the report (e.g., publication status), study (e.g., sample size; design), population, intervention/exposure (e.g. omega-3 fatty acid types and sources; intervention length), comparators (e.g., placebo), and co interventions (e.g. asthma medications, omega-6 fatty acids). Dual-assessor evaluations of RCTs' quality employed Jadad's validated items concerning randomization, double blinding, and the reporting of withdrawals and dropouts, [bib_ref] Assessing the quality of reports of randomized clinical trials: is blinding necessary?, Jadad [/bib_ref] and b) Schulz's validated question concerning the adequacy of allocation concealment [bib_ref] Empirical evidence of bias. Dimensions of methodological quality associated with estimates of..., Schulz [/bib_ref]. A total Jadad score below 3 (out of 5) indicates low study quality. Select items from a third validated instrument permitted the assessment of the quality of other designs, which are excluded from the present report [bib_ref] The feasibility of creating a checklist for the assessment of the methodological..., Downs [/bib_ref]. The applicability, or generalizability, of a study to a broadbased North American population was determined by the degree to which the study population included asthma patients who were otherwise "healthy," potentially exhibiting asthma concomitants (e.g., atopy), representing a somewhat broad demographic spectrum (e.g., sex, ethnicity), living a "typical" North American lifestyle, possibly receiving "typical" asthma treatment, and failing to exhibit major co morbidity.
# Results
Of the 1,010 bibliographic records entered into the initial screening for relevance, 851 were excluded. All but 5 of the remaining 159 reports were then retrieved and subjected to a more rigorous relevance assessment. The second relevance screening then excluded 122 reports. In total, 31 reports, describing 26 unique studies, were deemed relevant for the systematic review of treatment or prevention evidence, with 5 studies each described by two reports. Nineteen treatment studies were identified, ten of which were RCTs pertinent to the present report [fig_ref] Table 1: RCT Evidence of Omega-3 Fatty Acids to Improve Respiratory Outcomes in Asthma [/fig_ref] [ [bib_ref] Effect of fish-oil derived omega-3 fatty acid supplements on asthma control, Mcdonald [/bib_ref] [bib_ref] Dietary supplementation with fish oil rich in omega-3 polyunsaturated fatty acids in..., Nagakura [/bib_ref] [bib_ref] Effect of dietary intake of omega-3 and omega-6 fatty acids on severity..., Hodge [/bib_ref] [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] [bib_ref] Effects of dietary supplementation with n-3 fatty acids compared with n-6 fatty..., Okamoto [/bib_ref] [bib_ref] Dietary fish oil effects on seasonal hay fever and asthma in pollen-sensitive..., Thien [/bib_ref] [bib_ref] Evening primose oil and fish oil are ineffective as supplementary treatment of..., Stenius-Aarniala [/bib_ref] [bib_ref] Effect of eicosapentaenoic acid in asthma, Kirsch [/bib_ref] [bib_ref] Treatment of asthma with lipid extract of New Zealand green-lipped mussel: a..., Emelyanov [/bib_ref] [bib_ref] Effect of a fish oil diet on asthma: results of a 1-year..., Dry [/bib_ref]. Two treatment RCTs employed a crossover design [bib_ref] Effect of fish-oil derived omega-3 fatty acid supplements on asthma control, Mcdonald [/bib_ref] [bib_ref] Evening primose oil and fish oil are ineffective as supplementary treatment of..., Stenius-Aarniala [/bib_ref].
Nine treatment studies employed other designs, and seven studies investigated the possible primary preventive value of omega-3 fatty acid intake. The ten treatment RCTs were described in 13 reports. Hodge et al.'s pediatric trial results [bib_ref] Effect of dietary intake of omega-3 and omega-6 fatty acids on severity..., Hodge [/bib_ref] were first disseminated as an abstract [bib_ref] Effect of fish oil supplements on severity of asthma in children, Hodge [/bib_ref] Kirsch et al.'s adult treatment RCT had its clinical outcome data reported in one publication [bib_ref] Effect of eicosapentaenoic acid in asthma, Kirsch [/bib_ref] and its mediators of inflammation data in another [bib_ref] Alterations in human leukocyte function induced by ingestion of eicosapentaenoic acid, Payan [/bib_ref]. The lead author (Jonathan Arm) of two study reports with overlapping data recommended to us that, to avoid entering duplicate data into our review, we should consult the first report [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] for all data other than allergen-challenge results, which should be obtained from the second document [bib_ref] The effects of dietary supplementation with fish oil lipids on the airways..., Arm [/bib_ref]. To simplify matters, only one document per study is referred to in this report's text or table. Only one treatment RCT was not described by at least one published report [bib_ref] Effect of fish-oil derived omega-3 fatty acid supplements on asthma control, Mcdonald [/bib_ref]. All treatment RCT reports were published in English.
Two trials exclusively randomized children, [bib_ref] Dietary supplementation with fish oil rich in omega-3 polyunsaturated fatty acids in..., Nagakura [/bib_ref] [bib_ref] Effect of dietary intake of omega-3 and omega-6 fatty acids on severity..., Hodge [/bib_ref] one included both older adolescents and adults, [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] and the remainder randomized adults [bib_ref] Effect of fish-oil derived omega-3 fatty acid supplements on asthma control, Mcdonald [/bib_ref] [bib_ref] Effects of dietary supplementation with n-3 fatty acids compared with n-6 fatty..., Okamoto [/bib_ref] [bib_ref] Dietary fish oil effects on seasonal hay fever and asthma in pollen-sensitive..., Thien [/bib_ref] [bib_ref] Evening primose oil and fish oil are ineffective as supplementary treatment of..., Stenius-Aarniala [/bib_ref] [bib_ref] Effect of eicosapentaenoic acid in asthma, Kirsch [/bib_ref] [bib_ref] Treatment of asthma with lipid extract of New Zealand green-lipped mussel: a..., Emelyanov [/bib_ref] [bib_ref] Effect of a fish oil diet on asthma: results of a 1-year..., Dry [/bib_ref]. One report did not identify the age of its participants [bib_ref] Effect of a fish oil diet on asthma: results of a 1-year..., Dry [/bib_ref]. Given the differences in the clinical pictures of adult and (especially young) children's asthma, results obtained from studies enrolling the different populations were not combined.
Given the largely inconsistent picture within and across respiratory outcomes, it is impossible to conclude one way or another whether omega-3 fatty acids are an efficacious adjuvant or monotherapy in improving respiratory outcomes in asthmatic adults or children. This is perhaps best illustrated by what was observed with respect to the primary outcome variable, FEV 1 .
The adult RCTs revealed a somewhat contradictory picture of efficacy with respect to FEV 1 . One very small adult study conducted by , which compared "uncontrolled dosing" of perilla seed oil (i.e., pourable oil) and similarly uncontrolled dosing of corn oil over a short intervention period (4 weeks), reported a significant positive clinical effect [bib_ref] Effects of dietary supplementation with n-3 fatty acids compared with n-6 fatty..., Okamoto [/bib_ref]. However, two RCTs each observed no benefit. Kirsch et al.'s RCT compared high and low doses of EPA ethyl ester over 8 weeks in a small study (n = 12), [bib_ref] Effect of eicosapentaenoic acid in asthma, Kirsch [/bib_ref] whereas Emelyanov investigated the impact of low dose EPA+DHA (versus an olive oil placebo) over 8 weeks in the systematic review's highest quality RCT [bib_ref] Treatment of asthma with lipid extract of New Zealand green-lipped mussel: a..., Emelyanov [/bib_ref]. The latter study also randomized the second largest sample population (n = 46) included in our evidence review. Investigation of additional clinical outcomes for adults revealed that a) for AM peak expiratory flow (PEF), two studies reported a significant benefit, Proceeding from highest omega-3, or lowest omega-6/omega-3, fatty acid content of intervention; n-6 = omega-6 FAs; ALA = alpha linolenic acid; DHA = docosahexaenoic acid; EPA = eicosapentaenoic acid; Length = intervention length; Design = research design; n = sample size; pts = study participants; NR = not reported; wk = week(s); mo = month; RCT = randomized controlled trial; Jadad total quality score = /5; Applicability I = greatest applicability to North American population of asthma patients; Applicability II = somewhat narrow applicability; Applicability III = very narrow applicability; Applicability X = insufficient data to determine applicability [bib_ref] Effects of dietary supplementation with n-3 fatty acids compared with n-6 fatty..., Okamoto [/bib_ref] [bib_ref] Treatment of asthma with lipid extract of New Zealand green-lipped mussel: a..., Emelyanov [/bib_ref] while three reported no benefit, [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] [bib_ref] Dietary fish oil effects on seasonal hay fever and asthma in pollen-sensitive..., Thien [/bib_ref] [bib_ref] Evening primose oil and fish oil are ineffective as supplementary treatment of..., Stenius-Aarniala [/bib_ref] [bib_ref] Treatment of asthma with lipid extract of New Zealand green-lipped mussel: a..., Emelyanov [/bib_ref] whereas b) five studies failed to observe a significant benefit for PM PEF. Inconsistent results were observed with respect to each of three self-reported outcomes, including bronchodilator use, asthma symptom scores and asthma severity ratings. With regard to pediatric studies, one RCT observed no benefit with respect to FEV 1 [bib_ref] Effect of dietary intake of omega-3 and omega-6 fatty acids on severity..., Hodge [/bib_ref]. There were too few pediatric studies with which to observe any consistent patterns of result for any outcome.
No study, either involving adults or children, received an allocation concealment rating other than "Unclear." Only two adult studies, [bib_ref] Effects of dietary supplementation with n-3 fatty acids compared with n-6 fatty..., Okamoto [/bib_ref] [bib_ref] Evening primose oil and fish oil are ineffective as supplementary treatment of..., Stenius-Aarniala [/bib_ref] in addition to the RCT for which age data were not provided, [bib_ref] Effect of a fish oil diet on asthma: results of a 1-year..., Dry [/bib_ref] were found to exhibit low (total Jadad score: <3) study quality. There was very limited generalizability to a broad spectrum North American population of asthmatics for all but one of the studies, which was conducted in the United States [bib_ref] Effect of eicosapentaenoic acid in asthma, Kirsch [/bib_ref].
Several key observations made it inappropriate to consider conducting meta-analysis, including missing data (e.g., sample sizes; type and dose of omega-3 fatty acids; types and doses of concurrent asthma medication with the potential to confound results; placebo contents), poorly or heterogeneously defined populations (e.g., age, exact diagnosis, asthma severity, asthma triggers or concomitants), interventions (e.g., sources, types i.e. pharmaceutical versus health-food store grade fish oil, doses, and the degree of control over dosing of omega-3 fatty acids [capsules versus pourable oils]), intervention-comparator combinations, and outcomes. Some of the published studies used peak flow rate as their primary outcome measure, rather than the more reliable FEV 1 . In addition to typically small sample sizes, these study limitations also made it impossible to meaningfully assess, even qualitatively, the impact on clinical outcomes of co variables such as the source, type and dose of omega-3 fatty acids. This objective was further complicated by the fact that few significant effects were found. These study limitations likewise made it impossible to ascertain, based on four adult trials [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] [bib_ref] Effects of dietary supplementation with n-3 fatty acids compared with n-6 fatty..., Okamoto [/bib_ref] [bib_ref] Evening primose oil and fish oil are ineffective as supplementary treatment of..., Stenius-Aarniala [/bib_ref] [bib_ref] Effect of eicosapentaenoic acid in asthma, Kirsch [/bib_ref] and one pediatric RCT, [bib_ref] Effect of dietary intake of omega-3 and omega-6 fatty acids on severity..., Hodge [/bib_ref] whether omega-3 fatty acids consistently and positively influence the lipid mediators of inflammation in ways congruent with the biological model implicating the lipoxygenase and cyclooxoygenase pathways in asthma.
Six of eight adult RCTs [bib_ref] Effect of fish-oil derived omega-3 fatty acid supplements on asthma control, Mcdonald [/bib_ref] [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] [bib_ref] Dietary fish oil effects on seasonal hay fever and asthma in pollen-sensitive..., Thien [/bib_ref] [bib_ref] Evening primose oil and fish oil are ineffective as supplementary treatment of..., Stenius-Aarniala [/bib_ref] [bib_ref] Effect of eicosapentaenoic acid in asthma, Kirsch [/bib_ref] [bib_ref] Treatment of asthma with lipid extract of New Zealand green-lipped mussel: a..., Emelyanov [/bib_ref] and the two pediatric trials [bib_ref] Dietary supplementation with fish oil rich in omega-3 polyunsaturated fatty acids in..., Nagakura [/bib_ref] [bib_ref] Effect of dietary intake of omega-3 and omega-6 fatty acids on severity..., Hodge [/bib_ref] each reported safety data. Most of the adverse events were related to the capsule delivery of oils, rather than to the oils per se [bib_ref] Effect of fish-oil derived omega-3 fatty acid supplements on asthma control, Mcdonald [/bib_ref] [bib_ref] Dietary supplementation with fish oil rich in omega-3 polyunsaturated fatty acids in..., Nagakura [/bib_ref] [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] [bib_ref] Dietary fish oil effects on seasonal hay fever and asthma in pollen-sensitive..., Thien [/bib_ref]. On several occasions, difficulty swallowing capsules led to a withdrawal from the study. In two studies, participants may have had difficulties swallowing 18 capsules of oil per day, yet these difficulties were not reported [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] [bib_ref] Dietary fish oil effects on seasonal hay fever and asthma in pollen-sensitive..., Thien [/bib_ref]. The single most serious reaction noted was an undefined number of episodes of nausea and vomiting after ingesting fish oil capsules, which in turn led to study withdrawal [bib_ref] Dietary fish oil effects on seasonal hay fever and asthma in pollen-sensitive..., Thien [/bib_ref]. There were unspecified numbers of children and adults who experienced mild gastrointestinal upset, yet not all of these individuals had received the omega-3 fatty acid exposure. Fishy tasting eructations were reported by some study participants.
# Discussion
Although the use of omega-3 fatty acid supplementation is popular and its purported beneficial effects are biologically plausible, the existing body of evidence from clinical trials does not provide strong support for its clinical effectiveness. The adverse effects of nausea and abdominal discomfort do however appear to be consistent across studies and appear to be dose-related. There is no consistent evidence supporting an improvement in objective measures of ventilatory lung function in adults with the majority of studies of FEV 1 [bib_ref] Effects of dietary supplementation with n-3 fatty acids compared with n-6 fatty..., Okamoto [/bib_ref] [bib_ref] Effect of eicosapentaenoic acid in asthma, Kirsch [/bib_ref] [bib_ref] Treatment of asthma with lipid extract of New Zealand green-lipped mussel: a..., Emelyanov [/bib_ref] and PEF [bib_ref] Effect of dietary supplementation with fish oil lipids on mild asthma, Arm [/bib_ref] [bib_ref] Effects of dietary supplementation with n-3 fatty acids compared with n-6 fatty..., Okamoto [/bib_ref] [bib_ref] Dietary fish oil effects on seasonal hay fever and asthma in pollen-sensitive..., Thien [/bib_ref] [bib_ref] Evening primose oil and fish oil are ineffective as supplementary treatment of..., Stenius-Aarniala [/bib_ref] [bib_ref] Treatment of asthma with lipid extract of New Zealand green-lipped mussel: a..., Emelyanov [/bib_ref] failing to find a benefit of omega-3 fatty acid supplementation.
More often than not, other outcomes such as asthma symptoms, clinical severity scores, and 'rescue inhaler use' did not significantly improve. No RCT's, which investigated non-marine sources of omega-3 fatty acids, were found. In only the study by Hodge et al, was dietary manipulation of n-3 PUFA performed as part of the treatment phase [bib_ref] Effect of dietary intake of omega-3 and omega-6 fatty acids on severity..., Hodge [/bib_ref]. While potential changes to mediator generation may not be as marked with dietary manipulation as compared to supplement administration, an acceptable food-based approach may be better tolerated by individuals in the long run, should benefit be ultimately proven.
Since this systematic review has been completed, an additional study has been published that would have met our eligibility criteria [bib_ref] Fly AD: Protective effect of fish oil supplementation on exercise-induced bronchoconstriction in..., Mickleborough [/bib_ref]. Mickleborough and co-workers have studied the protective effect of fish oil supplementation on exercise-induced bronchoconstriction (EIB) in 16 adult asthmatic patients. Those whose diet was supplemented with fish oil had improved pulmonary function to below the diagnostic EIB threshold in post-exercise FEV1, a reduced fall in FEV1 15 minutes post exercise, reduced bronchodilator requirement, and a reduction of a variety of inflammatory mediators. The authors concluded that fish oil might represent a possible non-pharmacologic therapeutic strategy for asthmatic subjects with EIB. However, this observation does not alter our own overall conclusions [bib_ref] Fly AD: Protective effect of fish oil supplementation on exercise-induced bronchoconstriction in..., Mickleborough [/bib_ref].
The inability, due to strong inter-study clinical heterogeneity (i.e., populations; interventions; outcomes), to combine the results in a formal meta-analysis reduces the effective sample size from which to discern benefits or harm. It is not possible to determine if differences in clinical benefit between studies was caused by differences in characteristics of the study groups, interventions or outcomes, or simply due to chance. To clarify this, large randomized clinical trials are needed. A daily intervention of at least 3 grams of omega-3 fatty acids, which is considered a high adult dose, may well be required. Its source (e.g. marine, plant) and constituents (i.e., ALA, EPA, EPA+DHA) need to be well defined, although it is as yet unclear which combinations of omega-3 fatty acids should be employed. Not only should omega-3 supplementation be studied, but also dietary manipulation. Duration of intervention, minimal effective dosage, and formulation will need to be determined. Outcomes should reflect the dimensions of asthma; symptoms, use of rescue inhaler, ventilatory function, variability in function, and acute exacerbations. Exhaled nitric oxide is a newer measure of asthma inflammation. Exhaled breath condensate could be collected and tested for leukotrienes and prostaglandins, mediators of inflammation potentially influenced by omega-3 fatty acids. Induced sputum techniques could also be utilized to assess airway inflammatory cell and mediator characteristics.
Many sources of influencing and potentially confounding variables need to be controlled in the study design, measured during the study and perhaps adjusted for in the analysis.
Age and gender influence asthma severity and prevalence of atopy. Co-existing anti-inflammatory medications may overwhelm a similar but lesser effect of omega-3 fatty acids. Dietary omega-6 fatty acids need to be controlled for since they may compete with omega-3 fatty acids for enzymes and positions in cell membranes and neutralize the effect. Using encapsulated rather than pourable oils would add precision to the dose of omega-3 fatty acid supplementation. A relatively long follow-up period of one or two years would facilitate the assessment of exacerbation rates. Changes in exposure to indoor allergens and irritants may also influence asthma control. In conclusion, given the data available to date, and despite an acceptable safety profile, a recommendation as to whether omega-3 fatty acids are a viable treatment modality is at present unresolved. Recommendations for future research follow directly from observations of the problems and limitations observed in the included studies.
# Conclusion
Given the largely inconsistent picture within and across respiratory outcomes, it is impossible to determine whether or not omega-3 fatty acids are an efficacious adjuvant or monotherapy for children or adults. Based on this systematic review we recommend a large randomized controlled study of the effects of high-dose encapsulated omega-3 fatty acids on ventilatory and inflammatory measures of asthma controlling diet and other asthma risk factors. This review was limited because Meta-analysis was considered inappropriate due to missing data; poorly or heterogeneously defined populations, interventions, intervention-comparator combinations, and outcomes. In addition, small sample sizes made it impossible to meaningfully assess the impact on clinical outcomes of co-variables. Last, few significant effects were found.
## Competing interests
## Financial competing interests
Authors of this manuscript (and corresponding review) have not received any reimbursements fees, funding or salary form organizations that may in anyway gain or lose financially from the publication of this manuscript in the past five years prior to start of the corresponding review.
Authors do not hold any stocks or shares in an organization that may in any way gain or lose financially from the publication of this manuscript.
Authors do not hold or are currently applying for any patents relating to the content of the manuscript, nor they have received reimbursements, fees, funding, or salary from an organization that holds or has applied for patents relating to the content of the manuscript.
## Non-financial competing interests
Authors have no non-financial competing interests (political, personal, religious, ideological, academic, intellectual, commercial or any other) to declare in relation to this manuscript.
[table] Table 1: RCT Evidence of Omega-3 Fatty Acids to Improve Respiratory Outcomes in Asthma [/table]
|
Neoadjuvant apatinib plus S-1 in locally advanced pulmonary adenocarcinoma
Rationale: About one-third of the lung tumors are staged as locally advanced at the time of initial diagnosis; however, the optimal induction treatment before curative resection has not been elucidated. To date, the evidence regarding the preoperative apatinib plus S-1 for locally advanced pulmonary adenocarcinoma is scarce.Patient concerns: A 29-year-old female was admitted because of persistent cough, sputum, and chest distress for 2 months.Diagnoses: Primary pulmonary adenocarcinoma (cT3N2M0, IIIB) with unknown driver gene mutation status.Interventions: The patient had received 4 months of neoadjuvant therapy using oral apatinib (425 mg daily) plus S-1 (60 mg, twice daily for 4 weeks with a 2-week drug-free interval), followed by anatomical lobectomy with curative intent. Adjuvant apatinib (425 mg daily for a month, and 250 mg daily for another month) plus S-1 at the same dosage were administered for 2 months. Thereafter, maintenance of low-dose S-1 monotherapy (40 mg, twice daily for 4 weeks with a 2-week drug-free interval) was continued for 6 months.Outcomes: The adverse events were tolerable and well-controlled. A postoperative recurrence-free survival for 2 years and a half up to now was indicated.Lessons: Preoperative apatinib plus S-1 showed efficacy in locally advanced pulmonary adenocarcinoma. However, high-quality trials are warranted before the recommendation of this therapeutic regimen.Abbreviations: AEs = adverse events, CT = computed tomography, ECOG = Eastern Cooperative Oncology Group, NSCLC = nonsmall cell lung cancer, ORR = objective response rate, PFS = progression-free survival, RECIST 1.0 = response evaluation criteria in solid tumors version 1.0, TKIs = tyrosine kinase inhibitors, VEGFR = vascular endothelial growth factor receptor.
# Introduction
Lung cancer is the most commonly diagnosed cancer (11.6% of the total cases) and the leading cause of cancer death (18.4% of the total cancer deaths).The optimal management including neoadjuvant and adjuvant therapy for stage IIIA/N2 nonsmall cell lung cancer (NSCLC) is yet to be elucidated in the era of targeted therapy and immunotherapy. A network meta-analysis shows that neoadjuvant chemotherapy followed by surgery and adjuvant chemotherapy or radiotherapy has the greatest possibility to be the optimal regimen with the best overall survival and fewest treatment-related deaths for stage IIIA-N2 NSCLC.Apatinib, an oral tyrosine kinase inhibitor targeting vascular endothelial growth factor receptor-2, is effective for a broad range of solid tumors. S-1, an oral anticancer fluoropyrimidine derivative, is active and well tolerated as monotherapy for previously treated, advanced (clinical stage IIIB-IV) or relapsed NSCLC.S-1 monotherapy has demonstrated marked activity against NSCLC as well as gastric, colorectal, breast, cervical, and pancreatic cancers.First-line S-1, carboplatin, and antiangiogenetic bevacizumab followed by maintenance S-1 and bevacizumab had been reported to be active in advanced nonsquamous NSCLC.On the contrary, another trial revealed that the addition of bevacizumab to S-1 was not beneficial for patients with previously treated nonsquamous NSCLC.Therefore, it is important to clarify the most suitable agents for use with S-1 and the optimal timing of targeted therapy for lung cancer.
To the best of our knowledge, the available evidence regarding the application of apatinib plus S-1 for locally advanced pulmonary adenocarcinoma is rare. We herein presented a case of locally advanced pulmonary adenocarcinoma in which partial response was indicated after oral apatinib plus S-1 as induction therapy.
## Case presentation
In December 2016, a 29-year-old female nonsmoker was admitted for persistent cough, sputum, and chest distress for 2 months, without hemoptysis, hoarseness, chest pain, or significant loss of body weight. Her previous medical history was unremarkable. The Eastern Cooperative Oncology Group (ECOG) performance status was 0. Chest x-ray on admission revealed a mass in left lower lobe . In addition, laboratory tests showed elevated serum carcinoembryonic antigen (CEA), neuron-specific enolase (NSE), and cytokeratin-19 fragment (CYFRA 21-1). Further computed tomography (CT) indicated an irregular tumor measuring 70 mm  60 mm in size and enlarged mediastinal lymph nodes.
Bronchoscopic biopsy and pathological stain revealed the diagnosis of primary pulmonary adenocarcinoma. Distal metastasis was excluded by contrast-enhanced abdomen CT, cranial magnetic resonance, and whole-body bone emission CT. Then this case was staged as cT3N2M0, IIIB according to the 8th edition of TNM staging system for lung cancer.Meanwhile, the patient refused hospitalization and the standard first-line intravenous pemetrexed and carboplatin for personal reasons. However, genetic testing for the mutation status of epidermal growth factor receptor (EGFR), anaplastic lymphoma kinase, and programmed cell death protein 1 was not performed because it was not covered by her health insurance. Thus, EGFR-targeted agents or immunotherapy were not considered as the first therapeutic option. After a multidisciplinary evaluation, oral apatinib (425 mg daily) plus S-1 (120 mg per day for 4-week and 2-week withdrawal as her body surface area was > 1.5 m 2 ) was administered. During the induction therapy in outpatient clinic, CT and laboratory tests for serum CEA, NSE, and CYFRA21-1 were conducted regularly for efficacy evaluation according to Response Evaluation Criteria in Solid Tumors (RECIST 1.1), and the adverse events (AEs) were recorded in accordance with the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.0.
Encouragingly, the pulmonary adenocarcinoma indicated partial remission after 1 month of preoperative apatinib plus S-1 and stable disease during the next 3 months of the medical treatment . Similarly, the serum CEA, NSE, and CYFRA21-1 were decreased steadily. Grade 3 anemia, anorexia, hand-foot syndrome, and oral mucositis were observed and controlled effectively. No grade 4 toxicities were recorded during the neoadjuvant therapy.
On April 9, 2017, the pulmonary adenocarcinoma was slightly enlarged but it was considered to be resectable. Therefore, salvage lobectomy using fast-track protocol was scheduled as her ECOG score was 0. Anatomical lobectomy and mediastinal lymph node dissection was performed on April 13, . Chest x-ray and CT images of the pulmonary tumor during the induction treatment. A, X-ray on admission showed a bulky mass located in the left lower lobe. B, CT showed that the mass was 70 mm  60 mm in size. C, One month after oral apatinib plus S-1, the tumor indicated partial remission (PR) measuring 43 mm  54 mm with a necrotic cavity. D, Two months after induction therapy, the tumor showed stable disease (SD) measuring 44 mm  37 mm. E, The lesion was 41 mm  40 mm in size 3 months after treatment and SD was indicated. F, Four months after, the tumor was slightly enlarged measuring about 41 mm  42 mm before surgery. CT = computed tomography. 2017. Pulmonary vein-first approach was utilized during the surgery, with the aim of diminishing the risk of intraoperative tumor dissemination.Prophylactic ligation of the thoracic duct was not performed. A 26 French tube was used for chest drainage. The operating time was 75 minutes, with estimated blood loss of nearly 100 mL. Ultrasound-guided serratus anterior plane block was applied for pain relief. Her postoperative course was mainly uneventful, and she was discharged 5 days after surgery. R0 resection was achieved, and the maximal diameter of the tumor in specimen was 60 mm  35 mm. The pathological diagnosis was poorly differentiated lung adenocarcinoma with visceral pleura invasion (pT3N0M0, IIB).
Three weeks after the operation, adjuvant apatinib (425 mg daily for a month, and 250 mg daily for another month) plus S-1 (120 mg daily for 4 and 2-week withdrawal) were administered for 2 months, and then the apatinib was discontinued due to grade 3 hand-foot syndrome and grade 4 elevated serum aspartate aminotransferase and alanine aminotransferase. Thereafter, low-dose S-1 (40 mg, twice daily at the same schedule) as maintenance therapy was continued for another 6 months, although local recurrence or distant metastasis of pulmonary adenocarcinoma was not observed during the follow-up. The serum biomarkers of CEA, NSE, and CYFRA 21-1 were also in normal range. This patient demonstrated a PFS of 2 and a half years up to November 2019.
# Discussion
To the best of our knowledge, this is the first case report of firstline induction apatinib plus S-1 for locally advanced pulmonary adenocarcinoma, and this treatment regimen showed a promising effect on survival of the patient.
It is reported that one-third of the NSCLC patients are found to have locally advanced tumors at the time of initial diagnosis,and neoadjuvant therapy followed by surgery and adjuvant therapy might be the optimal treatment.Down-staging of primary tumor and/or mediastinal lymph nodal metastases after induction therapy are positive prognostic factors in selected patients.Another study indicates that there is a nonsignificant difference between the outcomes of neoadjuvant and adjuvant chemotherapy for IIIA NSCLC patients.Neoadjuvant tyrosine kinase inhibitor (TKI) erlotinib is well tolerated and might improve the resection rate of stage IIIA-N2 EGFR mutationpositive NSCLC, and the next-generation sequencing could be utilized to predict outcomes in these patients.Moreover, large neoadjuvant trials of immunotherapies and targeted therapies in advanced disease are underway.The optimal adjuvant therapy for clinical N2 NSCLC patients who undergo neoadjuvant chemotherapy/immunotherapy and surgery has not been elucidated.Brandt et alreport that neoadjuvant or adjuvant chemotherapy is not associated with an improvement in overall survival or PFS among patients with cT2∼4N0∼1M0 NSCLC after radical surgery. Another population-based study shows that patients with cN2 disease but postchemotherapy surgical nodal staging ypN0∼1 and/or lymph node ratio (LNR, which is defined as number of lymph nodes involved by tumor divided by total number of dissected nodes) < 15% do not benefit from adjuvant therapy, whereas the patients with persistent N2 disease and LNR > 15% who receive adjuvant chemoradiotherapy have improved survival,which indicates that aggressive therapy is beneficial to the patients with persistent or high nodal burden disease.
S-1 plus cisplatin in combination with radiotherapy results similar efficacy but better hematological tolerability (lower risk of leukocytopenia and neutropenia) as compared with standard concurrent chemoradiation regimens in locally advanced NSCLC.In addition, S-1 as a third-or fourth-line therapy for wild-type EGFR NSCLC demonstrates numerically better clinical outcomes than erlotinib.Furthermore, postoperative S-1 for 1 year seems feasible for stage IB-IIIA lung cancer with low incidence of AEs.However, there is no consensus regarding the benefit of S-1 maintenance therapy for squamous cell lung cancer. Apatinib has shown survival benefit in NSCLC trials with a favorable AEs profile.The previously reported studies of S-1 plus TKIs or bevacizumab for lung cancer are listed in .First-line S-1, carboplatin, and bevacizumab followed by maintenance S-1 and bevacizumab are active for advanced nonsquamous NSCLC.S-1 plus bevacizumab produces modest survival efficacy in second-line treatment for advanced nonsquamous NSCLC.First-line S-1 plus cisplatin with bevacizumab, and pemetrexed plus cisplatin with bevacizumab have similar activity and tolerability in patients with advanced nonsquamous NSCLC.Nevertheless, other trials show that S-1 plus bevacizumab does not provide any additional benefit in terms of PFS for nonsquamous NSCLC patients after failure of platinum-based chemotherapy.As for the researches regarding S-1 plus TKIs, a phase II trial shows that first-line concurrent carboplatin, S-1, and gefitinib is efficacious in advanced EGFR mutation-positive NSCLC patients.S-1 plus EGFR-TKIs shows synergistic efficacy in stage IIIB-IV NSCLC patients who have experienced prior EGFR-TKI failure because of acquired resistance.Another trial indicates that S-1 plus gefitinib is effective in EGFR mutationpositive pulmonary adenocarcinoma.Based on available reported studies and the presented case, targeted therapy in combination with S-1 might be an alternative option for locally advanced NSCLC. However, well-designed trials for convincing evidence are warranted before the implementation of TKIs or antiangiogenesis agents plus S-1 into therapeutic guideline. The registered trials evaluating the efficacy of TKIs such as gefitinib, anlotinib, and antiangiogenetic agents including bevacizumab and apatinib plus S-1 for lung cancer are summarized in .
From this case, there are several questions arise: Is there any reliable efficacy indicators of apatinib plus S-1 for patient selection? How to determine the optimal duration of induction therapy using antiangiogenetic agents? Is adjuvant apatinib plus S-1 necessary for pulmonary adenocarcinoma after R0 resection with ypN0 status (how to avoid over-treatment)? Similarly, did this patient really benefit from the 6-month maintenance therapy using S-1? The role of targeted agents needs to be validated in the era of immunotherapy. The reported clinical trials evaluating the efficacy of S-1 plus bevacizumab or TKI for lung cancer.
## First author (year)
Types of cancer Nishijima-Futami (2017)Advanced nonsquamous NSCLC 30 S-1+ bevacizumab Second-line ORR: 6.7% Anorexia (10%); infection (10%); diarrhea (6.7%) Yang (2018)Stage In summary, apatinib plus S-1 showed efficacy in locally advanced pulmonary adenocarcinoma. However, high-quality evidence is needed.
# Author contributions
Conceptualization: Chu Zhang, Dong Liu. The registered trials evaluating the efficacy of targeted agents plus S-1 for lung cancer.
## Registration identifier diseases
|
Flavivirus induces MHC antigen on human myoblasts: A model of autoimmune myositis?
Infection of human embryonic myoblasts by West Nile virus (WNV), a flavivirus, caused significant upregulation of class I and II MHC expression as determined by flow cytometry. After 48 hours at a multiplicity of infection of 5 pfukell, a sixfold increase in MHC class I expression was induced from initially low levels of expression. In contrast, MHC class II was induced de novo to five times the control fluorescence level. At least 70% of the cells were infected as determined using fluorescence microscopy and anti-WNV antibody labeling. Myoblasts were >90% pure as shown by anti-Leu-19 labeling. MHC class I (but not class II) was increased threefold after exposure to virus-inactivated supernatant from 48-hour-infected cells, indicating the presence of factor(s) contributing to the MHC class I increase. These findings may be important in establishing a link between viral infection of human cells and induction of inflammatory autoimmune disease. We discuss the possibility of using WNV as an in vivo model. 0 1992 John Wiley & Sons, Inc.
Major histocompatibility complex antigens (MHC class I and 11) are transmembrane glycoproteins. They bind viral antigen for recognition by cytotoxic (CD8+) T cells through MHC class I, and for presentation to T-helper (CD4+) cells through MHC class 11. The ensuing cellular immune response consists of the controlled, specific eradication of virus-infected cells,41 followed by suppression of this response. However, in certain individuals, autoimmune disease is thought to supervene.
Several different organs may be involved simultaneously in autoimmune attack, but in some diseases only a relatively narrow range of tissues may be affected. In the latter category, diseases such as polymyositis have been extensively investigated. Much work has uncovered evidence of humoral involvement in poly~nyositis,~~~ 1832 but during the last decade it has become increasingly clear that cellular immunity may play a significant part in disease pathogenesis. Thus, increases in MHC class 1' ' and 113* on muscle tissue have been found, as well as local infiltrations of CD4+ and CD8' T cells.',*,' Significantly, antecedent virus infection is also thought to be an important factor in the precipitation of p o l y m y o~i t i s .~~~~'~~~~~~~ Most viruses reduce cell surface MHC ex ression, either specifically or nonspecifically, 10,zK34 although some, for example, neurotropic coronaviruses, have been found to increase MHC expression on oligodendrocytes, a s t r~c y t e s "~~~~~~ and brain endothe-Viruses which induce MHC expression on normally MHC-negative cells would presumably be candidate etiologic agents in autoimmune polymyositis as well as other autoimmune diseases.
Recently, it has been found that flaviviruses increase MHC class I on fibroblastsz6 and class I1 on astrocytesz9 with little interference with cell metabolism for the first 24-48 hours. West Nile virus (WNV) is an arborvirus belonging to the family Flauiuiridae. Transmitted by mosquitoes, this hu-man pathogen causes significant morbidity and mortality throughout the world.
We report here that WNV infection of cultured human embryonic myoblasts increases MHC class I expression and induces de novo expression of MHC class I1 on these cells. We suggest that inappropriate expression and presentation of class I and 11 MHC by normally negative myoblasts/myotubes, particularly in the context of virus infection, could augment the pathogenesis of autoimmune disease. This virus may thus be useful as a tool to create a laboratory model for autoimmune muscle disease.
# Materials and methods
## Myoblast infection by wnv and wnv supernatant
Treatment. Human embryonic myoblasts, isolated from 14-17 week normal human embryos, were prepared as previously d e~c r i b e d ,~ and cultured for l week in Ham's F10 medium (CSL #733 2049) containing 10% fetal calf serum (HF1 O/FCS) under standard culture conditions (SCC). NH and MRC guidelines were strictly adhered to throughout. Myoblasts were then seeded into 50-mm plastic Petri dishes at 5 x lo5 cells/ dish. The cells were infected 24 hours later with a multiplicity of infection (moi) of 5 plaque-forming units (pfu) of' WNV/cell for 48 hours. The virus concentration and the time of infection was titrated to give maximum response for both class I and I1 MHC induction with minimum cytotoxicity . Infected cells were incubated under SCC for 48 hours. Control cells were mock-infected using HFIO/FCS only. A control for MHC induction consisted of myoblasts incubated with IFN-y (100 U/mL) for 48 hours. This was titrated for maximal MHC induction .
Supernatants from WNV-infected cultures were irradiated with UV light at 1600 pW/cm2 for 12 minutes. This destroys WNV without interfering with the activity of I F N s .~~,~' Supernatants were added to fresh myoblast cultures for 48 hours under SCC. Supernatants from mock-infected cultures were divided into two. Fresh WNV at 5 pfu/cell was added to half, the other was left untreated. Both were irradiated as above, and added to fresh cultures as controls.
# Results
## Mhc expression on normal embryonic myoblasts.
Previously, using quantitation by fluorescence microscopy, we reported low MHC class I expression and no class I1 expression on cultured normal human embryonic myoblast~.~ Here we provide a more accurate quantitation of MHC expression using FCM. [fig_ref] FIGURE 3: Flow cytometric analysis of MHC expression by human embryonic myoblasts [/fig_ref] illustrates the distribution of MHC class I expression on normal human myoblasts (b), compared with that of the isotype control antibody (anti-GFAP) (a). The difference in fluorescence is about tenfold. Thus, this level is not as [fig_ref] FIGURE 3: Flow cytometric analysis of MHC expression by human embryonic myoblasts [/fig_ref] shows the distribution of MHC class I1 expression on normal human myoblasts (b), compared with that of the isotype control antibody (a). In agreement with previous finding^,^ this demonstrates that there is no detectable MHC class I1 expression on these cells.
low as was previously thought. 7,22,31
## Effect of wnv on mhc expression by myoblasts.
MHC class 1 and I1 expression was measured by FCM on 48-hour WNV-infected (5 pfukell), mock-infected, and IFN-y-treated myoblasts. [fig_ref] FIGURE 3: Flow cytometric analysis of MHC expression by human embryonic myoblasts [/fig_ref] shows a sixfold increase in MHC class I expression by infected myoblasts (b) compared with mock-infected myoblasts (a). MHC class I expression on IFN-y-treated myoblasts (c) was almost 10-fold higher than on mock-infected cells. [fig_ref] FIGURE 3: Flow cytometric analysis of MHC expression by human embryonic myoblasts [/fig_ref] illustrates the de novo induction of MHC class I1 expression after WNV infection (b). The fluorescence intensity is about fivefold greater than mock-infected myoblasts (a), while the IFN-ytreated group (c) showed a 16-fold increase in fluorescence. This demonstrates that the capacity for MHC class I1 induction on myoblasts is significantly greater than previously reported. This may in part reflect differences in technique3 andlor tissue origin.lg
## Effect of supernatant from infected myoblasts on mhc expression by myoblasts.
To investigate whether soluble factor(s) secreted by infected myoblasts could upregulate MHC expression, UVirradiated supernatants from WNV-or mockinfected cultures were added to fresh myoblasts for 48 hours. [fig_ref] FIGURE 3: Flow cytometric analysis of MHC expression by human embryonic myoblasts [/fig_ref] shows that MHC class I expression was increased threefold (c), while the mock-infected control groups, either with (b) or without (a) UV-irradiated WNV, were unaffected. Thus, there was no apparent influence of irradiation breakdown products on MHC class I expression. The IFN-y-treated group (d) shows an 1 1-fold higher expression than mock-infected m yoblasts. [fig_ref] FIGURE 3: Flow cytometric analysis of MHC expression by human embryonic myoblasts [/fig_ref] demonstrates that there is no increase in MHC class I1 expression in the presence of irradiated supernatant (c), mock-infected with (b) and without (a) UV-irradiated WNV. The IFN-y-treated group (d) shows a fluorescence intensity greater than 20-fold that of non-UVirradiated WNV (a).
## Identification of myoblasts.
The purity of myoblast cultures is shown in [fig_ref] FIGURE 3: Flow cytometric analysis of MHC expression by human embryonic myoblasts [/fig_ref]. More than 90% of cells were positively labeled by the myogenic marker, antihuman Leu-19 MAb (b), compared with an isotype antibody control (a).
## Detection of wnv infection in myoblasts.
Infection of myoblasts on coverslips was detected using anti-WNV antibody, labeled with SAMIg-FITC and fluorescence microscopy. Infection was detectable in at least 70% of myoblasts in all experiments as a bright specific perinuclear fluorescence [fig_ref] FIGURE 4: Fluorescence micrographs of WNV-infected and mock-infected myoblasts [/fig_ref] compared with uninfected controls [fig_ref] FIGURE 4: Fluorescence micrographs of WNV-infected and mock-infected myoblasts [/fig_ref]. We report here that infection of human myoblasts by WNV significantly upregulates MHC class I and induces de novo class I1 expression. This was shown to be due, at least in part, to virusinduced secreted but unidentified soluble fact o r~, '~,~' since the virus-inactivated supernatants from WNV-infected myoblast cultures were able to induce increased MHC class I expression in fresh myoblast cultures. This may also be due to a direct virus effect.26 In contrast, MHC class I1 on myoblasts may be induced by the virus directly, as there was no induction of MHC class I1 in response to the virus-inactivated supernatant. The mechanism of this apparent direct induction by the active WNV is presently unknown. We think it may relate to the way some protein product(s) of viral transcription interact with the host cell. Investigation into possible mechanisms is presently being undertaken. Interestingly, and by contrast, UV-inactivated nonreplicating neurotropic coronavirus particles appear to be able to induce class [bib_ref] Monoclonal antibody analysis of mononuclear cells in myopathies. 11: Phenotypes of autoinvasive..., Engel [/bib_ref] MHC in a small subpopulation of astrocytes in ~i t r o .~' This suggests different mechanisms of induction of class I1 by these two viruses and, if true, would clearly have different implications for the respective antiviral immune responses.
As the concentration of MHC is directly related to the efficiency of both induction and execution of the cellular immune r e~p o n s e , '~.~~ increases in MHC expression due to IFN-y or virus (including WNV) are accompanied by increased lysis of target cells b virus-immune and alloimmune Tc cells. 5212,2432J In autoimmune muscle disease, long speculated to have a viral etiology, increased MHC class 1" and class Il3* expression on muscle fiber membranes is frequently found in biopsies, and is associated with an influx of CD8+ and CD4+ T Evidence suggests that CD8+ cells may mediate lysis of muscle cells." However, although there is an association in inflammatory muscle disease with several different viruses (e.g., retroviruses,8*"8 Coxsackie virus, l 7 pic o r n a v i r u~, '~ and infl~enza'~), direct viral involve-ment in the pathogenesis has been difficult to prove.
Hypothetically, viral infection resulting in a virus-specific cellular immune response would induce the release of IFN-y by T cells35 and thus induce an increase in MHC expression on local uninfected muscle cells.20 In susceptible individuals, this could facilitate crossreactive killing of uninfected muscle cells by antiviral T cells. Moreover, the induction of aberrant MHC expression might break self-tolerance and elicit an MHCrestricted antimuscle T-cell r e s~o n s e .~ Subsequent IFN-y induction of MHC by such autoimmune T cells would act as a positive feedback in this scenario.
Most viruses, however, in reducing cell surface MHC expression, would preferentially encoura e the formation of high-affinity antiviral Tc cells. These undoubtedly kill more efficiently if there is increased MHC expression on target cells that are infected.5724933 Nevertheless, the high affinity of these Tc cells for virus-infected target cells would not usually result in crossreactive killing of uninfected cells despite the increased MHC expression of the latter. In theory, viruses which increase MHC expression in infected cells in vivo would induce an antiviral cellular immune response consisting mostly of low-affinity Tc cells with a high degree of crossreactivit for MHC alone. This occurs with WNV in vitro.2 In addition, a tissue-specific, MHC-restricted response might simultaneously be induced in susceptible individuals in vivo. Such viruses would thus be candidate etiologic agents in autoimmune polymyositis and other autoimmune disease.
For reasons of hosthirus survival, virusinduced autoimmunity cannot be a common event, although the reasons for this are not well understood. Nevertheless, under certain conditions, particularly in geneticall susceptible individuals (e.g., HLA-B8 carriers *), it seems likely that autoimmune muscle disease could develop subsequent to such a viral infection. Ideally, we would like to be able to demonstrate a virus-specific and allospecific Tc-cell response in the human myoblast system reported here. However, although practicable in the murine system, there are severe ethical problems associated with doing this in the human embryonic system.
In conclusion, these findings suggest the use of WNV infection as a possible model for autoimmune muscle disease development in mice. Our preliminary findings suggest that murine embryonic myoblasts infected with WNV behave in the 1657 s 7 .
1276 WNV Increases Myoblast MHC same way as human myoblasts. It seems reasonable to suggest that investigation of such an animal model would shed useful light on events related to virus-induced autoimmune muscle disease pathogenesis in humans.
[fig] FIGURE 1 FIGURE 2: Titration of WNV multiplicity of infection on myoblasts. (A) MHC class I is maximally induced by 5 pfu/cell over a period of 24 hours (A-A), 48 hours (0-o), and 72 hours (0-0). (B) MHC class II is maximally induced by 5 pfukell over a period of 24 hours (A-A) and 72 hours (0-n), but maximally induced by 10 pfukell over 48 hours (0-0). The y-axis represents the peak channel fluorescence from FCM analysis of the myoblast population labeled for MHC class I (A) and MHC class II (6). The x-axis represents the WNV moi measured in pfukell. FCS. The cells were then incubated at 4°C for 45 minutes with 25 pL of a 1 :20 dilution of polyclonal anti-WNV antibody and washed again. WNV antibodies were labeled using 25 pL of 1 : 20 dilution of fluorescein isothiocyanate-conjugated sheep antimouse immunoglubulin (SAMIg-FITC) (Silenus). Hoechst 33342 (Sigma) was used at I pg/mL, 25 pL/coverslip, as a counterstain. The percentage of WNV-infected cells was calculated as a ratio of FITC-labeled (infected) cells to Hoechst-labeled nuclei x 100. Cells with double nuclei were counted as single cells. Preparation of Myoblasts for Flow Cytometry (FCM) Labeling. Myoblasts in Petri dishes were washed twice in CMF-HBSS. One milliliter of trypsin 30 c ward and side scatter measurements were within 25 -...................... Q"tlOl........... .......................... Titration of MHC class I and II induction by IFN-y on myoblasts. Class I (0--0) is maximally induced by 25-50 U/mL IFN-y, while class II (0-0) is maximally induced by 70-125 U/mL IFN-y. The y-axis represents the peak channel fluorescence from FCM analysis of the myoblast population labeled for MHC class I and II. The x-axis represents the concentration of IFN-y measured in units per milliliter. [/fig]
[fig] Log: Fluorescence (a.u.) [/fig]
[fig] FIGURE 3: Flow cytometric analysis of MHC expression by human embryonic myoblasts. (A) MHC class I antigen expression (b) is tenfold higher than isotype control antibody (a). (B) No difference between anti-HLA-DR antibody (b) and isotype control antibody (a). Thus, normal myoblasts express little or no MHC class II antigen. (C) Expression of MHC class I in myoblasts infected with WNV (b) is approximately sixfold higher than in mock-infected myoblasts (a), while expression in the IFN-?-treated myoblasts (c) is about tenfold higher. (D) Expression of MHC class II in WNV-infected myoblasts (b) is about fivefold higher than control levels of mock-infected myoblasts (a), while expression in the IFN-y-treated myoblasts (c) is nearly 16-fold higher. (E) Expression of MHC class I on myoblasts cultured in UV-inactivated supernatant from WNV-infected myoblast cultures (c) is more than threefold higher than myoblasts cultured in the supernatant from mock-infected with (b) and without (a) UV-inactivated fresh WNV. Myoblasts treated with IFN-y (d) show about a tenfold higher expression than (a) and (b). (F) Expression of MHC class II on myoblasts cultured in UV-inactivated supernatant from WNV-infected myoblast cultures (c) is similar to the expression on myoblasts cultured in the supernatant from mock-infected with (b) and without (a) UV-inactivated fresh WNV. Myoblasts treated with IFN-.I (d) show a greater than 20-fold higher expression than (a), (b), and (c). (G) More than 90% of cells are positively stained with antihuman Leu-I9 directly conjugated to phycoerythrin (PE) (b) compared to an irrelevant antibody control (a). The abscissas represent log fluorescence intensity (arbitrary units). The ordinates represent [/fig]
[fig] FIGURE 4: Fluorescence micrographs of WNV-infected and mock-infected myoblasts. (A) Absence of viral antigen fluorescence in mock-infected myoblasts stained with anti-WNV antibody. (6) Distribution of WNV antigens recognized by anti-WNV antibody in WNV-infected myoblasts. White bar = 50 Fm. [/fig]
[table] 0: 05%)-EDTA (1 mM) was added per dish and incubated at 24°C for 2 minutes, then 1 mL of HFlO/FCS was added to stop the digestion. The myoblasts were centrifuged at 4°C at 4009, resuspended in HFlO/FCS, and distributed into sample aliquots for labeling. of MHC Class I and Class II. Cells were incubated with 100 pL of a 1 : 20 dilution of monoclonal antibody (MAb) recognizing MHC class I (clone W6/32, Dakopatts) or class I1 (clone CR 3/43, Dakopatts) at 4°C for 45 minutes. Concentrations were titrated for maximal specific labeling. After incubation, myoblasts were centrifuged through 100 pL FCS, the supernatant was removed, and the cells resuspended in 100 pL of 1 :20 SAMIg-FITC for 45 minutes. Cells were then centrifuged through FCS as above and resuspended in 300 )IL HFlOIFCS for FCM. An isotype antibody recognizing glial fibrillary acidic protein (GFAP-Dako, Dakopatts) was used to control for nonspecific antibody labeling. [/table]
|
Evolution of SARS-CoV-2 Envelope, Membrane, Nucleocapsid, and Spike Structural Proteins from the Beginning of the Pandemic to September 2020: A Global and Regional Approach by Epidemiological Week
# Introduction
Coronavirus disease (COVID-was detected for the first time, in Wuhan, China, in December 2019. In January 2020, the responsible virus, acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was isolated and the complete viral genome was sequenced. Since then, SARS-CoV-2 has spread throughout the world in this ongoing pandemic. SARS-CoV-2 is a ß-coronavirus belonging to the Coronaviridae family, order Nidovirales. Coronaviruses (CoVs) are enveloped positive-sense RNA viruses with a large and nonsegmented genome of ∼30 kb length including five major open reading frames (ORFs), i.e., ORF1a, ORF1ab, and the four structural proteins plus a number of accessory genes. The first two overlapping ORFs (ORF1a and ORF1b) are located at the 5 end of the viral RNA, occupying about two-thirds of the genome, and encode proteins which are autoproteolytically processed into 16 non-structural proteins (nsp), which are involved in viral RNA replication and transcription. Interactions among several SARS-CoV-2 nonstructural and human proteins have been highlighted. The 3 end of the viral genome worldwide sequences collected from the Global Initiative on Sharing All Influenza Data (GISAID) from December 2019 toSeptember 2020. For this purpose, we used an in-house bioinformatics tool developed in our laboratory designed for genetic variability analysis of pathogens and proteins with biological or biomedical interest. We identified the most prevalent aa changes at a global, regional, and local level, and highlighted any changes with an increasing frequency in time or located in protein regions of special biological or structural interest.
# Materials and methods
All available complete and partial SARS-CoV-2 human genomic sequences deposited in the GISAID database (https://www.gisaid.org/) until 19 September 2020 were downloaded in nucleotides (nt) and classified according to the country of origin and to the epidemiological week (epiweek) by collection date. Epiweeks are a standardized method of counting weeks to allow for the comparison of data. By definition, the first epiweek of the year ends on the first Saturday of January, as long as it falls at least four days into the month. Each epiweek begins on a Sunday and ends on a Saturday. The analyzed SARS-CoV-2 sequences were deposited in GISAID from 29 December 2019 (epiweek 52, 2019) to 19 September 2020 (epiweeks 1 to .
For sequences analysis, we used an in-house bioinformatics tool previously designed and used in our laboratory for HIV genetic variability analysis and recently updated for SARS-CoV-2 sequences study. This tool is programmed in JAVA OpenJDK version 11.0.9.1 using IDE NetBeans version 12.2. Functions related to protein tracking, cutting, and aligning were tested with Mega X, and functions related to aa change identification were tested manually, using Excel 2019 version 19.0. With this program, the complete nt sequences from the four structural viral proteins were cut, aligned, and translated into the following amino acids (aa): spike or S (1273 aa), nucleocapsid or N (419 aa), membrane or M (222 aa), and envelope or E (75 aa). Wuhan SARS-CoV-2 was taken as the reference sequence (NCBI accession number NC 045512.2) to identify the aa changes in these proteins. The program detects any aa different from the reference sequence for each aa position and calculates the number and frequency of aa changes for that position. Nonsense mutations, gaps, and unknown amino acids (probably due to the low quality of some regions of the original sequences, failing to attribute a nucleotide with certainty) were not considered to calculate the mutation rate. This method allows the analysis of partial or low-quality genomes as long as the residue of the studied position is present enabling a much larger set of sequences to be studied. The statistical average of the aa changes was comparatively analyzed between epiweeks and regions.
To detect any emerging aa change, sequences were analyzed all together (global analysis) and by geographic region (regional analysis), establishing the following six regions of interest according to GISAID classification: Africa, Asia, Europe, North America (including Central America and the Caribbean), South America, and Oceania. Sequences were also analyzed by epiweeks to detect significant changes in time (increase or decrease of mutation rate). Since some SARS-CoV-2 sequences in GISAID only included the sampling year or month but not the complete date (day/month/year), these were used in the global and regional analysis but not in the epiweek approach. These sequences were 1127 in E (172 from Oceania, 25 from North America, 70 from Asia, and 860 from Europe), 1136 in S (67 from Asia, 848 from Europe, 26 from North America, and 168 from Oceania), 1146 in M (70 from Asia, 873 from Europe, 26 from North America, and 177 from Oceania), and 1102 in N (50 from Asia, 853 from Europe, 25 from North America, and 174 from Oceania). The number of sequences available for each region and epiweek was not evenly distributed, especially in the last epiweeks. For this reason, and to avoid overestimation of mutation rates due to one particular country or epiweek, selected mutations were analyzed individually by country of origin and epiweek, with only epiweeks with at least 10 available sequences being considered for the time evolution analysis. For the same reason, the global analysis must be considered to be the analysis of the whole sequence dataset, taking into account the different proportions of the sequences according to the region of origin. To assess the significance of the aa changes in positions 203 and 204 of the Nucleocapsid protein, we performed an ANOVA test and Sidak test for subsequent pairwise comparisons. To assess the effects of time on the main aa combinations of these positions, we performed an exponential linear regression. Both tests were performed using Stata 16.1. using the following formula Y = b0 × (eˆ(b1 × X)) OR ln(Y) = ln(b0) + (b1 × X).
Location of aa changes along the protein structure domains was done according to UniProtKB (https://www.uniprot.org) and RCSB Protein Data Bank (https://www.rcsb. org) annotation.
# Results
A total of 105,276 worldwide SARS-CoV-2 partial and complete sequences from 117 countries were downloaded from GISAID corresponding epiweek 52, 2019 and epiweeks 1 to 37, 2020. After bioinformatics processing, we recovered and studied the genetic variability of 101,100 spike, 101,376 envelope, 103,419 membrane, and 99,675 nucleocapsid complete sequences. The epiweeks with available sequences varied across the geographic regions established according to GISAID classification . Only Asia presented sequences in the 38 epiweeks under study, followed by Europe and North America (34 epiweeks each), Oceania (33 epiweeks), Africa (28 epiweeks), and South America (26 epiweeks). The number of analyzed sequences per protein and geographic region, the total number of mutated sequences, and the number, frequency, and nature of mutated residues in each protein are described in . The number of sequences available per country in each geographic region for each protein is listed in . The number of countries providing structural protein sequences in GISAID was 42 in Europe, 32 in Asia, 20 in Africa, 11 in North America (including Central America and the Caribbean), 9 in South America, and 3 in Oceania.shows the number of global and regional sequences with aa changes across S, E, M, and N SARS-CoV-2 complete structural proteins.shows the percentage of global sequences with aa changes for each residue across the four structural SARS-CoV-2 proteins and their location within the protein domains.
## Spike protein (s)
Global S aa conservation was 99.97%. Among the 101,100 analyzed sequences, 2671 (2.6%) aa changes were found in 1132 (88.9%) of 1273 spike residues. The most frequent aa change was D614G (81.5%, 82,183 sequences), located in S1, followed by S477N (4.1%), located in the receptor binding motif of the receptor binding domain. All other substitutions had a frequency below 1%. D614G was also the most frequent aa substitution in all regions as follows: Africa (94.7%), Asia (57.7%), Europe (82.9%), North America (84.3%), South America (93%), and Oceania (82%). D614G was found for the first time in epiweek 4 in Asia in two (1%) Chinese sequences, and in Oceania in one (11%) Australian sequence. In Europe, D614G appeared in epiweek 5 for the first time, mainly in Germany (41%), and in North America (31%), in three Canadian sequences. The last regions where this mutation appeared, in epiweek 9, were Africa (in three sequences from Nigeria, Senegal, and Morocco) and South America (in three sequences from Brazil). In epiweek 10, more than half (54.7%) of the total sequences showed this change, increasing to 97.9% in epiweek 37 . S477N in the S protein was present in all the geographic regions, but mainly in Oceania (56.8%, 3851 Australian sequences), where its frequency rose from 6% (epiweek 20) to 100% (epiweek 31). In the regional analysis, V1176F aa change stood out in South America (18.2%, 286 sequences), with all sequences but one belonging to Brazil. With a brown star, D614G, change from aspartic acid to glycine in residue 614 of spike protein, present in 81.5% of the global sequences, and located in S1 domain, before S1/S2 furin cleavage site. With an orange star, S477N, change from serine to asparagine in residue 477, present in 4.1% of the global sequences, and located in the receptor binding motif. In light blue, receptor binding domain (RBD). In dark blue, within the RBD, receptor binding motif (RBM); (B) Envelope protein (75 aa, 101,376 total sequences). With an orange star, S68F, change from serine to phenylalanine at position 68 of envelope protein, present in 0.2% of the global sequences. In light blue, PDZ-binding motif (PBM); (C) Membrane protein (222 aa, 103,419 total sequences). With an orange star, D3G, change from aspartic acid to glycine in residue 3, and T175 M, change from threonine to methionine in residue 75. D3G and T175M were present in 0.7% and 1% global sequences, respectively. In light blue, transmembrane domains (TM); (D) Nucleocapsid protein (419 aa, 99,657 total sequences). With a brown star, R203K, change from arginine to lysine in position 203 and G204R change from glycine to arginine in position 204. R203K and G204R were present in 37.3% and 37% of the global sequences, respectively, and located in the serine/arginine-rich (SR)- With a brown star, D614G, change from aspartic acid to glycine in residue 614 of spike protein, present in 81.5% of the global sequences, and located in S1 domain, before S1/S2 furin cleavage site. With an orange star, S477N, change from serine to asparagine in residue 477, present in 4.1% of the global sequences, and located in the receptor binding motif. In light blue, receptor binding domain (RBD). In dark blue, within the RBD, receptor binding motif (RBM); (B) Envelope protein (75 aa, 101,376 total sequences).
With an orange star, S68F, change from serine to phenylalanine at position 68 of envelope protein, present in 0.2% of the global sequences. In light blue, PDZ-binding motif (PBM); (C) Membrane protein (222 aa, 103,419 total sequences). With an orange star, D3G, change from aspartic acid to glycine in residue 3, and T175 M, change from threonine to methionine in residue 75. D3G and T175M were present in 0.7% and 1% global sequences, respectively. In light blue, transmembrane domains (TM); (D) Nucleocapsid protein (419 aa, 99,657 total sequences). With a brown star, R203K, change from arginine to lysine in position 203 and G204R change from glycine to arginine in position 204. R203K and G204R were present in 37.3% and 37% of the global sequences, respectively, and located in the serine/arginine-rich (SR)-linker. Color code as follows: white, 0% of sequences with aa changes; yellow, >0 to 0.1% of sequences with aa changes; light orange, >0.1 to 1% of sequences with aa changes; dark orange, >1 to 30% of sequences with aa changes; brown, >30% of sequences with aa changes. In light blue, serine/arginine-rich linker (SR-linker). SS, signal peptide; NTD, N-terminal domain; CTD, Cterminal domain; RBD, receptor binding domain; RBM, receptor binding motif; FP, fusion peptide; HR, heptad repeat; TM, transmembrane domain; PBM, PDZ-binding motif; SR-linker, serine/arginine-rich linker. Annotation according to UniProtKB (https://www.uniprot.org) and RCSB Protein Data Bank (https://www.rcsb.org).
Viruses 2021, 13, x FOR PEER REVIEW 7 of 18 . Global and regional frequency of D614G change in the spike protein over time. (A) Global and regional D614G frequency distribution in epidemiological weeks with, at least, 10 available sequences. The x-axis represents epidemiological week and the y-axis represents percentage of mutated sequences; (B) Global and regional number of sequences in spike's residue 614 harboring aspartic acid and glycine amino acids in epidemiological weeks with at least 10 available sequences. The x-axis represents epidemiological week and the y-axis represents the number of sequences harboring other aa than G, in grey color, and G, in blue color.
S477N in the S protein was present in all the geographic regions, but mainly in Oceania (56.8%, 3851 Australian sequences), where its frequency rose from 6% (epiweek 20) to 100% (epiweek 31). In the regional analysis, V1176F aa change stood out in South America (18.2%, 286 sequences), with all sequences but one belonging to Brazil.
## Envelope protein (e)
The global E aa conservation was 99.98%. Among the 101,376 E sequences analyzed, . Global and regional frequency of D614G change in the spike protein over time. (A) Global and regional D614G frequency distribution in epidemiological weeks with, at least, 10 available sequences. The x-axis represents epidemiological week and the y-axis represents percentage of mutated sequences; (B) Global and regional number of sequences in spike's residue 614 harboring aspartic acid and glycine amino acids in epidemiological weeks with at least 10 available sequences. The x-axis represents epidemiological week and the y-axis represents the number of sequences harboring other aa than G, in grey color, and G, in blue color.
## Envelope protein (e)
The global E aa conservation was 99.98%. Among the 101,376 E sequences analyzed, we found 142 aa changes in 65 (86.6%) positions of the 75 E residues. All mutations were extremely infrequent, present in less than 0.3% of the total sequences. The most prevalent aa change found in E was S68F, present in 221 (0.2%) global sequences, followed by L73F (122 sequences), R69I (92), P71L (68), T9I (56), and V62F (52), all with a frequency of 0.1%.
S68F was present in all regions except in South America, and mainly in Europe (86%, 177 sequences), specifically England (68.9%) where its frequency raised from epiweek 12 (0.6%) to epiweek 19 (3%), decreasing to 0.2% in the last epiweek available. This aa change first appeared in Oceania (epiweek 10), and later in Asia and Europe (epiweek 12), North America (epiweek 13), and Africa (epiweek 21). Most sequences with T9I, R69I, P71L, and L73F belonged to Europe. V62F was present in 70% of sequences from the USA in North America, where frequency increased in epiweeks 22 and 23 but dropped later.
In the analysis by geographic region, the mutation rate was less than 1% in all regions except in Africa, where V5F (1.5%) was present in 36 sequences (35 from Egypt), 92% of them belonging to the last epiweeks with African sequences available . In the analysis by epiweek, no steady increase over time globally or regionally was observed.
## Membrane (m) protein
The 103,419 M sequences analyzed presented 99.99% conservation, with 291 aa changes found in 165 (74.3%) positions of the 222 M aa. Most changes had a very low frequency (≤0.2%), except for D3G (0.7%, 724 sequences), and T175M (1%, 1026 sequences). D3G was the most frequent change in Africa (3.4%) and South America (2.8%) and T175M in Europe (1.6%). However, both aa changes were present in the six geographic regions. Neither aa change showed a steady increase over time globally or regionally. D3G first appeared in European sequences (Lithuania) in epiweek 5 and was not detected in other regions until epiweek 10 (North and South America) and 11 (Asia, Africa, and Oceania). T175M was detected for the first time in epiweek 9 in Europe (England and Netherlands), and later in Asia and South America (epiweek 10), North America and Oceania (epiweek 11) and, lastly, in Africa (epiweek 12).
Globally, we observed a significant change over time in the following six aa substitutions in the M protein: A2S, L17I, D209Y, H125Y, V23L, and V60L. Change A2S increased from 0.2% in epiweek 24 to 1.4% in epiweek 30, mainly due to Australian sequences, but no further increase was observed after this epiweek. Aa change L17I increased from 0.5% in epiweek 30 to 2.8% in epiweek 32, dropping its frequency in the last epiweeks. The increase was due to English sequences. Change D209Y showed a localized increase in frequency (from 0.2% to 1.1%) in epiweek 26. This increase was mainly due to sequences from the USA, but no further increase in global or American sequences was observed after week 27. H125Y increased during the last epiweeks available, from 0.4 in epiweek 31 to 1.3% in epiweek 34, mainly due to UK sequences, specifically English and Scottish.
V23L increased from 0.4% (epiweek 19) to 1.4% (epiweek 22). This aa change was mainly present in the UK and the increase was due to sequences belonging to Wales. Lastly, V60L frequency increased around epiweeks 27 and 28 (1.6 and 1.2%) due to European sequences, specifically from England and Switzerland, decreasing later and rising again in epiweek 34 (1.4%), mainly due to sequences from Scotland and Switzerland.
## Nucleocapsid (n) protein
A total of 99,657 N worldwide sequences were analyzed, finding 890 aa changes in 359 (85.7%) of the 419 aa residues in the N protein. The global aa conservation was 99.77%, slightly lower than the other structural proteins. Although most mutations had a low frequency, some positions showed mutations present in more than 1% of the total global sequences. It was the case for S197L (1.7%, 1686 sequences, 56% from Spain), P13L (1.8%, 1782 sequences, 62% from India and Singapore and 21% from Australia), D103Y
(1.9%, 1863 sequences, 89% from England), S194L (3.2%, 3194 sequences, 39% from England and Scotland, 29% from the USA, and 11% from India), and G204R (37%, 36,598 sequences) and R203K (37.3%, 36,876 sequences) with the highest global frequency.
G204R and R203K, both located in the SR-linker, tended to appear simultaneously in N protein and were the most frequent aa changes in the following six The global rate of the G204R and R203K combination rose from 23% in epiweek 10 to 81% in epiweek 30, dropping to 16% in epiweek 37.shows the occurrence of both aa changes in N by epiweek, globally, and in the six studied geographical regions. The exponential regression , the regional trend of this combination showed different behaviors in the other five regions, with important frequency fluctuations over time. To analyze if these regional fluctuations were related to the country of origin of the sequences or the uneven distribution of the sequences from each country along the epiweeks, the statistical average of the R203K and G204R combination was analyzed between epiweeks and countries. The number and frequency of sequences carrying the R203K and G204R combination in N protein by epiweek in each geographic region and country are described in , the aa combinations for positions 203 and 204 in each region and epiweek used for the scatter plots are available in .
In North America, the frequency increased until epiweek 23 (48.7%), and then decreased until epiweek 26, stabilizing at a frequency of around 20%. As 92% of the North American N sequences belonged to the USA , this regional curve probably describes what happened only in this country, where the R203K and G204R combination frequency reached ≈50% in epiweek 23, and then dropped to ≈20% in the following epiweeks, except for an isolated increase to 38% in epiweek 36. In Canada, the second country in this region with the most sequences, the aa combination had a median rate of 20% until epiweek 18, increasing to 73% in epiweek 21 (last epiweek with more than 10 sequences in Canada). The absence or a low number of N sequences per epiweek in the remaining North American countries excluded them for a complete similar analysis. However, when comparing data from the available epiweeks with more than 10 sequences, we also observed an increase in the frequency of the R203K and G204R combination in N sequences from Costa Rica (from 7.4% in epiweek 12 to 69.2% in epiweek 27) and Mexico (from 9.5% in epiweek 21 to 63.6% in epiweek 32); the global frequency of that combination in both countries was 28.1% and 19.3%, respectively .
In Europe, the R203K and G204R combination steadily increased until ≈85% around epiweek 30, decreasing to 3.2% in epiweek 37, where most of the sequences belonged to Wales. Most of the European sequences (72%) belonged to the UK, mainly to England, where these aa changes increased to 89% until epiweek 31, decreasing to 57% in epiweek 36 (last epiweek that met our criteria). Similarly, in Wales, the frequency raised until epiweek 33 (91%) dropping later (4% in epiweek 37), as in Scotland (93% in epiweek 32 and 29% in epiweek 36), but not in Northern Ireland (71% in epiweek 35). Although the available N sequences differed across European countries and epiweeks, this same increase-decrease tendency was observed in Italy, Denmark, and Switzerland, whereas in other countries the R203K and G204R combination frequency increased over time (as in Netherlands, Spain, and Sweden), or only in the last available epiweeks with 10 sequences (as in Germany). A steady increase in frequency was observed in Portugal, France, and Russia, whereas in Austria the frequency remained stable and in Belgium, it varied greatly without a clear tendency . In North America, the frequency increased until epiweek 23 (48.7%), and then decreased until epiweek 26, stabilizing at a frequency of around 20%. As 92% of the North In Africa, the frequency increased from 14% (epiweek 12) to 94.5% (epiweek 32), although a drop in frequency was observed in epiweeks 33 and 34, where only sequences from South Africa and Egypt were available. Most African sequences belonged to South Africa (61%), followed by the DRC (12%), Egypt (7%), and Senegal (5%) . South Africa showed a steady increase in the R203K and G204R combination frequency (≈90% frequency in epiweeks, in the DRC the frequency varied largely between epiweeks, and in Egypt it was very infrequent, only present in 4% of the total sequences, explaining the drop of the regional rate .
In Oceania, the R203K and G204R combination frequency steadily increased from epiweek 20-31. Of note, 96% of Oceania's sequences belonged to Australia where this combination was present in ≈90% sequences since epiweek 26. In New Zealand, it increased from 6% (epiweek 12) to 53% (epiweek 17), the only epiweeks with enough sequences for analysis (≥10), although its total frequency was lower than in Australia (11.8% vs. 68%) .
In Asia, the combination increased over time, except for epiweeks 26 and 27 and 31 and 32 where the frequency dropped. Most Asian sequences in these epiweeks (>50% in epiweek 26 and 100% in epiweeks 27 and 31) were from Singapore and South Korea, where the R203K and G204R combination was infrequent, explaining the frequency drop in these epiweeks. More than half of Asia's sequences were from India (31.3%), China (12.7%), and Singapore (11.5%) . The regional presence of the R203K and G204R combination changes in N varied greatly between countries . Considering those countries with >100 sequences, the higher frequencies were found in Bangladesh (87%), and Oman (75%), and the lowest in Malaysia (6%), Singapore (8%), China (9%), and South Korea and Thailand (10%). In India (total rate 36%), this combination frequency slowly increased until epiweek 25 (56%), dropping to ≈20% in the following epiweeks, and rising again to >80% in the last epiweeks that met our criteria (epiweeks 29, 33, and 34). In China, although the R203K and G204R combination total frequency was low, most sequences grouped up in the first epiweeks, where these changes were absent or extremely infrequent. However, in epiweek 14 (10 sequences) the frequency reached 40%, and in epiweeks 28-30 (next epiweeks that meet our criteria) this frequency increased to >80%, as happened in India. In Singapore, the R203K and G204R combination was absent until epiweek 28, when it increased from 4% to 47% in epiweek 36.
Finally, in South America, although some epiweeks did not meet our criteria for the time analysis, the R203K and G204R combination in N increased until epiweek 22, with great variations in frequency in the last epiweeks. Of note, the bulk of the sequences available belonged to Brazil (52%), followed by Colombia (13%), Chile (12%), and Peru (8%) . Brazil showed the highest frequency of this combination (88.7%, , and in epiweeks 10-19 (those that met our criteria), the R203K and G204R combination raised >90% since epiweek 15, observing a drop in the next epiweeks with <10 sequences, with the absence of that combination in the nine N sequences available in epiweek 30. In Colombia, Chile, and Peru, an increasing tendency was also observed in epiweeks meeting criteria (with ≥10 sequences). The observed increase was from 13.6% (epiweek 12) to 18.2% (epiweek 15) in Colombia, from 24% (epiweek 11) to 67% (epiweek 18) in Chile, and from 37% (epiweek 12) to 79.2% (epiweek 27) in Peru. In South America, in epiweeks 26 (3.6% regional rate) and 29 (14.3%), most of the available sequences belonged to Suriname, where these changes were extremely infrequent (6.67% total frequency), impacting in a low regional rate in these epiweeks. In the regional analysis, the most frequent N mutations were the already mentioned D103Y, S194L, and S197L in Europe (3.2%, 2.8%, and 2.3%, respectively), S194L in North America (4.3%); P13L and S194L in Asia (15% and 6.4%), and P13L and S197L in Oceania (5.7% and 4.8%, respectively). Other frequent mutations not mentioned before were S202N in Asia and Africa (2.5% and 3.8%, respectively), I292T in South America (25%), Q384H in Africa (7%), and L230F in Oceania (2.7%).
# Discussion
Monitoring SARS-CoV-2 genetic diversity and emerging mutations in this ongoing pandemic is crucial for understanding the evolution of this new coronavirus and assuring the performance of new diagnostic tests, vaccines, and therapies against COVID-19. This descriptive study analyzes the aa variability of the four SARS-CoV-2 structural proteins in the largest set of worldwide sequences for over 9 months since the beginning of the pandemic, describing the most frequent aa changes and their evolution in time by geographic region and epiweek.
We observed that, despite the extremely high conservation of SARS-CoV-2 structural proteins (99.99% in M, 99.98% E, 99.97% S, and 99.77% N), all proteins presented mutated residues across isolates, with different temporal trends. The S protein harbored the highest proportion of aa positions admitting changes along its sequence (88.9% of 1273 aa), followed by the E protein (86.6% of 75 aa), N protein (85.7% of 419 aa), and the M protein (74.3% of 222 aa), most of them presenting extremely infrequent changes.
In the S protein, the most remarkable aa change was D614G, present in eight out of 10 global spike analyzed sequences. This mutation has been increasing its rate over time both globally and regionally, spreading in all of the six studied regions and establishing as the dominant polymorphism. However, any aa change frequency taking place in the last epiweeks should be interpreted with caution, since it can be due to the low number of SARS-CoV-2 spike sequences for advanced epiweeks in the study sequence set, impacting the regional data. D614G has been previously reported in several studies, some of them also including a large number of sequences, but its biological impact has not been elucidated yet.
Some reports suggest that D614G spread could be explained by a founder effect, without the need for a selective advantage, especially at the beginning of an epidemic, when most individuals are susceptible. However, other reports hypothesize that D614G enhances viral fitness. It has been associated with lower RT-PCR cycle thresholds, and therefore greater infectivity, as well as infecting ACE2-expressing cells more efficiently in vitro, showing greater transmissibility. A recent study has reported that D614G increases infectivity in human lung cells or cells with bat or pangolin ACE2 receptor in cell cultures, and that D614G shifts the S protein conformation toward an ACE2-binding fusion-competent state.
Previous reports hold that the E and M proteins are conserved across ß-coronaviruses, showing high structural similarity to pangolin and bat CoVs isolates. Despite nearly identical conservation (99.98% vs. 99.99%), E harbored fewer aa changes than M (142 changes in 65 of 75 E residues vs. 291 changes in 165 of 222 M residues), S68F being the most prevalent, in only 0.2% of the global sequences. The CoVs E secondary structure consists of a short , hydrophilic N-terminal domain (NTD), a large hydrophobic transmembrane domain (25 aa) with a large proportion of valine and leucine, and a hydrophilic C-terminal domain (CTD). Although infrequent, most mutations with a global rate ≥0.1% were found in the CTD, including S68F. The last four aa of CoVs CTD harbor a post-synaptic density protein-95/discs large/zonula occludens-1 (PDZ)binding motif (PBM) that has been reported as a major determinant of virulence in SARS-CoV. The PBM is essential for the SARS-CoV virus, as mutation or deletion of PBM or full-length E causes revertant mutants with PBM restoration. Of note, SARS-CoV-2 reference sequence harbors in the PBM the same four aa (D72, L73, L74, and V75) as the SARS-CoV reference isolate (NCBI accession number NC_004718.3). In this study, we found only 221 sequences of the total 101,376 E sequences (0.2%) with aa changes in the PBM, being L73F the most prevalent, and with no frequency increase over time. Further studies are required to establish the effect of C-terminal and PBM mutations in SARS-CoV-2 E protein structure and protein-protein interactions, as well as future temporal analysis to detect any increase in their frequency.
The M protein was the most conserved protein (99.99%), but it also presented aa changes in the global approach, D3G and T175M being the most prevalent of the 291 aa changes found. The M protein consists of a short NTD ectodomain, three transmembrane domains, and a long CTD situated inside the virion particle. The D3G mutation belongs to the exposed NTD, whereas T175M is located in the CTD. It has been stated that mutations in NTD could play an important role in host-cell interaction. Of note, glycine in position 3 has been also found in bat and pangolin CoVs.
The N protein showed the lowest conservation rate (99.77%), presenting 890 changes in 359 of 419 N residues in the analyzed sequence set. This protein is divided into two main domains, i.e., the N-terminal RNA-binding domain (NTD) and the C-terminal dimerization domain (CTD), divided by a central serine/arginine-rich (SR)-linker responsible for phosphorylation. Phosphorylation of this SR-link motif in SARS-CoV modulates nucleocapsid multimerization, translational inhibitory activity and cellular localizationThe main finding in this protein was the simultaneous increase in the frequency of both G204R and R203K, the most frequent mutations in the N protein (37% and 37.3%, respectively) at the global and regional levels. Although the K203 and R204 combination showed a globally increasing incidence as compared with the R203 and G204 combination (aa in Wuhan reference sequence NC 045512.2), this behavior could have been distorted by the large number of global sequences belonging to Europe in the studied sequence set. When analyzing each geographical region individually, different temporal trends of the G204R + R203K pair were observed according to the country of origin of the N sequences and even between epiweeks. In Oceania and Africa, the rate seemed to be increasing, although not enough sequences were available for the last epiweeks' analysis. In most European countries, the G204R and R203K combination seemed to have steadily increased in time during the beginning of the pandemic, however, this upward trend may be disappearing in the last epiweeks for some countries. Further analysis should be conducted in the following epiweeks to check the incidence of these mutations in each country and geographic region with new and recent sequences when available.
G204R and R203K are located in the SR-linker of the N protein (aa 180-210), as well as the other two aa changes found (S194L and S197L) with >1% global rate. Therefore, the SR-linker was the most variable region within the N protein, with a median substitution rate of 2.7% vs. 0.03% in CTD and 0.04% in NTD. The SR-linker forms a phosphorylationdependent binding domain for protein 14-3-3, a signaling molecule involved in various cellular processes, such as cell cycle, survival, and death. It has been reported that mutations in this region could have an important biological impact, including aa change S197L, present in 1686 N sequences in our study, and representing 1.7% of the complete sequence set. S197L is a phosphorylation site for kinases involved in the control of the cell cycle. Furthermore, a change in this position would also affect the adjacent residue T198, another phosphorylation site. A recent study suggested that the co-occuring mutations R203K and G204R may decrease the overall structural flexibility of SARS-COV-2 N protein. Further analysis is needed to evaluate the impact of N SR-linker mutations in SARS-CoV-2 transmissibility and virulence, as well as of the observed global increase of the G204R and R203K combination.
The main limitation of the temporal analysis performed at the regional level is the uneven country and epiweek distribution of available sequences, mainly in Europe (with a large set of sequences belonging to England), North America (mostly sequences from the USA), and South America (with most sequences belonging to Brazil). In fact, in some countries, none or just a few sequences are available in the GISAID database, as we reported in . Moreover, in some regions, none or just a few SARS-CoV-2 sequences were available for advanced epiweeks, impacting the regional data. However, these limitations could be easily overcome by periodically checking the GISAID database as more sequences are uploaded daily, and therefore reinforce economic support for viral sequencing in countries. Therefore, we encourage worldwide upload of SARS-CoV-2 genomic sequences in this widely used database and continuous analysis of their variability in order to be able to detect any emerging mutation early.
Although SARS-CoV-2 presents a mutation rate around 10 times lower than for other RNA viruses (around 33 genomic mutations per year), this virus still acquires some mutations as it spreads from host to host, as our presented genomic data support. As more data is made available during the pandemic, we expect this study will facilitate ongoing investigation of SARS-CoV-2 variability and its consequences in COVID-19 evolution. Further analysis of linked mutations, virus-host protein interactions, and protein structure are crucial to understand the global and regional implications of the aa changes previously described. The presented data provide useful knowledge for future diagnostic, therapeutic, or vaccination approaches directed to these structural SARS-CoV-2 proteins, which could help disease control and prevention efforts, and for a better understanding of how this virus expands in certain countries or geographic areas.
Supplementary Materials: The following are available online at https://www.mdpi.com/1999-4 915/13/2/243/s1,: Number of global and regional sequences with amino acid changes across SARS-CoV-2 structural proteins. (A) Spike protein (1273 aa, 101,100 total sequences). With an asterisk, D614G, change from aspartic acid to glycine in residue 614 of spike protein. D614G was the most prevalent aa substitution present in 82,183 (81.5%) of the global sequences; (B) Envelope protein (75 aa, 101,376 total sequences). With an asterisk, S68F, change from serine to phenylalanine at position 68 of envelope protein. S68F was the most prevalent change found, present in 221 (0.2%) global sequences; (C) Membrane protein (222 aa, 103,419 total sequences). With an asterisk, D3G, change from aspartic acid to glycine in residue 3 and T175 M, change from threonine to methionine in residue 75. D3G and T175M were present in 724 (0.7%) and 1026 (1%) global sequences, respectively; (D) Nucleocapsid protein (419 aa, 99,657 total sequences). With an asterisk, R203K, change from arginine to lysine in position 203 and G204R change from glycine to arginine in position 204. R203K and G204R were present in 36,876 (37.3%) and 36,598 (37%) global sequences, respectively. Color code, red (global), green (Africa), purple (Asia), dark blue (Europe), orange (North America), light blue (South America), pink (Oceania). D, aspartic acid; G, glycine; S, serine; F, phenylalanine; T, threonine, M, methionine; R, arginine; K, lysine, : Epiweeks with available sequences in each structural SARS-CoV-2 protein across geographic regions. Epiweek, epidemiological week; S, spike protein; E, envelope protein; M, membrane protein; N, nucleocapsid protein, : Analyzed sequences, number of aa changes, and mutated aa positions in each SARS-CoV-2 structural protein by geographic region. S, spike protein; E, envelope protein; M, membrane protein; N, nucleocapsid protein, : Available sequences per country and SARS-CoV-2 structural proteins. S, Spike protein; E, envelope protein; M, membrane protein; N, nucleocapsid protein, : Number and frequency of sequences carrying R203K and G204R aa change combination in N protein by epiweek in each geographic region and country. Epiweek, epidemiological week; N, nucleocapsid protein; N.A., North America; S.A., South America. Funding: This research was supported by fundraising activities and donations from the Bomberos Ayudan Association and Fundación Alonso. This study is also included in the "Subprograma de Inmigración y Salud" from CIBERESP (Spain). P.T. was funded by ISCIII-Programa Estatal de Promoción del Talento-AES Río Hortega exte. CM19/00057. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
Institutional Review Board Statement: Not applicable.
## Informed consent statement: not applicable.
Data Availability Statement: All the sequences dataset used in this study are available in the public GISAID database (https://www.gisaid.org). All data regarding results are available in the supplementary information.
|
Dengue Virus Infection and Associated Risk Factors in Africa: A Systematic Review and Meta-Analysis
# Introduction
Dengue is an important arboviral disease, with the highest incidence in tropical and subtropical regions, with a potential to spread into other geographical areas. In the past four decades, dengue has caused a significant impact on human health and national economies. Approximately 390 million people are infected with dengue virus (DENV) annually. Of these, 96 million develop clinical manifestations that lead to 500,000 hospitalizations and 25,000 deaths, annually. Dengue is caused by an RNA virus of the family Flaviviridae that is transmitted to humans through a bite of infected Aedes mosquitoes.
## Quality assessment
The methodological quality of selected prevalence studies was evaluated by two reviewers using a quality assessment checklist adapted from Hoy and others. Risk of bias was assessed using nine domains: target population, sampling frame, sample selection method, likelihood of non-response bias, data source, case definition, study instrument that measured the parameter of interest, mode of data collection, and numerator and denominator of the parameter of interest. The risk of bias levels was low (score = 0) or high (score = 1), and the overall risk of bias was defined as low (score 0-3), moderate Viruses 2021, 13, 536 3 of 17, and high. Any discrepancy was resolved through discussion with a third reviewer.
# Statistical analysis
The extracted data were pooled using MetaXL version 5.3 software (EpiGear International Pty Ltd., Queensland, QLD, Australia). A random effect model was used to estimate the overall prevalence and 95% confidence intervals (CI), and results were presented in forest plots. The percentage of heterogeneity between studies was quantified using I 2 and chi-square tests, and I 2 ≥ 50% was considered significant. Sensitivity analysis to test the effect of each study on summary prevalence, by excluding each study step by step, was used to evaluate the robustness of overall prevalence. A funnel plot and Egger's regression test were used to detect publication bias. All results with p-values < 0.05 were considered statistically significant. Descriptive statistics, narrative synthesis, and relevant figures were used to summarize the information where statistical pooling was not possible.
# Results
## Search results and characteristics of selected studies
A total of 2170 records were retrieved from database searches. After duplicates removal and screening, 43 studies were finally included in the review. The methodological quality of studies ranged from low (0-3 score, 37 studies) to moderate (4-6 score, 6 studies). No study had a high risk of bias, six (13.9%) studies had moderate risk, and 37 (86.1%) had low risk. Out of 43 studies, 34 were prospective crosssectional, six were retrospective cross-sectional, two were prospective cohort, and one was a case-control study. denominator of the parameter of interest. The risk of bias levels was low (score = 0) or high (score = 1), and the overall risk of bias was defined as low (score 0−3), moderate (4−6), and high (7−9). Any discrepancy was resolved through discussion with a third reviewer.
# Statistical analysis
The extracted data were pooled using MetaXL version 5.3 software (EpiGear International Pty Ltd., Queensland, QLD, Australia). A random effect model was used to estimate the overall prevalence and 95% confidence intervals (CI), and results were presented in forest plots. The percentage of heterogeneity between studies was quantified using I 2 and chi-square tests, and I 2 ≥ 50% was considered significant. Sensitivity analysis to test the effect of each study on summary prevalence, by excluding each study step by step, was used to evaluate the robustness of overall prevalence. A funnel plot and Egger's regression test were used to detect publication bias. All results with p-values < 0.05 were considered statistically significant. Descriptive statistics, narrative synthesis, and relevant figures were used to summarize the information where statistical pooling was not possible.
# Results
## Search results and characteristics of selected studies
A total of 2170 records were retrieved from database searches. After duplicates removal and screening, 43 studies were finally included in the review. The methodological quality of studies ranged from low (0−3 score, 37 studies) to moderate (4−6 score, 6 studies). No study had a high risk of bias, six (13.9%) studies had moderate risk, and 37 (86.1%) had low risk. Out of 43 studies, 34 were prospective cross-sectional, six were retrospective cross-sectional, two were prospective cohort, and one was a case-control study.
## Dengue virus outbreaks and serotype distribution
Since 1964, 45 dengue outbreaks were reported in 14 countries. Most of the outbreaks occurred in East (17/45) and West (16/45) Africa. DENV-1 and DENV-2 were dominant serotypes in most of the outbreaks. During the past decade (2010-2020), there was an expansion of multiple DENV serotypes occurrence in Africa.
## Dengue prevalence
Overall, the prevalence of DENV in Africa was 14% (95% CI, 9−19%), N = 15,807). Substantial heterogeneity was found between studies during outbreak (I 2 = 99%, p < 0.01) and non-outbreak (I 2 = 95%, p < 0.01) periods. Subgroup meta-analysis showed that the prevalence of DENV was 29% (95% CI, 20−39%, N = 8966) and 3% (95% CI, 1−5%, N = 6841) during outbreak and non-outbreak periods, respectively.
## Dengue prevalence
Overall, the prevalence of DENV in Africa was 14% (95% CI, , N = 15,807). Substantial heterogeneity was found between studies during outbreak (I 2 = 99%, p < 0.01) and non-outbreak (I 2 = 95%, p < 0.01) periods. Subgroup meta-analysis showed that the prevalence of DENV was 29% (95% CI, 20-39%, N = 8966) and 3% (95% CI, 1-5%, N = 6841) during outbreak and non-outbreak periods, respectively. Sensitivity analysis based on prospective cross-sectional studies (n = 34/43) showed that 33/34 studies had good precision on overall DENV prevalence (14% (95% CI, 9-20%). One study had relatively low precision (12% (95% CI, 8-18%). Funnel plot asymmetryand Egger's regression test (p = 0.0022) indicated an evidence of publication bias in the outbreak studies. No evidence of publication biaswas detected in the non-outbreak studies (Egger's regression test, p = 0.2633).. Funnel plot asymmetryand Egger's regression test (p = 0.0022) indicated an evidence of publication bias in the outbreak studies. No evidence of publication biaswas detected in the non-outbreak studies (Egger's regression test, p = 0.2633).
## Severe dengue and co-morbidities
Between 2011 and 2019, a total of 176 severe dengue cases were reported in six countries: Burkina Faso, Côte d'Ivoire, Djibouti, Republic of Sudan, Senegal, and Tanzania. The majority of cases were reported in the Republic of Sudan (126/176) and Burkina Faso (38/176). Malaria and dengue co-infections were the most prevalent (78%, 554/711), followed by dengue and chikungunya co-infections (16%, 114/711). Other co-morbidities of dengue were yellow fever, measles, pancreatitis, and hepatitis E (6%, 43/711).
## Risk factors
Evidence from 21 reports published between 2007 and 2020 showed that old age, lack of mosquito control, living in urban areas, climate change, and history of recent travel were the leading risk factors of dengue. Other risk factors were type of occupation, lack of education, low income, and known diabetes mellitus status.
# Discussion
This systematic review reports the distribution of outbreaks and the prevalence of dengue in Africa during the outbreak and non-outbreak periods. Our results show that dengue has been reported in 24 of 54 countries and has become endemic, with repeated outbreaks in most of them. Since 1960, all four DENV serotypes (DENV-1-4) caused epidemics in all African sub-regions, with DENV-1 and DENV-2 dominating. Laboratory confirmed outbreaks were reported in 13 African countries, with the East Africa region contributing over 50% of the epidemics. These observations support evidence previously documented. After 2010, severe dengue cases have been increasingly reported in different countries, including Burkina Faso, Côte d'Ivoire, Djibouti, the Republic of Sudan, Senegal, and Tanzania. The previous report shows that these countries have experienced continuous active DENV transmission in the past decade.
Our analysis revealed an increased occurrence of multiple DENV serotypes in Africa during the past decade (2010-2020), with a greater proportion of serotypes reported in East and West Africa. Concurrent infections with multiple serotypes may pose a risk of severe dengue because lifelong immunity against primary infection by one serotype does not cross-protect subsequent infections by a different serotype. In secondary infection, antibody-dependent enhancement facilitates viral multiplication in the host cells, resulting in severe disease. Expansion of multiple DENV serotypes in Africa may be caused by several factors. International travel of infected people from epidemic and endemic countries has been associated with the introduction of DENV-1 and DENV-3 serotypes in several African countries. Spill-over of sylvatic DENV-2 strains from forest Aedes mosquitoes into a human transmission cycle possibly facilitates the spread into urban or new geographical areas with the potential to cause epi-demics. Further, increasing recognition of DENV as the cause of undifferentiated febrile syndromes, and the availability of more sensitive and specific molecular-based laboratory tests in the past decade, may have contributed to more detection and reporting of DENV serotypes.
Meta-analysis results show that the overall prevalence of dengue virus in Africa is 14%. This prevalence is relatively higher than the 7% reported in previous meta-analysis. The discrepancy could be due to differences in a number of prevalence studies included in the meta-analysis. More studies included in this review possibly contributed to an increased number of dengue positive cases. In addition, our review included studies conducted during outbreaks, thus, large studies with a higher proportion of dengue positive cases were expected. During an outbreak, the prevalence of dengue virus was 29%. Our results agree with the 30% prevalence reported in the previous meta-analysis that included prospective cross-sectional studies conducted during epidemics or following recognized epidemics in the Republic of Sudan. These observations indicate a high burden of dengue of up to 39% in febrile patients during an outbreak, highlighting the need for routine laboratory dengue diagnosis in tropical Africa.
Low dengue virus prevalence of 3% in febrile patients was found during the nonoutbreak period. In comparison, our results were relatively lower than values reported in a previous meta-analysis byinvolving febrile patients from studies conducted during the non-outbreak period. This difference could be due to the selection and epidemiological contexts of included studies. For instance, inclusion of studies conducted during ongoing epidemics or following epidemics are likely to contribute to a higher number of dengue positive cases. As a result, the overall prevalence could have been overestimated. Despite the low prevalence observed during the non-outbreak period, a burden of up to 5% in febrile patients is still of a public health concern that needs appropriate interventions. Persistent occurrence of sporadic dengue cases may indicate endemicity, therefore, routine laboratory dengue diagnosis and enhanced mosquito surveillance could help to detect cases early and identify hotspots, respectively. Despite substantial variability (I 2 = 98.89) between prospective cross-sectional studies, more than 90% of the studies had good precision on the overall dengue prevalenceand low risk of methodological quality bias. The presence of publication bias (Egger's test, p = 0.0022) in the outbreak studies, could be due to small studies with non-significant results not being published.
Co-existing unrecognized co-morbidities can complicate dengue diagnosis and patient management. The findings from previous studiesshow that co-morbidities increase the risk of severe disease and fatal outcomes among dengue patients. Our results show that in the past decade (2010-2020), malaria and dengue co-infections were the most prevalent, followed by dengue and chikungunya co-infection. A similar occurrence pattern of malaria and dengue co-infection dominance followed by dengue and chikungunya co-infection was previously reported. In Africa, co-morbidities of dengue are not usually diagnosed due to a lack of diagnostic capacity to differentiate dengue from other mosquito-borne acute febrile illnesses such as chikungunya, Zika, and yellow fever. The diseases develop similar non-specific clinical signs and can be co-transmitted with dengue. These findings underscore the need to enhance differential diagnosis of non-malaria febrile illnesses in Africa.
Results from this review show that increasing age, lack of mosquito control, living in an urban area, climate change, and recent history of travel were the leading risk factors of dengue in Africa. A high risk of contracting DENV in the old age group may be explained by continuous exposure to Flaviviruses. The presence and abundance of Aedes mosquito vectors are known to increase risk of dengue exposure. Evidence from some studies shows that people living in areas surrounding waste dump sites, opening windows at night, presence of stagnant water at home, households with indoor bathrooms, and living with open water containers were associated with a high risk of dengue. Other potential risk factors included occupation type, lack of education, low income, and known diabetes mellitus status. These findings disclose gaps in individual and environmental practices that could limit Aedes mosquito abundance and spread in African settings.
This review had some limitations. First, we could not establish a meta-analysis of DENV NS1 prevalence due to an inadequate number of studies reporting NS1 prevalence alone. Most studies had overlap data between NS1 and RT-PCR test. Second, a small number of studies (n < 10) limited subgroup meta-analysis of dengue prevalence based on setting (community versus healthcare facility), geographical sub-regions, and the design other than prospective cross-sectional. Third, it is possible that some individuals were asymptomatic and could not be detected in the included studies, thus, the number of dengue positive cases may be higher than reported in this review. Despite the limitations, we are confident that our findings provide useful information for clinical practice and public health policy decisions.
# Conclusions
In conclusion, this review reveals a high burden of dengue infection and highlights an increased risk of severe disease in Africa due to the increasing circulation of multiple dengue virus serotypes. We advocate for the need of routine laboratory dengue diagnosis in Africa to facilitate early detection of cases, provision for appropriate patient care, identification of serotypes/genotypes, and outbreak preparedness. It is important to implement effective mosquito surveillance to identify hotspots, and control through the promotion of education on individual behaviours and environmental management practices that can limit the spread of dengue infection in Africa.
# Supplementary materials:
The following are available online at https://www.mdpi.com/1999-4 915/13/4/536/s1,: Quality assessment checklist,: Risk of bias scores,: Funnel plot for the outbreak studies,: Funnel plot for the non-outbreak studies.
|
Self-regulation versus social influence for promoting cooperation on networks
If T − 1 > −S, then the temptation to defect is stronger than the disadvantage of being defected. This game will be hereafter called T-driven.- If T − 1 < −S, then the disadvantage of being defected is stronger than the temptation to defect. This game will be hereafter called S-driven.
## Preliminaries
## The evolutionary game equation on networks (egn) and self games
Let V = {1, 2, . . . , N } be the set of players. Each player is placed in a vertex of an undirected graph, defined by the adjacency matrix A = {a v,w } ∈ {0, 1} N ×N with (v, w) ∈ V 2 . Specifically, a v,w = 1 when v is connected to w, 0 otherwise. It is also assumed that a v,v = 0. The degree of player v is defined as the cardinality of his neighborhood, namely:
[formula] k v = N w=1 [/formula]
a v,w .
In the literature on evolutionary game theory, it is assumed that, at each time, one individual uses a pure strategy in a given set while playing games with connected individuals. These games are often modeled as Prisoner's dilemma games, where the set of pure strategies contains only two elements: cooperation (C) and defection . The outcome of those games is described by the following payoff matrix:
[formula] B = R S T P , [/formula]
where R is the reward when both players cooperate, T is the temptation to defect when the opponent cooperates, S is the sucker's payoff earned by a cooperative player when the opponent defects, and P is the punishment for mutual defection. More specifically, for a Prisoner's dilemma game, the reward is a better outcome than the punishment (R > P ), the temptation payoff is higher than the reward (T > R), and the punishment is preferred to the sucker's payoff (P > S). Without loss of generality, we assume R = 1 and P = 0, thereby normalizing the advantage of mutual cooperation over mutual defection to 1 . Under this assumption, T > 1 and S < 0. It is straightforward to note that T − 1 quantifies the temptation to defect, while −S is related to the disadvantage of being defected. Therefore, we can distinguish two cases:
At each time instant, an individual v will play k v continuous prisoner's dilemma games with his neighbors ; specifically, he chooses his own level of cooperation indicated by x v ∈ [0, 1]. Notice that discrete strategies C and D are special cases obtained for x v = 1 and x v = 0, respectively. When any two connected players v and w take part in a game, the payoff for v is defined by the continuous function φ : [0, 1] × [0, 1] → R (see :
[formula] φ(x v , x w ) = (R − T + P − S)x v x w + (S − P )x v + (T − P )x w + P = (1 − T − S)x v x w + Sx v + T x w .(1) [/formula]
The total payoff φ v of player v is the sum of all outcomes of two-player games with neighbors. Formally, the payoff function φ v : [0, 1] N → R is defined as follows:
[formula] φ v (x) = N w=1 a v,w φ(x v , x w ), [/formula]
where x is the vector of all the x v variables. Moreover, given the vector x, we define the following payoff of pure strategies C (x v = 1) and D (x v = 0):
[formula] p C v (x) = N w=1 a v,w φ(1, x w ) = N w=1 a v,w [(R − S)x w + S] = N w=1 a v,w [(1 − S)x w + S] p D v (x) = N w=1 a v,w φ(0, x w ) = N w=1 a v,w [(T − P )x w + P ] = N w=1 a v,w T x w . [/formula]
Following , the EGN equation for two-strategy games reads as follows:
[formula] x v = x v (1 − x v )∆p v (x),(2) [/formula]
where
[formula] ∆p v (x) = p C v (x) − p D v (x) = N w=1 a v,w [(1 − T − S)x w + S]. [/formula]
It is clear that the level of cooperation of player v increases (decreases) when ∆p v (x) is positive (negative). In other words, the player v will be more cooperative over time as long as the payoff he can earn using the pure strategy C is better than the payoff he can earn using the pure strategy D.
This evaluation of the benefits provided by the available strategies can be formulated in an alternative way. Specifically, suppose that player v is able to appraise whether a change of his strategy x v produces an improvement of his payoff φ v . This means that, if the derivative of φ v with respect to x v is positive (negative), the player would like to increase (decrease) his level of cooperation. According to this idea, notice that:
[formula] ∂φ v (x) ∂x v = N w=1 a v,w ∂φ(x v , x w ) ∂x v = N w=1 a v,w [(1 − T − S)x w + S] = ∆p v (x). [/formula]
Thus, the EGN equation (2) can be rewritten as follows:
[formula] x v = x v (1 − x v ) ∂φ v (x) ∂x v .(3) [/formula]
Since in (3) x v (1 − x v ) ≥ 0, then the sign ofẋ v depends only on the term ∂φ v /∂x v , which involves the states x w of all neighbors, rather than the current state x v of player v himself. Then, if this term is positive (negative), player v would like to increase (decrease) his level of cooperation x v . Of course, when it is null, player v has no incentives to change his mind.
It is worthwhile to notice that, while the replicator equation is used to describe the dynamics of population where strategies correspond to the phenotypes of the individuals, its extension on graphs, the EGN equation, is suitable for analyzing the dynamics of individuals arranged on a network, and able to choose their strategies in the continuous set [0, 1].
The EGN equation (3), as well as most of the models presented in the literature, assumes that the strategy dynamics of a generic player v is driven only by external factors. Indeed, ∂φv ∂xv depends only on the state of neighboring players, not on the current state x v of player v himself. Inspired by mechanisms describing self-regulation in animal societies reported in , we overcome this issue by introducing the Self-Regulated EGN equation (SR-EGN); this new model is obtained by adding a self-regulating term f v to the EGN equation, balancing the external feedback ∂φv ∂xv . The SR-EGN equation reads as:
[formula] x v = x v (1 − x v ) ∂φ v (x) ∂x v − β v f v (x v ) ,(4) [/formula]
where the parameter β v is used to tune the effectiveness of the introduced self-regulation mechanism. Specifically, we assume that this self-regulation term embodies a game that a given individual plays against himself. To describe this self game, consider two generic players, choosing feasible strategies y and z. As already mentioned, the first player can assess whether a change of his strategy y can lead to an improvement of the payoff φ(y, z). In particular, the assessment is based on the sign of the partial derivative
[formula] ∂φ(y, z) ∂y = (1 − T − S)z + S. [/formula]
In the particular case of individuals representing both the first and second player at the same time, the derivative reads as follows:
[formula] ∂φ(y, z) ∂y y=xv,z=xv = (1 − T − S)x v + S. [/formula]
Therefore, the self-regulating term is defined as:
[formula] f v (x v ) = ∂φ(y, z) ∂y y=xv,z=xv = (1 − T − S)x v + S. [/formula]
Notice that function f v (x v ) in equation (4) depends on x v . Thus, the self game introduces a feedback mechanism regulated by the parameter β v ∈ R. In particular, in equation (4), β v > 0 represents a negative feedback, β v < 0 stands for a positive feedback, while β v = 0 refers to situations where the player v does not play a self game.
## Steady states and linearization
A steady state x * is a solution of equation (4) satisfyingẋ v = 0 ∀v ∈ V. In order to be feasible, the components of a steady state must belong to the set [0, 1]. Formally, the set of feasible steady states is:
[formula] Θ = {x * ∈ R N :ẋ * v = 0 ∧ x * v ∈ [0, 1] ∀v ∈ V}. [/formula]
It is clear that all points such that for all v, x * v = 0 or x * v = 1 are in the set Θ. We remark that set Θ may contain also other steady states, exhibiting partial levels of cooperation. Among the pure steady states, particularly relevant are the followings:
[formula] x * ALLC = [1, 1, . . . , 1] [/formula]
, and
x * ALLD = [0, 0, . . . , 0] . Indeed, they represent a population composed by full cooperators and full defectors, respectively, and thus they describe the spread of cooperation, or alternatively its extinction in a given population.
The dynamical properties of these two pure steady states is fundamental for the emergence of cooperation. In particular, their stability can be locally analyzed by linearizing system (4).
The Jacobian matrix of system (3), J(x) = {j v,w (x)}, is defined as follows:
[formula] j v,w (x) = ∂ẋ v ∂x w = x v (1 − x v )(1 − T − S), if a v,w = 1 (1 − 2x v ) ∂φ v (x) ∂x v − β v f v (x v ) − β v x v (1 − x v )(1 − T − S), if w = v 0, otherwise . [/formula]
It is easy to show that the Jacobian matrix reduces to a diagonal one for both x * ALLC and x * ALLD . Moreover, observe that:
[formula] ∂φ v (x) ∂x v x * ALLC = N w=1 a v,w [(1 − T − S) · 1 + S] = N w=1 a v,w (1 − T ) = k v (1 − T ), ∂φ v (x) ∂x v x * ALLD = N w=1 a v,w [(1 − T − S) · 0 + S] = N w=1 a v,w S = k v S, f v (1) = (1 − T − S) · 1 + S = 1 − T, f v (0) = (1 − T − S) · 0 + S = S. [/formula]
Therefore:
[formula] j v,v (x * ALLC ) = (1 − 2 · 1) ∂φ v (x) ∂x v x * ALLC − β v f v (1) = (T − 1)(k v − β v ) for x * ALLC , and j v,v (x * ALLD ) = (1 − 2 · 0) ∂φ v (x) ∂x v x * ALLD − β v f v (0) = S(k v − β v ) for x * ALLD . [/formula]
The system (4) may have steady states with some components in the set (0, 1). Consider a steady state x * and suppose that one component x * v belongs to the set (0, 1). Therefore:
[formula] ∂φ v (x) ∂x v − β v f v (x v ) x=x * = 0.(5) [/formula]
Notice that:
[formula] ∂φ v (x) ∂x v − β v f v (x v ) = N w=1 a v,w [(1 − T − S)x w + S] − β v ((1 − T − S)x v + S) . = (1 − T − S) N w=1 a v,w x w + N w=1 S − β v ((1 − T − S)x v + S) = k v (1 − T − S) 1 k v N w=1 a v,w x w + k v S − β v ((1 − T − S)x v + S) = k v [(1 − T − S)x v + S] − β v [(1 − T − S)x v + S],(6) [/formula]
where
[formula] x v = 1 k v N w=1 a v,w x w , [/formula]
represents an equivalent player, incorporating the average decisions of all neighbors of player v.
For x = x * , we have that equation (5) can be rewritten as:
[formula] k v [(1 − T − S)x * v + S] − β v [(1 − T − S)x * v + S = 0, yielding to x * v = k v β v x * v − S 1 − T − S 1 − k v β v . [/formula]
The following theoretical results on the feasibility of x * v hold.
[formula] Theorem 1. Consider an T-driven game. Let ρ = 1−T S and let Q v = (ρ−1)kvx * v +kv ρ . Then β v ∈ (Q v , ρQ v ) ⇐⇒ x * v ∈ (0, 1). [/formula]
Proof. First of all, T > 1, S < 0 and 1 − T > −S (T-driven game), then ρ = 1−T S > 1.
Secondly, notice that:
[formula] Q v = (ρ − 1)k v x * v + k v ρ > 0, since k v ≥ 1, x * v ∈ [0, 1] and ρ > 1. Therefore, β v is positive when it belongs to the set (Q v , ρQ v ). Since β v < ρQ v , then ρQ v − β v > 0. Dividing the last inequality by (ρ − 1)β v , we get: ρQ v − β v (ρ − 1)β v > 0.(7) [/formula]
Similarly, since β v > Q v , then ρQ v < ρβ v . Moreover, subtracting β v from both left and right side of the inequality, we get
[formula] ρQ v − β v < ρβ v − β v , or equivalently, ρQ v − β v < (ρ − 1)β v . [/formula]
Dividing the last inequality by (ρ − 1)β v , we get:
[formula] ρQ v − β v (ρ − 1)β v < 1.(8) [/formula]
The proof is concluded by observing that:
[formula] x * v = k v β v x * v − S 1 − T − S 1 − k v β v = k v β v x * v − 1 − T − S S −1 1 − k v β v = k v β v x * v − 1 − T S − 1 −1 1 − k v β v = k v β v x * v − (ρ − 1) −1 1 − k v β v = k v β v x * v − 1 ρ − 1 1 − k v β v = k v β v x * v − 1 ρ − 1 β v − k v β v = (ρ − 1)k v x * v + k v − β v (ρ − 1)β v = ρQ v − β v (ρ − 1)β v .(9) [/formula]
Therefore, inequalities (7) and (8) imply that:
[formula] x * v ∈ (0, 1). Theorem 2. Consider an S-driven game. Let ρ = S 1−T and let Q v = (1 − ρ)k v x * v + ρk v . Then β v ∈ Q v ρ , Q v ⇐⇒ x * v ∈ (0, 1). [/formula]
Proof. First of all, T > 1, S < 0 and 1 − T < −S (S-driven game), then ρ = S 1−T > 1.
Secondly, notice that:
[formula] Q v = (1 − ρ)k v x * v + ρk v = k v ((1 − ρ)x * v + ρ) = k v (x * v + ρ(1 − x * v )) > 0, since k v ≥ 1, x * v ∈ [0, 1] and ρ > 1. Therefore, β v is positive when it belongs to the set Qv ρ , Q v . Since β v > Qv ρ , then Q v − ρβ v < 0. Dividing the last inequality by (1 − ρ)β v , we get: Q v − ρβ v (1 − ρ)β v > 0,(10) [/formula]
where the last inequality holds since
[formula] (1 − ρ)β v < 0. Similarly, since β v < Q v , then β v − ρβ v < Q v − ρβ v , or equivalently, (1 − ρ)β v < Q v − ρβ v . Dividing the last inequality by (1 − ρ)β v , we get: Q v − ρβ v (1 − ρ)β v < 1.(11) [/formula]
The proof is concluded by observing that:
[formula] x * v = k v β v x * v − S 1 − T − S 1 − k v β v = k v β v x * v − 1 − T − S S −1 1 − k v β v = k v β v x * v − 1 − T S − 1 −1 1 − k v β v = k v β v x * v − 1 ρ − 1 −1 1 − k v β v = k v β v x * v − ρ 1 − ρ 1 − k v β v = k v β v x * v − ρ 1 − ρ β v − k v β v = (1 − ρ)k v x * v + ρk v − ρβ v (1 − ρ)β v = Q v − ρβ v (1 − ρ)β v .(12) [/formula]
Therefore, inequalities (10) and (11) imply that:
x * v ∈ (0, 1).
The stability of a steady state x * v ∈ (0, 1) can be studied by observing that the sign ofẋ v in equation (4) depends only on the term ∂φv(x)
∂xv − β v f v (x v ); according to equation (6), this term is linear with respect to x v , and hence the stability of x * v depends on the slope of this straight line. Specifically, the slope is equal to
[formula] −β v (1 − T − S). [/formula]
Since β v > 0, then we can have two cases:
- the slope is positive when the game is T-driven. In this case, the point x * v acts as a repeller;
- the slope is negative when the game is S-driven. In this case, the point x * v acts as an attractor.
These results are reported in of the main paper.
## Global emergence of cooperation in the egn equation with self-regulations
The global emergence of cooperation is reached when all members of a social network turn their strategies to cooperation. Therefore, the asymptotic stability of x * ALLC , as well as the instability of x * ALLD , have a fundamental role in this context. In order to study the stability of steady states x * ALLC and x * ALLD , we start by analyzing their linear stability. Moreover, an appropriate Lyapunov function is found, for proving that, under certain hypotheses, x * ALLC is also globally asymptotically stable. Finally, different hypotheses are used for identifying a Lyapunov function proving that x * ALLD is globally asymptotically stable.
## Asympotic stability of x *
## Allc
Recall that the spectrum of J(x * ) characterizes the linear stability of any steady state x * .
The following results hold.
Theorem 3. If β v > k v ∀v ∈ V, then x * ALLC is asymptotically stable. Proof. As shown before, the Jacobian matrix evaluated for x * ALLC is diagonal. Then, the elements on the diagonal of the Jacobian matrix correspond to its eigenvalues and they are defined as follows:
[formula] j v,v (x * ALLC ) = λ v = (T − 1)(k v − β v ). [/formula]
Since β v > k v ∀v ∈ V and T > 1, all the eigenvalues are negative. Thus, x * ALLC is asymptotically stable.
Theorem 4. If ∃v ∈ V : β v > k v , then x * ALLD is unstable. Proof. The eigenvalues of the Jacobian matrix relative to the steady state x * ALLD are:
[formula] j v,v (x * ALLD ) = λ v = S(k v − β v ). [/formula]
If the hypothesis of the theorem are fulfilled, since S < 0, then there is at least one positive eigenvalue, implying that x * ALLD is an unstable steady state. These results are summarized as follows: defection dominates over cooperation. Then, if the system does not present any internal feedback mechanism (i.e. β v = 0 ∀v ∈ V), the whole social network will converge to x * ALLD (cooperation vanishes). Anyway, using β v > k v for all the members of the population, x * ALLD is destabilized and x * ALLC becomes attractive.
## Global asymptotic stability of x *
## Allc
Theorems 3 and 4 prove that under suitable condition, x * ALLC is asymptotically stable and x * ALLD is unstable. Anyway, this is not sufficient to prove the global emergence of cooperation. Indeed, there can be some other steady states in Θ which may be also attractive. Nevertheless, a Lyapunov function for the steady state x * ALLC on the set x ∈ (0, 1] N can be found.
Adapting the approach presented in [S10, S11] to the SR-EGN equation, we consider the following function:
[formula] V (x) = − N v=1 log(x v ), for x ∈ (0, 1] N . Notice that V (x * ALLC ) = 0, and V (x) > 0 ∀x = x * ALLC . [/formula]
Moreover, the time derivative of V (x) is defined as follows:
[formula] V (x) = ∂V (x) ∂t = N v=1 ∂V (x) ∂x vẋ v = = − N v=1 1 x v x v (1 − x v ) ∂φ v (x) ∂x v − β v f v (x v ) = = N v=1 (x v − 1) ∂φ v (x) ∂x v − β v f v (x v ) .(13) [/formula]
Clearly,V (x * ALLC ) = 0.
Starting from these premises, ifV (x) < 0 for all x ∈ (0, 1] N \ {x * ALLC }, then V (x) is a Lyapunov function.
Let's introduce the following quantities:
[formula] ψ = inf y∈(0,1] [(1 − T − S)y + S] ,(14)ξ = sup y∈(0,1] [(1 − T − S)y + S] ,(15) [/formula]
and ρ = ψ ξ .
It is easy to show that:
[formula] ψ = min{1 − T, S} = − max{T − 1, −S}, and ξ = max{1 − T, S} = − min{T − 1, −S}, [/formula]
and hence:
[formula] ρ = max{T − 1, −S} min{T − 1, −S} . [/formula]
Interestingly, the parameter ρ is equal to the ratio between the maximum and minimum of two "driving forces", namely the temptation to defect (T − 1) and the disadvantage of being defected (−S). Therefore, we can distinguish two cases:
- for a T-driven game, since T − 1 > −S, then ψ = 1 − T , ξ = S and ρ = 1 − T S ;
- for a S-driven game, since T − 1 < −S, then ψ = S, ξ = 1 − T and ρ = S 1 − T .
The following result holds.
Theorem 5. If β v > ρk v ∀v ∈ V, then V (x) is a Lyapunov function.
Proof. It is straightforward to observe that:
[formula] ∂φ v (x) ∂x v = N w=1 a v,w [(1 − T − S)x w + S] ≥ N w=1 a v,w ψ = k v ψ.(16) [/formula]
Similarly, since β v > 0, we get that:
[formula] β v f v (x v ) = β v [(1 − T − S)x v + S] ≤ β v ξ.(17) [/formula]
Joining (16) and (17) together, we get that:
[formula] ∂φ v (x) ∂x v − β v f v (x v ) ≥ k v ψ − β v ξ. [/formula]
Moreover, notice that:
[formula] β v > ρk v ⇒ β v > ψ ξ k v ⇒ k v ψ − β v ξ > 0, and hence ∂φ v (x) ∂x v − β v f v (x v ) > 0 ∀v ∈ V.(18) [/formula]
According to equations (13) and (18), since x v − 1 < 0 for all x ∈ (0, 1] N \ {x * ALLC }, we guarantee thatV (x) < 0 for all x ∈ (0, 1] N \ {x * ALLC }. Hence, V (x) is a Lyapunov function.
[formula] Corollary 1. If β v > ρk v , then lim t→∞ x v (t) = 1 . [/formula]
Proof. This is a direct consequence of inequality (18).
## Global asymptotic stability of x *
## Alld
Consider the following function for the steady state x * ALLD :
[formula] V (x) = − N v=1 log(1 − x v ), for x ∈ [0, 1) N . Notice that V (x * ALLD ) = 0, and V (x) > 0 ∀x = x * ALLD . [/formula]
The time derivative of V (x) is defined as follows:
[formula] V (x) = ∂V (x) ∂t = N v=1 ∂V (x) ∂x vẋ v = = N v=1 1 1 − x v x v (1 − x v ) ∂φ v (x) ∂x v − β v f v (x v ) = = N v=1 x v ∂φ v (x) ∂x v − β v f v (x v ) .(19) [/formula]
Clearly,V (x * ALLD ) = 0.
Starting from these premises, ifV (
[formula] x) < 0 for all x ∈ [0, 1) N \ {x * ALLD }, then V (x) is a Lyapunov function. [/formula]
The following result holds.
Theorem 6. If β v < 1 ρ k v ∀v ∈ V, then V (x) is a Lyapunov function. Proof. It is straightforward to see that:
[formula] ∂φ v (x) ∂x v = N w=1 a v,w [(1 − T − S)x w + S] ≤ N w=1 a v,w ξ = k v ξ.(20) [/formula]
Similarly, since β v > 0, we get that:
[formula] β v f v (x v ) = β v [(1 − T − S)x v + S] ≥ β v ψ.(21) [/formula]
Joining (20) and (21) together, we get that:
[formula] ∂φ v (x) ∂x v − β v f v (x v ) ≤ k v ξ − β v ψ. [/formula]
Moreover, notice that:
[formula] β v < 1 ρ k v ⇒ β v < ξ ψ k v ⇒ k v ξ − β v ψ < 0, and hence ∂φ v (x) ∂x v − β v f v (x v ) < 0 ∀v ∈ V.(22) [/formula]
According to equations (19) and (22), since x v > 0 for all x ∈ [0, 1) N \{x * ALLD }, we guarantee thatV (x) < 0 for all x ∈ [0, 1) N \{x * ALLD }. Hence, V (x) is a Lyapunov function.
[formula] Corollary 2. If β v < 1 ρ k v , then lim t→∞ x v (t) = 0 . [/formula]
Proof. This is a direct consequence of inequality (22).
## Game transitions
The previous results highlight the effectiveness of the self-regulating term in promoting cooperation. In this Section we show that this fact is due to transitions between different games, occurring when the system parameters are changed. We start from the generic formulation of the replicator equation for the two-strategy case, as proposed in :
[formula] x = x(1 − x) [(σ C + σ D )x − σ D ] ,(23) [/formula]
where σ C and σ D are the elements of the diagonal payoff matrix
[formula] B = σ C 0 0 σ D . [/formula]
The sign of the parameters σ C and σ D rules the dynamics of equation (23) as follows:
- σ C < 0 and σ D > 0: Prisoner's dilemma game (PD), for which defection is the only dominant strategy;
- σ C > 0 and σ D > 0: Stag Hunt game (SH), where cooperation is the best response to cooperation, and defection is the best response to defection;
- σ C < 0 and σ D < 0: Chicken game (CH), where cooperation is the best response to defection, and vice versa;
- σ C > 0 and σ D < 0: Harmony game (HA), for which cooperation is the only dominant strategy.
It is worthwhile to notice that σ C < 0 models the temptation to defect, while σ D > 0 models the fear to be betrayed.
We can rewrite the SR-EGN equation according to the structure of equation (23):
[formula] x v = x v (1 − x v ){k v [(1 − T − S)x v + S] − β v [(1 − T − S)x v + S]} = x v (1 − x v ) [(σ v C + σ v D )x v − σ v D ] [/formula]
. Then, we have the following relationships:
[formula] σ v C + σ v D = −β v (1 − T − S) σ v D = − {k v [(1 − T − S)x v + S] − β v S} ⇒ ⇒ σ v C = k v [(1 − T − S)x v + S] − β v (1 − T ) σ v D = −k v [(1 − T − S)x v + S] + β v S .(24) [/formula]
σ v C and σ v D characterize the equivalent game of player v. They depend on game parameters T and S, degree k v and self-regulating factor β v . Interestingly, they depend also on the strategy of the equivalent player x v . This fact implies that the values of σ v C and σ v D change not only by changing the system's parameters, but also over time, according to the dynamics of x v , thus producing dynamical game transitions. Hereafter, we report the conditions for which these transitions occur.
2.4.1 T-driven case: σ v C We establish the conditions for which the sign of σ v C is independent or dependent on the on the level of cooperation of his neighborhood x v . According to equation (24), we have
[formula] σ v C = k v [(1 − T − S)x v + S] − β v (1 − T ). [/formula]
We distinguish three cases.
[formula] 1. σ v C > 0 for any x v ∈ [0, 1] if k v [(1 − T − S)x v + S] > β v (1 − T ) [/formula]
. Using equation (14), since in this case ψ = 1 − T , we have that:
[formula] k v [(1 − T − S)x v + S] ≥ k v ψ = k v (1 − T ) ∀x v ∈ [0, 1]. [/formula]
Then
[formula] σ v C ≥ k v (1 − T ) − β v (1 − T ) = (1 − T )(k v − β v ) ∀x v ∈ [0, 1]. [/formula]
Since T > 1, we get the following result:
[formula] β v > k v ⇒ σ v C > 0 ∀x v ∈ [0, 1]. 2. σ v C < 0 if k v [(1 − T − S)x v + S] < β v (1 − T ) for any x v ∈ [0, 1]. [/formula]
Using equation (15), since in this case ξ = S, we have that:
[formula] k v [(1 − T − S)x v + S] ≤ k v ξ = k v S ∀x v ∈ [0, 1]. [/formula]
Since ρ = 1−T S , then:
[formula] σ v C ≤ k v S − β v (1 − T ) = S(k v − β v ρ) ∀x v ∈ [0, 1]. [/formula]
By hypothesis, S < 0, and hence we get the following result:
[formula] β v < k v ρ ⇒ σ v C < 0 ∀x v ∈ [0, 1]. [/formula]
3. The sign of σ v C depends on x v when β v ∈ kv ρ , k v . Noticing that
[formula] σ v C > 0 ⇔ k v ((1 − T − S)x v + S) − β v (1 − T ) > 0, [/formula]
and dividing by S < 0, since ρ = 1−T S , we get:
[formula] k v ((ρ − 1)x v + 1) − β v ρ < 0 ⇒ x v < β v ρ − k v k v (ρ − 1) . [/formula]
Summarizing:
[formula] - σ v C > 0 for x v < β v ρ − k v k v (ρ − 1) ; - σ v C < 0 for x v > β v ρ − k v k v (ρ − 1) . [/formula]
Additionally, by the definition of σ v C we have that
[formula] ∂σ v C ∂x v = k v (1 − T − S) < 0, [/formula]
then for the T-driven case σ v C decreases as x v increases.
## D
We establish the conditions for which the sign of σ v D is independent or dependent on the on the level of cooperation of his neighborhood x v . According to equation (24), we have
[formula] σ v D = −k v [(1 − T − S)x v + S] + β v S. [/formula]
We distinguish three cases. (15), since ξ = S, we have that
[formula] 1. σ v D > 0 for any x v ∈ [0, 1] if k v [(1 − T − S)x v + S] < β v S. Using equationk v [(1 − T − S)x v + S] ≤ k v ξ = k v S ∀x v ∈ [0, 1]. [/formula]
Then,
[formula] σ v D ≥ −k v S + β v S = S(β v − k v ) ∀x v ∈ [0, 1]. [/formula]
Since S < 0, we get the following result:
[formula] β v < k v ⇒ σ v D > 0 ∀x v ∈ [0, 1]. 2. σ v D < 0 for any x v ∈ [0, 1] if k v [(1 − T − S)x v + S] > β v S. [/formula]
Using equation (14), since ψ = 1 − T we have that:
[formula] k v [(1 − T − S)x v + S] ≥ k v ψ = k v (1 − T ) ∀x v ∈ [0, 1], [/formula]
Then, since ρ = 1−T S , we have:
[formula] σ v D ≤ −k v (1 − T ) + β v S = −S(k v ρ − β v ) ∀x v ∈ [0, 1]. [/formula]
Since, moreover, S < 0, we get the following result:
[formula] β v > k v ρ ⇒ σ v D < 0 ∀x v ∈ [0, 1]. 3. For β v ∈ [k v , k v ρ] the sign of σ v D depends on x v . Starting from σ v D > 0 ⇔ k v ((1 − T − S)x v + S) − β v S < 0, [/formula]
and dividing the inequality by S < 0, since ρ = 1−T S , we get:
[formula] k v ((ρ − 1)x v + 1) − β v > 0 ⇒ x v > β v − k v k v (ρ − 1) . [/formula]
Summarizing:
[formula] - σ v D > 0 for x v > β v − k v k v (ρ − 1) ; - σ v D < 0 for x v < β v − k v k v (ρ − 1) . [/formula]
Additionally, by the definition of σ v D we have that:
[formula] ∂σ v D ∂x v = −k v (1 − T − S) > 0 [/formula]
for the T-driven case, and hence σ v D increases as x v increases.reports a summary of the results on game transitions for the T-driven case.
Condition We establish the conditions for which the sign of σ v C is independent or dependent on the on the level of cooperation of his neighborhood x v . According to equation (24), we have
[formula] σ v C σ v D Set of v Equivalent game 0 < β v < k v ρ < 0 > 0 D PD k v ρ < β v < k v ∈ R > 0 U PD or SH k v < β v < k v ρ > 0 ∈ R U SH or HA β v > k v ρ > 0 < 0 C HAσ v C = k v [(1 − T − S)x v + S] − β v (1 − T ). [/formula]
We distinguish three cases.
[formula] 1. σ v C > 0 for any x v ∈ [0, 1] if k v [(1 − T − S)x v + S] > β v (1 − T ) [/formula]
. Using equation (14), since ψ = S, we have
[formula] k v [(1 − T − S)x v + S] ≥ k v ψ = k v S ∀x v ∈ [0, 1]. [/formula]
Therefore, since ρ = S 1−T , we have
[formula] σ v C ≥ k v S − β v (1 − T ) = (1 − T )(k v ρ − β v ) ∀x v ∈ [0, 1]. [/formula]
Moreover, since T > 1, we get the following result:
[formula] β v > k v ρ ⇒ σ v C > 0 ∀x v ∈ [0, 1]. 2. σ v C < 0 for any x v ∈ [0, 1] if k v [(1 − T − S)x v + S] < β v (1 − T ) [/formula]
. Using equation (15), since ξ = 1 − T , we have
[formula] k v [(1 − T − S)x v + S] ≤ k v ξ = k v (1 − T ) ∀x v ∈ [0, 1]. [/formula]
Then:
[formula] σ v C ≤ k v (1 − T ) − β v (1 − T ) = (1 − T )(k v − β v ) ∀x v ∈ [0, 1]. [/formula]
By hypothesis, T > 1, and hence we get the following result:
[formula] β v < k v ⇒ σ v C < 0 ∀x v ∈ [0, 1]. 3. For β v ∈ [k v , k v ρ] [/formula]
, the sign of σ v C depends on x v . Starting from
[formula] σ v C > 0 ⇔ k v ((1 − T − S)x v + S) − β v (1 − T ) > 0, [/formula]
and dividing the inequality by 1 − T < 0, since ρ = S 1−T , we get:
[formula] k v ((1 − ρ)x v + ρ) − β v < 0 ⇒ x v > k v ρ − β v k v (ρ − 1) . [/formula]
Summarizing:
[formula] - σ v C > 0 for x v > k v ρ − β v k v (ρ − 1) ; - σ v C < 0 for x v < k v ρ − β v k v (ρ − 1) . [/formula]
Moreover, since:
[formula] ∂σ v C ∂x v = k v (1 − T − S) > 0 [/formula]
for the S-driven case, then σ v C increases as x v increases.
## S-driven case: σ v d
We establish the conditions for which the sign of σ v D is independent or dependent on the level of cooperation of his neighborhood x v . According to equation (24), we have
[formula] σ v D = −k v [(1 − T − S)x v + S] + β v S. [/formula]
We distinguish three cases.
[formula] 1. σ v D > 0 for any x v ∈ [0, 1] if k v [(1 − T − S)x v + S] < β v S. [/formula]
Using equation (15), since ξ = 1 − T , we have
[formula] k v [(1 − T − S)x v + S] ≤ k v ξ = k v (1 − T ) ∀x v ∈ [0, 1]. [/formula]
Therefore, recalling that ρ = S 1−T , we get:
[formula] σ v D ≥ −k v (1 − T ) + β v S = (1 − T )(β v ρ − k v ) ∀x v ∈ [0, 1]. [/formula]
Since T > 1, we get the following result:
[formula] β v < k v ρ ⇒ σ v D > 0 ∀x v ∈ [0, 1]. 2. σ v D < 0 for any x v ∈ [0, 1] if k v [(1 − T − S)x v + S] > β v S. [/formula]
Using equation (14), since ψ = S, we have that:
[formula] k v [(1 − T − S)x v + S] ≥ k v ψ = k v S ∀x v ∈ [0, 1]. [/formula]
Then:
[formula] σ v D ≤ −k v S + β v S = −S(k v − β v ) ∀x v ∈ [0, 1]. [/formula]
Since S < 0, we get the following result: 3. For β v ∈ kv ρ , k v , the sign of σ v D depends on x v . Starting from
[formula] β v > k v ⇒ σ v D < 0 ∀x v ∈ [0, 1]. σ v C σ v D Set of v Game 0 < β v < k v ρ < 0 > 0 D PD k v ρ < β v < k v < 0 ∈ R U PD or CH k v < β v < k v ρ ∈ R < 0 U CH or HA β v > k v ρ > 0 < 0 C HAσ v D > 0 ⇔ k v ((1 − T − S)x v + S) − β v S < 0, [/formula]
and dividing the inequality by 1 − T < 0, since ρ = S 1−T , we get:
[formula] k v ((1 − ρ)x v + ρ) − β v ρ > 0 ⇒ x v < ρ(k v − β v ) k v (ρ − 1) . [/formula]
Summarizing:
[formula] - σ v D > 0 for x v < ρ(k v − β v ) k v (ρ − 1) ; - σ v D < 0 for x v > ρ(k v − β v ) k v (ρ − 1) . [/formula]
Additionally, by the definition of σ v D we have that:
[formula] ∂σ v D ∂x v = −k v (1 − T − S) < 0 [/formula]
for the T-driven case, and hence σ v D decreases as x v increases.reports a summary of the results on game transitions for the S-driven case.
|
Fluid Overload, Pulse Wave Velocity, and Ratio of Brachial Pre-Ejection Period to Ejection Time in Diabetic and Non-Diabetic Chronic Kidney Disease
Fluid overload is one of the characteristics in chronic kidney disease (CKD). Changes in extracellular fluid volume are associated with progression of diabetic nephropathy. Not only diabetes but also fluid overload is associated with cardiovascular risk factors The aim of the study was to assess the interaction between fluid overload, diabetes, and cardiovascular risk factors, including arterial stiffness and left ventricular function in 480 patients with stages 4-5 CKD. Fluid status was determined by bioimpedance spectroscopy method, Body Composition Monitor. Brachial-ankle pulse wave velocity (baPWV), as a good parameter of arterial stiffness, and brachial pre-ejection period (bPEP)/brachial ejection time (bET), correlated with impaired left ventricular function were measured by ankle-brachial index (ABI)-form device. Of all patients, 207 (43.9%) were diabetic and 240 (50%) had fluid overload. For non-diabetic CKD, fluid overload was associated with being female (b = -2.87, P = 0.003), heart disease (b = 2.69, P = 0.04), high baPWV (b = 0.27, P = 0.04), low hemoglobin (b = -1.10, P,0.001), and low serum albumin (b = -5.21, P,0.001) in multivariate analysis. For diabetic CKD, fluid overload was associated with diuretics use (b = 3.69, P = 0.003), high mean arterial pressure (b = 0.14, P = 0.01), low bPEP/ET (b = -0.19, P = 0.03), low hemoglobin (b = -1.55, P = 0.001), and low serum albumin (b = -9.46, P,0.001). In conclusion, baPWV is associated with fluid overload in non-diabetic CKD and bPEP/bET is associated with fluid overload in diabetic CKD. Early and accurate assessment of these associated cardiovascular risk factors may improve the effects of entire care in late CKD.
# Introduction
Cardiovascular disease (CVD) is the major cause of morbidity and mortality in patients with chronic kidney disease (CKD). The presentation of fluid overload is often noticed in patients with CKD, and excess fluid status induces elevated arterial pressure, left ventricular hypertrophy, and associated cardiovascular sequelae [bib_ref] Dilemma of assessing volume state-the use and the limitations of a clinical..., Wizemann [/bib_ref] [bib_ref] Influence of hydration state on plasma volume changes during ultrafiltration, Wizemann [/bib_ref]. Hung et al. indicated a significant association of fluid overload with cardiovascular risk factors, such as diabetes, systolic blood pressure, and arterial stiffness in CKD patients not on dialysis [bib_ref] Volume overload correlates with cardiovascular risk factors in patients with chronic kidney..., Hung [/bib_ref]. Previous studies reported that fluid overload was a predictor of cardiovascular mortality in patients on dialysis [bib_ref] Longer treatment time and slower ultrafiltration in hemodialysis: associations with reduced mortality..., Saran [/bib_ref] [bib_ref] The mortality risk of overhydration in haemodialysis patients, Wizemann [/bib_ref] [bib_ref] NT-proBNP, fluid volume overload and dialysis modality are independent predictors of mortality..., Paniagua [/bib_ref] [bib_ref] Association between high ultrafiltration rates and mortality in uraemic patients on regular..., Movilli [/bib_ref]. Fluid overload is not only a characteristic but also a clinical indicator of cardiovascular burden.
On the other hand, diabetic CKD patients have a greater risk of commencing dialysis, and higher all-cause and cardiovascular mortality than non-diabetic CKD patients [bib_ref] Chronic Kidney Disease Prognosis Consortium: Associations of kidney disease measures with mortality..., Fox [/bib_ref]. This is probably the result that more advanced atherosclerotic change of vascular or cardiac level in diabetic CKD or vascular disease in nondiabetic CKD is not necessarily atherosclerotic. Additionally, diabetics are more likely to have fluid overload than non-diabetics [bib_ref] Is fluid overload more important than diabetes in renal progression in late..., Tsai [/bib_ref] , and progression of diabetic nephropathy would contribute to the increase in extracellular fluid volume [bib_ref] Disassociation between glomerular hyperfiltration and extracellular volume in diabetic rats, Tucker [/bib_ref]. An interaction between fluid overload, diabetes, and vascular injury or cardiac dysfunction might exist in CKD. Accumulating evidence shows that pulse wave velocity (PWV), which can be easily measured by a clinical device, the ankle-brachial index (ABI)-form, has been regarded as a clinical indicator of arterial stiffness [bib_ref] Pulse wave velocity in lower-limb arteries among diabetic patients with peripheral arterial..., Yokoyama [/bib_ref] [bib_ref] Validity, reproducibility, and clinical significance of noninvasive brachial-ankle pulse wave velocity measurement, Yamashina [/bib_ref]. Cardiac dysfunction is frequently evaluated by echocardiography; however, its application in predicting cardiovascular events is limited because echocardiography is time-consuming and operatordependent [bib_ref] A systolic parameter defined as the ratio of brachial pre-ejection period to..., Chen [/bib_ref]. The ratio of brachial pre-ejection period (bPEP) and brachial ejection time (bET), measured easily by ABI device, was reported to have a significant correlation with impaired left ventricular systolic function [bib_ref] Myocardial performance index derived from brachial-ankle pulse wave velocity: A novel and..., Su [/bib_ref]. Chen et al. found that bPEP/ bET was an independent predictor for all-cause and cardiovas-cular mortality in CKD patients on or not on dialysis [bib_ref] A systolic parameter defined as the ratio of brachial pre-ejection period to..., Chen [/bib_ref] [bib_ref] A new systolic parameter defined as the ratio of brachial pre-ejection period..., Chen [/bib_ref]. Hence, the aim of this study is to evaluate the relationship between fluid overload, diabetes, and baPWV or bPEP/bET and whether baPWV or bPEP/bET could be used as simple clinically available measures for risk stratification in late CKD.
# Materials and methods
## Study participants
All 612 of CKD stages 4-5 patients were invited to participate in the study from January 2011 to December 2011 at one hospital in Southern Taiwan. The study protocol was approved by the Institutional Review Board of the Kaohsiung Medical University Hospital. All patients had been enrolled in our integrated CKD program for more than 3 months (30.9627.0 months). CKD was staged according to K/DOQI definitions and the estimated glomerular filtration rate (eGFR) was calculated using the equation of the 4-variable Modification of Diet in Renal Disease (MDRD) Study (CKD stage 3, eGFR: 30,59 ml/min/1.73 m 2 ; CKD stage 4, eGFR: 15,29 ml/min/1.73 m 2 ; CKD stage 5, eGFR,15 ml/ min/1.73 m 2 ). Of all patients, we excluded 115 with disabilities, 12 with impaired skin integrity, and 5 with pacemaker implantation. Four hundred and eighty patients were enrolled and scheduled for a study interview after informed consent.
# Ethics statement
The study protocol was approved by the Institutional Review Board of the Kaohsiung Medical University Hospital (KMUH-IRB-990125). Informed consents were obtained in written form from patients and all clinical investigations were conducted according to the principles expressed in the Declaration of Helsinki. The patients gave consent for the publication of the clinical details.
## Measurement of fluid status
Fluid status was measured once by a bioimpedance spectroscopy method, Body Composition Monitor (BCM, Fresenius Medical Care) at enrollment. BCM has been validated extensively against all available gold-standard methods in the general and dialysis populations [bib_ref] Importance of whole-body bioimpedance spectroscopy for the management of fluid balance, Wabel [/bib_ref] [bib_ref] Application of body composition monitoring to peritoneal dialysis patients, Crepaldi [/bib_ref] [bib_ref] Body fluid volume determination via body composition spectroscopy in health and disease, Moissl [/bib_ref]. The information of normohydrated lean tissue, normohydrated adipose tissue, and extracellular fluid overload in whole body based on the difference of impedance in each tissue is provided by BCM. Fluid overload can be calculated from the difference between the normal expected and measured extracellular water (ECW) [bib_ref] Effect of fluid management guided by bioimpedance spectroscopy on cardiovascular parameters in..., Hur [/bib_ref]. Fluid overload value, ECW, intracellular water (ICW) and total body water (TBW) were determined from the measured impedance data based on the model of Moissl et al [bib_ref] Bioimpedance-based identification of malnutrition using fuzzy logic, Wieskotten [/bib_ref] [bib_ref] Fluid status in peritoneal dialysis patients: the European Body Composition Monitoring (EuroBCM)..., Van Biesen [/bib_ref]. The relative hydration status (gHS = fluid overload/ECW) has been used as an indicator of fluid status in a previous study [bib_ref] The mortality risk of overhydration in haemodialysis patients, Wizemann [/bib_ref]. In the present study, ''Fluid overload'' was defined as DHS of 7% or more corresponding to the value of 90 th percentile for the normal reference population when the fluid status was measured with the same technology [bib_ref] Volume overload correlates with cardiovascular risk factors in patients with chronic kidney..., Hung [/bib_ref] [bib_ref] Bioimpedance-based identification of malnutrition using fuzzy logic, Wieskotten [/bib_ref] [bib_ref] Fluid status in peritoneal dialysis patients: the European Body Composition Monitoring (EuroBCM)..., Van Biesen [/bib_ref].
## Assessment of bapwv, bpep, and bet
Brachial-ankle pulse wave velocity (baPWV), as a good parameter of arterial stiffness, was measured by the ABI-form device, which automatically and simultaneously measured blood pressures in both arms and ankles using an oscillometric method [bib_ref] Pulse wave velocity in lower-limb arteries among diabetic patients with peripheral arterial..., Yokoyama [/bib_ref] [bib_ref] Validity, reproducibility, and clinical significance of noninvasive brachial-ankle pulse wave velocity measurement, Yamashina [/bib_ref] , at the same time of fluid measurement for each patient. For measuring PWV, pulse waves obtained from the brachial and tibial arteries were recorded simultaneously, and the transmission time (DTba), which was defined as the time interval between the initial increase in brachial and ankle waveforms, was determined. The transmission distance from the brachium to ankle was calculated according to body height. The path length from the suprasternal notch to the brachium(Lb) was obtained using the following equation: Lb = 0.21956height of the patient (in cm)-2.0734. The path length from the suprasternal notch to the ankle(La) was obtained using the following equation: La = (0.81296height of the patient [in cm]+12.328). Finally, the following equation was used to automatically obtain baPWV: baPWV = (La-Lb)/DTba [bib_ref] Decrease in ankle-brachial index over time and cardiovascular outcomes in patients with..., Chen [/bib_ref]. After obtaining bilateral PWV values, the higher one was used as representative for each subject. Additionally, bPEP, and bET was also measured by the ABI-form device.
The bET was automatically measured from the foot to the dicrotic notch (equivalent to the incisura on the downstroke of the aortic pressure wave contour produced by the closure of aortic valve) of the pulse volume waveform. The total electromechanical systolic interval (QS2) was measured from the onset of the QRS complex on the electrocardiogram to the first high-frequency vibration of the aortic component of the second heart sound on the phonocardiogram. The bPEP was also automatically calculated by subtracting the bET from the QS2 [bib_ref] A systolic parameter defined as the ratio of brachial pre-ejection period to..., Chen [/bib_ref] [bib_ref] A new systolic parameter defined as the ratio of brachial pre-ejection period..., Chen [/bib_ref] [bib_ref] Significant correlation between ratio of brachial pre-ejection period to ejection time and..., Chen [/bib_ref].
## Data collection
Demographic and clinical data of patients were obtained from interviews and medical records at enrollment. Blood and urine samples were obtained at the same time of fluid status measurement. Patients were asked to fast for at least 12 hours before blood sample collection for the biochemistry study. Protein in urine was measured using an immediate semiquantitative urine protein dipstick test and graded as negative, trace, 1+, 2+, 3+, or 4+. The body mass index was calculated as the ratio of weight in kilograms divided by square of height in meters. Blood pressure was recorded as the mean of two consecutive measurements with 5-minute intervals, using one single calibrated device. Mean arterial pressure was calculated as 2/3 diastolic blood pressure plus 1/3 systolic blood pressure. Diabetes was defined as those with a medical history through chart review. Cardiovascular disease was defined as a history of heart failure, acute or chronic ischemic heart disease, and myocardial infarction. Information regarding patient's medications including diuretics, HMG-CoA reductase inhibitors (statins), anti-hypertensive drugs, including calcium channel blocker, b-blocker, and angiotensin converting enzyme inhibitor, and angiotensin II receptor blocker within 3 months before enrollment was obtained from medical records.
# Statistical analysis
The study population was further classified into two groups according to the presence or absence of diabetes. Continuous variables were expressed as mean 6 SD or median (25 th , 75 th percentile), as appropriate, and categorical variables were expressed as percentages. Skewed distribution continuous variables were log-transformed to attain normal distribution. The significance of differences in continuous variables between groups was tested using independent t-test or the Mann-Whitney U analysis, as appropriate. The difference in the distribution of categorical variables was tested using the Chi-square test. Linear regression models were utilized to evaluate the determinants of fluid overload in diabetes and non-diabetes. All the variables in [fig_ref] Table 1: The clinical characteristics of study subjects stratified by Diabetes Mellitus [/fig_ref] tested by univariate analysis and those variables with Pvalue less than 0.05 were selected in multivariate analysis. Statistical analyses were conducted using SPSS 18.0 for Windows (SPSS Inc., Chicago, Illinois) and some graphs were made by GraphPad Prism 5.0 (GraphPad Software Inc., San Diego CA, USA). Statistical significance was set at a two-sided p-value of less than 0.05.
# Results
## Characteristics of entire cohort
The comparison of clinical characteristics between groups based on the presence or absence of diabetes is shown in [fig_ref] Table 1: The clinical characteristics of study subjects stratified by Diabetes Mellitus [/fig_ref]. A total of 211 (44.0%) had diabetes. The mean age was 65.4612.7 years. Higher baPWV, and lower bPEP/bET were found in diabetic patients than non-diabetic patients. Uric acid levels and the degree of proteinuria were higher and serum albumin was lower in diabetic patients than in non-diabetic patients. Different zones can be identified in the plot of mean arterial pressure (Y-axis) versus relative hydration status (DHS) (X-axis), of normovolemic and normotensive patients (40.9%, zone A), fluidoverloaded and hypertensive patients (14.0%, zone B), fluidoverloaded but normo-or hypotensive patients (36.0%, zone C), hypertensive and normovolemic patients (9.1%, zone D), in [fig_ref] Figure 1: Scatter plot of brachial-ankle pulse wave velocity [/fig_ref]. [fig_ref] Figure 1: Scatter plot of brachial-ankle pulse wave velocity [/fig_ref] shows a substantial scatter of baPWV quartiles, and the proportion of baPWV quartile 4 was the highest in zone B (36.4%) than other zones (P,0.001). [fig_ref] Figure 1: Scatter plot of brachial-ankle pulse wave velocity [/fig_ref] shows that the proportion of diabetes was higher in zone B (59.7%) and C (51.7%) than other zones (P,0.001). There was no significant difference of gender distribution among zones [fig_ref] Figure 1: Scatter plot of brachial-ankle pulse wave velocity [/fig_ref]. [fig_ref] Table 2: The determinants of relative hydration status in non-diabetic chronic kidney disease patients [/fig_ref]. DHS correlated positively with age, heart disease, diuretics use, anti-hypertension drug use, baPWV, and urine protein, but negatively with the male gender, body mass index, eGFR, serum albumin, cholesterol, and hemoglobin levels in univariate linear regression analysis. Further multivariate analysis showed that increased DHS was associated with being female (b = -2.87, P = 0.003), heart disease (b = 2.69, P = 0.04), high baPWV (b = 0.27, P = 0.04), low hemoglobin (b = -1.10, P, 0.001), and low serum albumin (b = -5.21, P,0.001). [fig_ref] Table 3: The determinants of relative hydration status in diabetic chronic kidney disease patients [/fig_ref]. DHS correlated positively with diuretics use, anti-hypertensive drug use, mean arterial pressure, and urine protein, but negatively with age, male, body mass index, bPEP/bET, eGFR, serum albumin, and hemoglobin levels in univariate linear regression analysis. Further multivariate analysis showed that increased DHS was associated with diuretics use (b = 3.69, P = 0.003), high mean arterial pressure (b = 0.14, P = 0.01), low bPEP/bET (b = -0.19, P = 0.03), low hemoglobin (b = -1.55, P = 0.001), and low serum albumin (b = -9.46, P,0.001).
## Determinants of fluid overload (relative hydration status, dhs) in non-diabetic ckd
## Determinants of fluid overload (relative hydration status, dhs) in diabetic ckd
# Discussion
The aim of this study was to evaluate the relationship between fluid overload, diabetes and baPWV or bPEP/bET in patients with stages 4-5 CKD. Fluid overload was associated with baPWV, a clinical indicator of arterial stiffness, in non-diabetic CKD. Conversely, instead of baPWV, fluid overload was associated with bPEP/bET, a marker of left ventricular systolic function, in diabetic CKD.
Fluid overload may have an influence on vasculature and lead to vascular remodeling, characterized by dilatation of muscular and elastic type arteries and increased wall thickness, thereby inducing arterial stiffness and consequent cardiovascular sequelae. Previous reports demonstrated that baPWV was associated with fluid overload in CKD [bib_ref] Volume overload correlates with cardiovascular risk factors in patients with chronic kidney..., Hung [/bib_ref] [bib_ref] Association of Fluid Overload With Kidney Disease Progression in Advanced CKD: A..., Tsai [/bib_ref]. On the other hand, diabetes has profound effects on vasculature, and baPWV appears to be associated with clinical variants of diabetes and microvascular and macrovascular complications [bib_ref] Preferential stiffening of central over peripheral arteries in type 2 diabetes, Kimoto [/bib_ref] [bib_ref] Central versus peripheral arterial stiffness in association with coronary, cerebral and peripheral..., Tsuchikura [/bib_ref] [bib_ref] The association between regional arterial stiffness and diabetic retinopathy in type 2..., Kim [/bib_ref]. There might be an interaction between fluid overload, baPWV, and diabetes. In the Hence, we divided all subjects into non-diabetes and diabetes groups to analyze the relationship between fluid overload and baPWV respectively and found a significant association between fluid overload and baPWV in non-diabetic CKD, not in diabetic CKD. Interestingly, instead of baPWV, low bPEP/bET, which has a significant correlation with impaired left ventricular ejection fraction [bib_ref] Myocardial performance index derived from brachial-ankle pulse wave velocity: A novel and..., Su [/bib_ref] , is correlated with fluid overload in diabetic CKD patients. This study also used the echocardiographic examination to evaluate cardiac function in 184 subjects and found that left ventricular ejection fraction was significantly correlated with fluid overload in diabetic CKD (b = -0.21, P = 0.002), not in nondiabetic CKD (b = -0.02, P = 0.4). These findings suggest that fluid overload may have different levels of influences on CKD patients depending on the presence or absence of diabetes. Fluid overload is correlated with vascular level alterations in nondiabetic CKD, and with cardiac level modifications in diabetic CKD.
Cardiovascular dysfunction progresses with arterial-cardiac interactions [bib_ref] Arterial disease in chronic kidney disease, Moody [/bib_ref]. The arterial-cardiac compensatory adaptations maintain cardiac performance with enhanced contractility [bib_ref] Coupled systolic-ventricular and vascular stiffening with age: implications for pressure regulation and..., Chen [/bib_ref]. Fluid overload alters and blunts the arterial-cardiac response, leading to hemodynamic instability [bib_ref] Coupled systolic-ventricular and vascular stiffening with age: implications for pressure regulation and..., Chen [/bib_ref]. The difference of rates of progression of dysfunction between the heart and the vasculature may explain our results. Probably, the phenomenon of fluid overload directly affecting cardiac function beyond vasculature in diabetic CKD exists. Arterial stiffness may occur in early diabetic CKD and then fluid overload may subsequently have effect on left ventricular systolic dysfunction in late diabetic CKD. Further study is needed to evaluate the mechanisms between fluid overload, arterial stiffness and cardiac dysfunction in diabetic CKD.
Due to the strong association of fluid overload and adverse outcomes [bib_ref] Association of Fluid Overload With Kidney Disease Progression in Advanced CKD: A..., Tsai [/bib_ref] , precise measurement of fluid status is important in clinical practice of CKD patients. Traditional physical examination is not enough to detect slight variations of fluid status. Accumulating evidence suggests that multifrequency spectroscopic bioimpedance can provides information of increases in fluid status and associated body composition. Besides, using a non-invasive ABI-form device, clinicians can easily obtain the value of baPWV, a reliable marker of arterial stiffness, and bPEP/bET, a surrogate of left ventricular systolic function. These inexpensive and convenient tools might assist clinicians in assessing the risk of renal progression and cardiovascular morbidity and mortality earlier.
The present study has some limitations that must be considered. This study was conducted at a single center. Fluid status, baPWV, bPEP/bET, clinical parameters, and the use of drugs were measured only once at enrollment. The association of time-varying baPWV, bPEP/bET, clinical parameters, and the use of drugs with time-varying fluid status could not be estimated. Additionally, our findings show the relative weak correlation between bPEP/ bET and fluid overload in diabetic CKD. The weak correlation between fluid overload and bPEP/bET is probably related to the relatively small sample size of diabetic patients. Based on the association of bPEP/bET with cardiac function, we performed subgroup analysis to analyze the relationship between bPEP/bET and fluid overload in the heart disease group and the result was consistent (b = -0.46, P = 0.03). However, there was no significant correlation between bPEP/bET and fluid overload in the nonheart disease group. We need a large population study to evaluate the interaction between bPEP/bET, cardiac function and diabetes in a CKD cohort.
In conclusion, our study evaluates a relationship between fluid overload, diabetes, and arterial stiffness or cardiac function in late CKD patients. Fluid overload is correlated with arterial stiffness in non-diabetic CKD, and with left ventricular dysfunction in diabetic CKD. Fluid overload probably has distinct arterialcardiac influences on CKD patients in the presence or absence of diabetes. Early and accurate assessment of these associated cardiovascular risk factors may improve the effects of entire care in late CKD patients.
[fig] Figure 2: shows a positive association of baPWV and fluid overload in non-diabetic CKD (r = 0.488, P,0.001). The determinants of fluid overload in non-diabetic patients are reported in [/fig]
[fig] Figure 3: shows a negative association of bPEP/bET and fluid overload in diabetic CKD (r = -0.251, P = 0.01). The determinants of fluid overload in diabetic patients are reported in [/fig]
[fig] Figure 1: Scatter plot of brachial-ankle pulse wave velocity (1A), diabetes (1B), and gender (1C) between relative hydration status (%) in the X-axis and mean arterial pressure (mmHg) in the Y-axis in all study subjects. doi:10.1371/journal.pone.0111000.g001Fluid Overload in Diabetic and Non-Diabetic CKD PLOS ONE | www.plosone.org [/fig]
[table] Table 1: The clinical characteristics of study subjects stratified by Diabetes Mellitus. [/table]
[table] Table 2: The determinants of relative hydration status in non-diabetic chronic kidney disease patients. [/table]
[table] Table 3: The determinants of relative hydration status in diabetic chronic kidney disease patients. [/table]
|
Results from a difference‐in‐differences evaluation of health facility HIV and key population stigma‐reduction interventions in Ghana
Introduction: Stigma undermines all aspects of a comprehensive HIV response, as reflected in recent global initiatives for stigma-reduction. Yet a commensurate response to systematically tackle stigma within country responses has not yet occurred, which may be due to the lack of sufficient evidence documenting evaluated stigma-reduction interventions. With stigma present in all life spheres, health facilities offer a logical starting point for developing and expanding stigma reduction interventions. This study evaluates the impact of a "total facility" stigma-reduction intervention on the drivers and manifestations of stigma and discrimination among health facility staff in Ghana. Methods: We evaluated the impact of a total facility stigma-reduction intervention by comparing five intervention to five comparable non-intervention health facilities in Ghana. Interventions began in September 2017. Data collection was in June 2017 and April 2018. The primary outcomes were composite indicators for three stigma drivers, self-reported stigmatizing avoidance behaviour, and observed discrimination. The principal intervention variable was whether the respondent worked at an intervention or comparison facility. We estimated intervention effects as differences-in-differences in each outcome, further adjusted using inverse probability of treatment weighting (IPTW). Results: We observed favourable intervention effects for all outcome domains except for stigmatizing attitudes. Preferring not to provide services to people living with HIV (PLHIV) or a key population member improved 11.1% more in intervention than comparison facility respondents (95% CI 3.2 to 19.0). Other significant improvements included knowledge of policies to protect against discrimination (difference-in-differences = 20.4%; 95% CI 12.7 to 28.0); belief that discrimination would be punished (11.2%; 95% CI 0.2 to 22.3); and knowledge of and belief in the adequacy of infection control policies (17.6%; 95% CI 8.3 to 26.9). Reported observation of stigma and discrimination incidents fell by 7.4 percentage points more among intervention than comparison facility respondents, though only marginally significant in the IPTW-adjusted model (p = 0.06). Respondents at intervention facilities were 19.0% (95% CI 12.2 to 25.8) more likely to report that staff behaviour towards PLHIV had improved over the last year than those at comparison facilities. Conclusions: These results provide a foundation for scaling up health facility stigma-reduction within national HIV responses, though they should be accompanied by rigorous implementation science to ensure ongoing learning and adaptation for maximum effectiveness and long-term impact.
through interpersonal acts of discrimination that can range from verbal and physical abuse to social isolation and gossip . Perceived stigma is the perception of the prevalence of stigmatizing attitudes and behaviours, for example within the general community or in a health facility . Anticipated stigma is the fear that stigma will happen, if for example a family member learns a person is living with HIV, whether or not it is actually experienced. Perceived and anticipated stigma are sometimes described together as felt stigma. Individuals facing stigma may also internalize stigma. Intersecting stigma occurs when individuals who have identities linked to multiple marginalized groups, for example based on socio-economic status, race, sexual orientation, gender identity or health conditions face multiple intersecting stigmas . Embedded within human rights imperatives, recognition of the necessity and urgency of responding to stigma and resulting discrimination is critical to achieving global targets for HIV testing, linkage to care and viral suppression. As stated in the Global Partnership for Action to Eliminate All Forms of HIV-Related Stigma and Discrimination:"Without addressing HIV-related stigma and discrimination, the world will not achieve the goal of ending AIDS as a public health threat by 2030" (p. 5).
Stigma occurs and needs to be addressed in all life spheres and institutionsin the home, community, workplace, places of worship, schools, health care facilities, etc. The health facility offers a logical place to develop and scale-up a national response to stigma for multiple reasons. It is, in most places, the gateway into HIV treatment and often prevention. As a chronic condition, HIV requires lifelong engagement with the health system. In addition, health care workers are members of their communities and are often looked to for guidance on health and other issues. Health workers are a potentially powerful force not only to reduce stigma in their health facilities, but to be change agents in their homes and communities. Stigma also undermines the health workforce, impacting the health and wellbeing of health workers. The importance of tackling stigma in health facilities is underscored by the UNAIDS-led global Agenda for Zero Discrimination in Health-care Settingsand corresponding policy guidance . Global initiatives like the Fast Track Cities have also made addressing stigma a key pillar of their response, including a focus on health facilities.
The prevalence, formsand consequencesof health facility HIV stigma are well documented across the globe, as are key drivers of that stigma, including fear of contracting HIV in the workplace, lack of awareness and understanding of stigma, attitudes and the health facility institutional environment ]. Yet despite this evidence and global recognition of the need to tackle stigmaespecially in health facilitiesthere is little concrete evidence of concerted efforts by countries and donors to scale-up stigma-reduction interventions in health facilities within national HIV responses. This may be in part due to the still nascent body of evaluated intervention approaches to stigma-reduction in health facilities.
In Ghana in particular, stigma and discrimination remain a pervasive issue, as documented both through measurement of stigmatizing attitudes in the general populationand through the experiences of people living with HIV (PLHIV) . In 2014, only 8.0% of women and 14.1% of men in the general population expressed accepting attitudes as measured by four standardized indicators (caring for a family member in one's home, buying fresh vegetables from PLHIV, allowing a teacher living with HIV to continue working, and keeping a family member's positive HIV status secret). PLHIV reported that experienced stigma and discrimination was significant, including social exclusion and gossip, while noting that fear of transmission through casual contact was a key driver of stigma. A health facility stigma assessment in Ghanaidentified the physical layout of healthcare services; lack of in-depth knowledge about HIV; fear of contracting the disease; and low levels of respect and dignity based on cultural and moral grounds as contributing factors for stigma and discrimination experienced by PLHIV in health facilities. HIV stigma in Ghana has been documented as a barrier to antiretroviral therapy (ART) services; voluntary testing and counselling (VCT) uptake; and to parents informing children living with HIV of their seropositive status.
In its contribution to literature outlining potential interventions within health facility stigma-reduction, this study evaluates the impact of a "total facility" stigma-reduction intervention on the drivers and manifestations of stigma and discrimination among health facility staff in Ghana.
# | methods
## | setting and intervention
We evaluated the impact of a total facility stigma-reduction intervention in five health facilities in Ghana compared to five non-intervention facilities selected to be broadly comparable with regard to facility size (staff complement and patient load); subnational region; and type of facility (district-level hospitals). The initial phase of the study collected stigma data from the four highest HIV-caseload civilian facilities in five regions -Ashanti, Brong Ahafo, Eastern, Greater Accra and Westernfor a total of 20 facilities. From these, a subset of 10 facilities were selected for the second phase, with one intervention facility and one comparison facility selected from each region based on staff size (medium-sized, district-level facilities) and facility interest to participate. There was no randomization of facilities.
After baseline data collection in June 2017 from health facility staff, we conducted participatory data dissemination and review workshops with staff from all 20 first-phase facilities in August/September 2017. In these workshops, staff identified stigma and discrimination challenges and then generated potential solutions. Non-intervention facilities received no further intervention via the study during this period, though some facilities independently conducted stigma-reduction activities. After the evaluation reported in this manuscript, the non-intervention facilities are in the process of receiving the intervention approach through scale-up led by a local Ghanaian organization with support from The Global Fund.
Targeting the whole facility beyond HIV services, the intervention approach included a two-day participatory stigma-reduction training for all staff levels (clinical and non-clinical) with delivery by staff and clients from the facilities who were trained as stigma-reduction facilitators. Trainings were delivered between September and November 2017 to all categories of staff, with a target of reaching 70% of the facility workforce. The trainings were based on pre-existing global training materials that are available online. These are targeted towards actionable HIV stigma drivers and were adapted to the Ghanaian context through an in-country stakeholder workshop. Final Ghanaian training materials are available online . Ultimately, 1228 staff members were trained -79% of staff at intervention facilities (with individual facilities ranging from 61% to 97% training coverage). Each facility also created an eight to ten member "champion team, " which was provided $5000 USD to develop facility-specific ancillary activities, including launch events, anti-stigma-and-discrimination banners and posters, additional staff trainings, printed codes of ethics, reporting mechanisms, and staff nametags to enable identification and reporting of stigma and discrimination. More detail on the intervention is available online in a final project report . We conducted follow-up surveys in April 2018 and disseminated results to staff at the 10 study facilities as well as at a national dissemination meeting for policymakers, donors and community stakeholders.
## | sampling
We conducted repeated cross-sectional surveys, stratified by health facility and job category, of a random sample of staff at the five intervention and five comparison facilities two to three months before and five to six months after the training portion of the intervention. The total sample size was 2308 participants (1154 in each of the pre-and post-intervention periods). To preserve anonymity, we did not attempt to identify the same participants for the follow-up survey. The sample was designed to provide 80% power to detect a 10-percentage point change in the only indicator for which pre-baseline estimates were available: observation of colleagues gossiping about PLHIV. The sampling frame consisted of a complete listing of all staff from which an off-site statistician generated the target sample. If respondents were unavailable or declined to participate, substitutes were chosen in order from a predetermined list. The survey was designed to be self-administered if participants could read and to be administered by interviewers to participants who could not read or preferred the administered survey.
## | instruments and variables
Survey instruments are provided in Data S1. The survey included modules to capture demographic information; three actionable stigma and discrimination drivers (perceptions of health facility policies, fears of contracting HIV in the workplace, and attitudes towards PLHIV and key populations); and stigma and discrimination manifestations (self-reported unnecessary infection control behaviours and observed discrimination by other healthcare workers). The follow-up survey also included items about exposure to the intervention and perceived changes since baseline.
Survey questions were developed by adapting a globally validated tool for health worker stigma and discrimination . Items were adapted during a one-day participatory workshop with Ghanaian collaborators and other stakeholders and then tested during a rapid pilot at one health facility which had not been selected for further surveys. Surveys were administered in English, Dangme, Akuapem Twi and Ga. English surveys were translated by professional translators and then checked by mother-tongue speakers on the research term and any issues resolved with the translators.
The primary outcomes outlined below were composite indicators for each domain, dichotomized from a series of items.
## | facility policies
One variable for reporting knowledge that facility guidelines protect PLHIV, men who have sex with men (MSM), and sex workers from discrimination; one variable for believing that the respondent would be punished if discriminating against PLHIV and all of four key populations (MSM, sex workers, sexually active adolescents and persons who inject drugs); and one variable for agreeing or strongly agreeing that the facility has adequate infection control policies and personal protective equipment and knowing that the facility has both a posted post-exposure prophylaxis (PEP) policy and PEP access.
## | fear
Reporting being a little worried, worried, or very worried to conduct one or more of four care activities with clients living with HIV, for fear of HIV acquisition.
## | attitudes
Three variables for agreeing or strongly agreeing with one or more stigmatizing statement about PLHIV in general, women living with HIV, and MSM; and one variable for agreeing or strongly agreeing that the respondent would prefer not to provide services to one or more key population.
## | self-reported provision of care in a stigmatizing manner
Reporting usually avoiding physical contact, wearing double gloves, wearing gloves during all aspects of care, or using extra precautions with PLHIV but not other patients.
## | observed stigma and discrimination
Observing one or more instances of healthcare workers at the same facility being unwilling to care for, providing poorer care quality to, talking badly about, or disclosing status about PLHIV or one of four key populations in the last six months.
A secondary outcome was participant's general perception about whether behaviour towards clients had changed in the facility over the year prior to the endline survey, dichotomized as a little/much better versus no change or a little/much worse.
The principal intervention variable was whether the respondent worked at an intervention or comparison facility. Analyses were by intention to treat, so participants were classified as at intervention facilities even if they did not individually receive training, as intention-to-treat analysis was expected to capture the effect of average programme implementation. We also collected demographic variables for use in adjusted analyses: age (18 to 24 years, 10 year increments from 24 to 54, and 55+); region; sex; staff category (senior medical, other medical, and administrative or support staff); tenure at the facility (<2 years, two to five years, 5+ years); whether the respondent has ever worked in an HIV-specific department; and quintiles of the number of PLHIV the respondent reports personally providing care to in the past month.
## | statistical analyses
We present basic descriptive statistics of the sample in. In, we report the number and proportion of respondents in the follow-up sample who reported exposure to the main intervention components: trainings in infection control; patients' rights to informed consent, privacy, and confidentiality; HIV stigma and discrimination; and key population stigma and discrimination; or who participated in any other stigma and discrimination reduction activity since August 2017 when the intervention began.
Intervention effects are presented inas differences-in-differences: the difference in before-to-after changes between intervention and comparison areas for each composite indicator. To estimate differences-in-differences, we fit saturated linear probability models using generalized linear models with an identity link function and binomial error distribution. Each outcome was regressed on indicator variables for intervention (intervention vs. comparison); time (pre vs. post-intervention); and their interaction, with the coefficient on the interaction term being the estimated difference-in-differences (DID). Primary analyses were conducted on the full sample, and secondary analyses were restricted only to clinical providers. All analyses were by intention-to-treat, and we treated facilities designated as comparison but which partially implemented as comparison facilities in the analysis.
In addition to unadjusted estimates, we estimate intervention effects using inverse probability of treatment weighting (IPTW). The IPTW models generated weights to balance the following covariates across the four intervention-bytime groups: subnational region and staff members' age, sex, employment category, employment tenure, work history in an HIV-specific department and number of PLHIV personally treated. Separate IPT weights were constructed for the full sample and the clinical staff subsample. We then fit DID models identical to the primary analyses except incorporating IPT weights. We provide full details of the IPTW approach and balance diagnostics in Data S2. All analyses adjust standard errors for clustering by facility.
In the main analysis, we did not adjust for multiple comparisons because all outcomes were theoretically informed and predetermined and no analyses were added post hoc. Because some readers might expect adjustment; however, we present an analysis in Data S4 that reports risk of false discovery because of multiple outcomes. We present false discovery rates as q-values produced via Simes' method. We used Stata v15.1 for all analyses and statistical code and output are provided in Data S3.
The institutional review boards at Health Media Labs and the Ghana Health Services Ethical Review Board approved the study, and all participants provided written informed consent.
# | results
Characteristics of the analytical sample of 2308 participants (1154 in each of the pre-and post-intervention periods) are in. They are broadly comparable across the four intervention-by-time groups, and balance of characteristics before and after IPT weighting is provided in Data S2. IPT-weighted models met generally accepted criteria for balance of potential confounders.
As expected, exposure to trainings and other activities to reduce stigma and discrimination was higher among intervention than comparison staff during the post-intervention period, with each of the trainings reaching at least three-quarters of intervention respondents, and about two-thirds being exposed to at least one additional intervention activity. More than half of comparison respondents reported receiving training on infection control and universal precautions (65%) and informed consent, privacy, and confidentiality (57%) during the same time period. A lower but still substantial percentage of comparison respondents reported receiving HIV stigma and discrimination training (45%), key population stigma and discrimination training (33%) and other relevant activities (32%).
Using intention-to-treat analyses, we observed favourable intervention effects for all outcome domains except for stigmatizing attitudes in both the full sample and the medical staff-only subsample. In the IPTW-adjusted analyses of the full sample, preferring not to provide services to PLHIV or a key population member improved by 11.1 percentage points (95% CI 3.2 to 19.0) more in intervention than comparison facility respondents. Knowledge of policies to protect against discrimination (DID = 20.4 percentage points [pp], 95% CI 12.7 to 28.0); belief that discrimination would be punished (DID = 11.2pp, 95% CI 0.2 to 22.3); and knowledge of and belief in the adequacy of infection control policies (DID = 17.6pp, 95% CI 8.3 to 26.9) all improved significantly. Observing stigma and discrimination incidents improved by 7.4 percentage points more among intervention than comparison facility respondents (95% CI À 0.3 to 15.2), though this result was only marginally significant (p = 0.06). Results were very similar in the unadjusted analyses.
We found comparable improvements in the subsample composed only of medical staff. Here too, there were no significant improvements in stigmatizing attitudes, but all other outcome domains improved significantly with similar intervention effect estimates to those identified in the full sample. In addition to the domains measured in the full sample, two domains were limited only to medical staff. Fear of conducting at least one care activity improved by 24.7 more percentage All estimates are constructed such that a positive DID is an improvement (even if the original measure was scaled so that higher numbers were an undesirable outcome). Fear and avoidance outcomes are restricted only to clinical staff because the sizable majority of non-clinical staff answered "not applicable" to the single item applicable to them. DID, difference-in-differences; IPTW, inverse probability of treatment weighting; MSM, men who have sex with men; PEP, post-exposure prophylaxis.
When asked whether there had been a change in behaviour towards PLHIV over the last year, intervention facility respondents were 19.0 percentage points (95% CI 12.2 to 25.8) more likely to say it was better in IPTW-adjusted analyses. Results were similar in the medical staff subpopulation (18.0pp; 95% CI 11.2 to 24.9) and in unadjusted analyses. Substitution of false discovery rate q-values for p-values does not meaningfully change conclusions.
# | discussion
Evidence continues to grow of the negative relationship of stigma to pre-exposure prophylaxis, HIV testing, linkage and retention in care, medication adherenceand viral load suppression. More visible global recognition of stigma in fuelling the HIV epidemic and in undermining the HIV response has been forthcoming through recent global declarations ]. This includes a recognition of the importance of addressing stigma in health facilities [72], yet concerted efforts to tackle stigma at any scale within countries are lacking, apart from Thailand. One reason for such inaction may be the small, but growing, body of evidence of effective intervention models for HIV-stigma reduction, particularly in health facilities. This paper adds to that literature. This evaluation of the Ghana "total facility" intervention found strong and consistent evidence of improvements, across indicator domains, apart from stigmatizing attitudes. Fear of acquiring HIV while providing care for clients living with HIV improved significantly in intervention versus comparison facilities, as did associated stigmatizing avoidance behaviours such as double gloving. These findings mirror results of interventions in Chinaand Vietnamthat also included addressing fear of HIV transmission in the health facility as a key component of stigma-reduction interventions. A further study in China underscores the importance of addressing fear directly and through a focus on standard precautions as a key strategy in reducing health facility stigma and discrimination . There was no statistically significant DID between the intervention and control facilities on the composite stigmatizing attitudes variable. This is not unexpected as changing deeply held attitudes may require a longer and more intense intervention than changing fear and behaviours based in incorrect knowledge or strengthening the facility environment to facilitate delivery of non-stigmatizing care. Additionally, more items in the composite attitude variables results in less change being expected because all of a respondent's stigmatizing attitudes would have to be reduced to alter a single "none-or-any" composite. That changing (and measuring change in) stigmatizing attitudes is potentially more complicated than for other immediately actionable drivers of stigma is also highlighted by a study in Vietnam, where the arm of the study that received a focus on social stigma (attitudes) had a similar drop in stigmatizing attitudes to the arm that received only a focus on fear-based stigma.
There are limitations to this study that should be noted. The short-time frame of the evaluation did not allow for assessment of longer-term intervention effects and therefore, could overestimate more transient effects of the intervention or underestimate those that require longer intervention periods. The nature of the setting, design and funding did not allow for randomization of the facilities to intervention or comparison. However, facility selection, the DID approach and IPTW should substantially account for differences between the intervention and comparison groups. Some stigma-reduction activities did occur in the comparison group of facilities, which likely diminishes some of the intervention effect. All facilities received their baseline results before the intervention began in the intervention sites, as required by ethical guidelines, which may have spurred some comparison facilities to begin addressing their baseline results. As presented in, respondents in the comparison arm did report participating in a range of training related to specific stigma drivers (e.g. Infection control and standard precautions) and other general, non-specified stigma-reduction activities. While the percent of staff reporting such exposure was significantly lower in comparison to intervention sites, the data does indicate some activities were happening in the comparison facilities. Social desirability bias is always present in stigmareduction studies. One way that the study worked to reduce this potential bias was by implementing data collection strategies to reassure health care staff respondents that their responses were anonymous, including self-filled questionnaires (where literacy was not a challenge); placing of questionnaires by respondents in envelopes which they sealed and that were put in a locked box; and not collecting any personal identifiers on the questionnaire. That the attitudes indicator, one that we would expect to be particularly sensitive to social desirability bias, was the one indicator that did not change significantly perhaps indicates a mitigation of social desirability in the study.
The rationale for addressing stigma for an effective HIV response and to achieve national and global targets is clearly underscored by the evidence of the negative link between stigma and testing, treatmentand viral load suppression. The importance and urgency of responding to stigma in a consistent and scaled manner is visible in recent global calls for stigma-reduction initiatives . Translating these global calls into action on the ground will require national HIV responses to include stigmareduction as a core feature, alongside prevention and treatment. Health facilities are a logical and feasible place to initiate this work, as demonstrated by this study and others. Ghana has begun building on this pilot intervention, expanding the intervention to other facilities with The Global Fund's support, while Thailand is in the midst of expanding approaches beyond a pilot study, with their stigma-reduction intervention currently implemented in over 100 health facilities. (Personal Communication, Dr. Taweesap Siraprapasiri, Thailand MOPH).
# | conclusion
Addressing HIV stigma is not only a public health imperative but a human rights one. Tackling HIV stigma in the health system is a logical place to develop and scaleup a national response to HIV stigma. This paper adds to the small but growing literature demonstrating that HIV stigma can effectively be addressed in health facilities and is a challenge welcomed by facility staff and management. This evidence provides a solid foundation for developing and testing the feasibility and efficacy of scaling up health facility stigma-reduction within countries. It will be critical that these efforts are accompanied by rigorous implementation science to ensure ongoing learning and adaptation to maximize effectiveness and long-term impact. Additional support for the analysis was provided by RTI, International. The content of this manuscript is the sole responsibility of the authors and does not necessarily reflect the views or policies of the Global Fund, the U.S. Agency for International Development or the U.S. President's Emergency Plan for AIDS Relief (PEPFAR) and does not imply endorsement by the U.S. Government. We are grateful to all the study respondents, health facility staff and clients, for sharing their time and insights and to management of intervention and non-intervention facilities, who supported the participation of staff and clients. We are especially grateful to the stigma-reduction facilitators and champion team members for their dedication to stigma-reduction. In-country implementation of both research and intervention were carried out by a team of talented and dedicated staff from the Educational Assessment Research Center's (EARC) who worked in partnership with the Ghana AIDS Commission (GAC) and the National AIDS Control Program (NACP), with support from colleagues at HP+. eliminate all forms of HIV related stigma and discrimination. 22nd Annual International AIDS Conference. Amsterdam; 2018 74. Li L, Liang LJ, Lin C, Wu Z. Addressing HIV stigma in protected medical settings. AIDS Care. 2015;27:1439-42.
## Supporting information
Additional information may be found under the Supporting Information tab for this article.
Data S1. Survey questionnaires. Data S2. Inverse probability of treatment weighting models. Data S3. Replication log for Ghana facility intervention Diff-in-Diff. 21 August 2019**.
|
Do beneficiaries’ views matter in healthcare purchasing decisions? Experiences from the Nigerian tax-funded health system and the formal sector social health insurance program of the National Health Insurance Scheme
Background: Purchasing is a health financing function that involves the transfer of pooled resources to providers on behalf of a covered population. Little attention has been paid to the extent to which the views of that population are reflected in purchasing decisions. This article explores how purchasers in two financing mechanisms: the Formal Sector Social Health Insurance Programme (FSSHIP) operating under the Nigerian National Health Insurance Scheme (NHIS), and the tax-funded health system perform their roles in light of their responsibilities to the populations. Methods: A case study approach was adopted in which each financing mechanism is a case. Sixteen (16) in-depth interviews with purchasers and eight (8) focus group discussions with beneficiaries were held. Agency and organizational behavioural theories were used to characterise the purchaser-citizen relationships. A deductive framework approach was used to assess whether actions identified in a model of 'ideal' strategic purchasing actions were undertaken in each case.Results: For both cases, mechanisms exist to reflect people's health needs in purchasing decisions, including quantitative and qualitative needs assessment, mechanisms to raise awareness of benefit entitlements and allow choice. However, purchasers do not use the mechanisms to effectively engage with and hold themselves accountable to the people. In the tax-funded system, weak information systems and unclear communication channels between the purchaser and citizens constrain assessment of needs; while timeliness of health information and poor engagement practices of Health Maintenance Organisations (HMOs) are the main constraints in FSSHIP. Inadequate information sharing in both mechanisms limits beneficiaries' awareness of entitlements. Although beneficiaries of FSSHIP can choose providers, lack of information on the quality of services offered by providers constrains rational decision-making and the inability to change HMOs reduces HMO responsiveness to beneficiary needs.
# Background
Purchasing is a key function of health care financing and is the process through which purchasers, on behalf of the population, transfer pooled resources to health care providers in exchange for health care services [bib_ref] A descriptive framework for country-level analysis of health care financing arrangements, Kutzin [/bib_ref]. That is, purchasers act as agents for people in the purchase of health care services. Accordingly, to fulfil this agency role effectively, it is important for purchasers to ensure there are effective mechanisms in place to determine and reflect people's needs, preferences and values in purchasing decisions, and to hold themselves accountable to the population for which they are responsible.
Many of the studies that examine health care purchasing have emphasised the relationship between purchasers and providers of services by examining contracts and provider payment mechanisms [bib_ref] Strategies to support quality-based purchasing: A review of the evidence, Dudley [/bib_ref] [bib_ref] Quality-based purchasing in health care, Waters [/bib_ref]. However, little attention has been given to the roles and responsibilities of purchasers in relation to the people they represent. This article focuses on the purchaser-citizen relationship for two health care financing mechanisms that operate in Enugu State, Nigeria: the tax-funded health system and the Formal Sector Social Health Insurance Programme (FSSHIP) that operates under the National Health Insurance scheme (NHIS). Specifically, the purpose of the article is to examine how the purchaser in each system performs their roles and responsibilities to the population they serve based on a model of 'ideal' strategic purchasing actions.
## The tax-funded system
Under the Enugu State's tax-funded system, the State Ministry of Health (SMoH) purchases and provides public health services for the entire state population. Residents of the state can elect to use these services. The SMoH defines a minimum package of health care services which covers promotive, preventive, and curative care at primary and secondary care levels, and includes services for communicable and non-communicable diseases, child survival, safe motherhood, nutrition, health education, laboratory services and community mobilisation. SMoH also engages a number of private providers to deliver selected services, including mortuary and immunisation services. Services that are not on the SMoH's package of services are mostly paid out-of-pocket with user fee required at the point of service utilization.
Health services are predominantly funded through general tax revenue, which the State Government allocates to SMoH as part of the State Budget. All government owned primary and secondary providers are involved in service delivery. Although primary health care falls under the purview of the Local Government Area (LGA), SMoH releases resources to Primary Health Centres (PHCs) for vertical programmes and pays salaries of state public servants and some senior level health workers at the LGAs. The LGA, in turn, pays salaries to all other PHC workers.
## The formal sector health insurance programme (fsship)
FSSHIP is a Social Health Insurance programme established under the Nigerian National Health Insurance Scheme (NHIS). FSSHIP is a mandatory scheme for employees in the formal sector. As of 2012, coverage was estimated to be about 3% of the population of Nigeria. [bib_ref] Scaling up National Health Insurance in Nigeria: learning from case studies of..., Dutta [/bib_ref]. Beneficiary contributions to the scheme are based on a proportion of earnings and are accompanied by contributions from employers (employers pay 3.25% of the employee's consolidated salary while employees pay 1.75%). The NHIS pools funds at the federal level through the National Health Insurance Fund (NHIF). NHIS contracts private, for-profit Health Maintenance Organisations (HMOs) to administer the purchasing system and channel resources to providers. Healthcare providers receive capitation payments for primary healthcare services and fee-for-service for secondary services. A mix of NHIS-accredited public and private health care providers are contracted to deliver services under the FSSHIP.
FSSHIP provides a standard benefit package that specifies primary, secondary and tertiary services, with some exclusions for high technology investigations (CT scans, MRI, etc.) and occupational diseases. FSSHIP members are allocated to specific HMOs, but can choose their primary health care providers from an NHIS accredited provider list.
# Methods
This study was undertaken as part of a multi-country study that examined purchasing functions in 10 countries in Africa and Asia . The multi-country study employed a case study approach in which individual purchasing mechanisms were the cases. The Nigerian study selected two mechanisms for examination: the tax-funded system in Enugu state and the Formal Sector Social Health Insurance Programme (FSSHIP). The citizen-purchaser relationship was the unit of analysis for this paper. The study examined whether there was a difference in the citizen-purchaser relationship between the tax-funded system, where both purchasers and providers belong to the same organisation, and the social health insurance system, where there is organisational separation between purchasers and providers. FSSHIP is financed at the federal level but operates at the state level with beneficiaries being federal government employees and other formal sector workers. FSSHIP and the tax-funded system are the main health purchasing mechanisms operating in Nigeria.
## Study setting
The study was conducted in Enugu state in 2014. Enugu, with a population of about 3.9 million in 2013, is one of 36 states in Nigeria and is located in the South-eastern zone with four other states. The South-east is one of six (6) constitutionally recognised geopolitical zones in Nigeria.
States are grouped into zones based on cultural, historical, and political relationships; the states in each zone have broadly similar patterns and levels of development. In Enugu, there areLGAs, 291 political wards, 471 communities and an urban-rural population ratio of 5:12.
## Data collection
Document review, individual in-depth interviews (IDIs) and focus group discussions (FGDs) were used to collect data for the study. For the tax-funded system, sixIDIs were held with respondents from the Ministry of Health and six (6) FGDs were conducted with community members from each of the three purposively selected communities (2 FGD per community). One community was selected from each of the three senatorial zones. For the FSSHIP, a total of ten (10) IDIs were held; six (6) with HMOs and four (4) with NHIS employees. Two (2) FGDs (one male and one female group) were held with beneficiaries of the scheme.
The interview respondents were purposively selected to include the key actor groups involved in healthcare purchasing in the two mechanisms. The conceptual framework described in the following section guided the mapping of the key actor groups. FGD participants in the FSSHIP were beneficiaries of the scheme selected from federal institutions operating in the state. In the tax system, FGD participants were male and female community members recruited from the three senatorial zones. At least one participant was a Facility Health Committee (FHC) member. The officer in charge of health facility that operated in the study community assisted in the recruitment of community members.
Initial mapping of actors based on information areas was used to determine the possible number of interviews. However, interviews were discontinued when no new information was obtained from the focus groups [bib_ref] Planning and recruiting the sample for focus groups and in-depth interviews, Macdougall [/bib_ref]. In addition, the tax funded system has wider coverage hence the need to conduct interviews in all three senatorial zones. The FGDs and IDIs were used to assess beneficiary views on key strategic purchasing actions in the relationship between citizens and purchasers.
Both IDIs and FGDs were conducted by members of the research team; a moderator and a note taker attended each interview. Note takers ensured that the key points for each question were covered and that follow-up questions and prompts were explored. FGDs were primarily run in the native language [Igbo] while IDIs were held in English with respondents at liberty to communicate in either of the two languages. Individual interviews and FGDs were audio recorded with the consent of participants. All interviews were transcribed and those in Igbo were translated into English. Transcribers were well versed in both the Igbo and English languages. The moderators and note takers checked the accuracy of the transcripts against the audio recordings.
A review of grey literature including the 2010-2015 Strategic Health Development Plan, and 2014 Health Policy documentation from the Ministry of Health, the Ministry of Finance and the Planning Department. Government records, including the 2013 and 2014 expenditure reports, budget performance reports, health financing policy documents, health sector reports and presentations at meetings, were examined to determine the context for and expected function of purchasing arrangements as described in the policy design.
## Analytical framework
The multi-country study used a conceptual framework that identified principal-agent relationships between: purchasers and providers; purchasers and citizens; and purchasers and the government. Under agency theory, agents have specific roles to play for their relevant principals [bib_ref] The economics of agency, Arrow [/bib_ref]. The multiple relationship purchasing framework was used to identify a list of ideal strategic purchasing actions which was compared with policy design and actual practice in order to identify key design and implementation gaps.
This article focuses on the relationship between citizens and purchasers and considers the actions that purchasing organisations, as agents, are expected to undertake in order to: be responsive to the needs and preferences of the population they cover (principals); accountable to the people; and ensure people access their entitlements. A broad conceptual framework for strategic purchasing was developed to identify key actions that purchasers, in their role as people's agent, are expected to undertake , including:
1. Assessment of the service needs, preferences and values of the population and use of this information to specify service entitlements/benefits; 2. Informing the population of their entitlements and obligations; 3. Ensuring that the population can access their entitlements; 4. Receiving and responding to complaints and feedback from the population; and 5. Allowing people to choose/exit a purchaser and/or provider.
Organisational behavioural theory suggests that, of the above-listed actions, 1, 4, and 5 can be viewed as either 'voice' or 'exit' mechanisms to influence purchaser decision-making. While 'voice' mechanisms, such as public consultation, advocacy groups, complaint mechanisms, and formal representation of people on purchasing boards, − allow people to express opinions; 'exit' strategies allow people to choose purchasers and/or providers and express dissatisfaction by leaving a specific purchasing mechanism or changing an assigned provider.
Determination of how a particular purchaser-citizen relationship is functioning involves the examination of voice and exit mechanisms, and mechanisms that hold purchasers accountable to people (action 2 above) and ensure that people's entitlements are fulfilled (action 3 above).
# Data analysis
Interview transcripts were the primary data analysed. QSR NVivo 10 was used to organize and manage data. A deductive framework approach was used to analyse the data. The study used the list of actions identified above as the 'theoretical ideal' and determined whether those actions were reflected in the policy design in each case (i.e. the tax-funded system and FSSHIP) and how the mechanisms functioned in actual practice in order to identify gaps in both the policy design and implementation.
# Results
Using the analytical framework presented in the Methods section, the findings for each purchasing mechanism are presented according to five themes: (1) assessment of the health needs of beneficiaries; (2) ensuring beneficiary awareness of entitlements and obligations; (3) receiving and addressing complaints; (4) ensuring access to entitlements; and (5) strategies for promoting choice and right to exit.
Assessment of health needs and update of entitlements to reflect needs, preferences, and values of the population In both mechanisms, two main needs assessment strategies were identified: (1) processes that aim to collect routine epidemiological data on service utilization; and (2) formal engagement with beneficiaries to elicit their needs and preferences for the services they receive.
## Routine health data collection
In the tax-funded system, the MoH collects routine health service data through the Health Management Information System (HMIS). Reports and statistics on services utilized at health facilities are intended to inform health planning and priority setting. However, MoH respondents indicated that routine data are weak and unreliable for determining needs and making meaningful decisions and plans. Data quality and subsequent usage are compromised due to problems associated with the capacity of human resources, the timeliness of collection and completeness of data reporting:
"The information is not coming when it is due. In many of the [health] centres, some of the information officers are not functioning as required; to get statistical information is a problem, even if you get the information, what about the quality? You can't be sure of the data you are getting. So, we have been working hard to improve on all those things so that at least it will help us in planning." (IDI 6, MoH respondent)
In the FSSHIP, HMOs are expected to compile and transmit patient encounter data to NHIS on a quarterly basis. The data is gathered at the health facility level and is intended to highlight disease burden and utilization rates. According to a respondent from an HMO: "There is an encounter data form, which is an instrument we use to collect data on utilization and service delivery. The encounter data is used in various ways. The major thing is for you to ascertain the level of utilization. For instance, do you have up to 13% of enrolees coming to a hospital; is it 3% or 50%? It shows what percentage of the enrolees you have that are coming. It also shows the disease burden in that particular region." (IDI 1, HMO respondent) There was consensus that the required information is not transmitted as designed because many providers do not compile and make these data available. This was partly attributed to the broad engagement of providers with (multiple) HMOs, and larger health facilities, in particular, are unlikely to deliver monthly patient documentation:
"Some [providers] are complying but many don't. The problem is that majority of them are from very big hospitals where they have a lot of activities with other HMOs and a lot of things to do [...] But whenever they come, we compile them and send." (IDI 4, HMO respondent)
## Engagement mechanisms
Facility Health Committees (FHCs) serve as a means to engage citizens in the tax-funded health system. The FHC is a community-based arrangement that is established to offer a sustainable channel for promoting citizen voice and to encourage greater accountability and responsiveness in the health system. Committee members are selected through community-based democratic processes (voting at an open community meeting); and based on their trustworthiness and stated willingness to work for their community. Committee members are primarily responsible for: regular interaction with health workers at their designated facilities; regular interaction with members of their local community to understand health needs; providing feedback and liaising with the MoH on their community's health. These committees have been established in most primary and secondary health facilities in the state. Most FHCs carry out their responsibilities in consulting both health workers and community members to identify needs, but health services are often not improved and identified needs remain unmet:
"They ask questions to know the state of things; after asking those questions, at the end of the day, we won't see any improvements." (FGD 1, beneficiary of taxfunded system) Problems relating to unclear communication channels between the FHCs and MoH and the bureaucratic protocols involved in meeting policy makers limit FHC performance:
"What we are talking about is, the protocols that are involved in seeing them [policy makers] are the issue." (FGD 6, beneficiary of tax-funded system).
"What we want is to have an effective avenue for communication. Now that we have laid our complaints, how do we get them to the Ministry of Health or those in authority so that the problems can be solved? That is our problem." (FGD 3, beneficiary of tax-funded system).
Although, by design, volunteerism and stewardship to community are emphasized values for FHC committee members, some provisions have been made for minor, non-monetary incentives, such as recognition in the community and, in the past, some non-governmental agencies provided monetary incentives to FHC members. Currently, the absence of formal incentives further constrains active participation by some members of the FHC committee.
"You know that it's not easy to pull out a family man to spend hours on free work. If you call him 2-3 times, the next time he'll tell you that he is on the farm.... So, if the ministry gave him stipend at the end of the month, it would be of great help." (FGD 2, beneficiary of tax-funded system) Under the FSSHIP, HMOs are mandated to conduct quarterly seminars/interactive sessions to educate beneficiaries on the scope of benefit entitlements, rights, and privileges. In addition, these forums provide the opportunity to assess beneficiaries' satisfaction with services delivered, and receive complaints, suggestions, and recommendations. Most HMOs report conducting these seminars and engaging with beneficiaries to understand their needs, but beneficiaries expressed mixed responses, with the majority suggesting inadequate and limited engagement with HMOs: It was also acknowledged that the nature of the engagement does not allow for the views of all to be represented as the forums involve few people from selected institutions:
[formula] " [/formula]
"Like I have just said, there are 64 in the office, and I was the only person that came for that meeting. I believe that it would have been okay if others were given opportunity to air their views." (FGD 2, FSSHIP beneficiary)
In general, beneficiaries assert that their preferences for service delivery are not met. For example, problems with long waiting times, quality of services, drugs and overall poor treatment of scheme patients when compared to feepaying patients have been repeatedly expressed but not yet addressed. However, one HMO noted that input from beneficiary engagement was the basis for the revision of the initial benefit package to include some of the services previously on the exclusion list:
"The comments we get, we try to feedback to NHIS that these are what people are saying. For instance, the issue of Myomectomy -at the commencement, it wasn't there, that is a fibroid operation. That has been changed. The issue of maternity care, before it was limited to a particular group of people but now it's open to all women to have four live births irrespective of the number of births they had before they joined the scheme." (IDI 6, HMO respondent).
## Mechanisms through which purchasers ensure beneficiary awareness of entitlements and obligations
In the tax-funded system, Enugu State health law recognises the right to health as Government's obligation to the citizens. The State Patient Rights Charter, an elaboration of the health law, specifies the rights, entitlements, and obligations of patients as service users. Posters indicating the contents of the charter are expected to be displayed in strategic locations at all public health facilities. The charter is intended to stimulate a change in citizen behaviour by provision of adequate knowledge of legal rights. Although posters are displayed in most public health facilities, the information provided has had little impact; patients remain largely unaware of their rights and do not demand that they be upheld: "Our awareness of rights is poor. We have what is called Patients' Rights Charter. We paste it on the wall. But I am almost 2 years here now and I have not gotten a patient who said, "doctor, see this person refuses to explain the cost of this drug to me"; "It is in the charter"; or "This person is supposed to be on duty or is supposed to be attending but he comes to work at 10am." That's why I said our rights awareness is still very poor." (IDI 4, public provider respondent)
One policy maker observed arrangements were being made to inform the public about the services available at hospitals:
"In our district hospitals, we now have community mobilization officers working with us. So they will mobilize communities to take-up the health services offered in their community." (IDI 4, MoH respondent) For the FSSHIP, the rights of beneficiaries are stipulated in the scheme's operational guidelines, including the specification that patients can exercise their rights. Through continuous education, HMOs are required to ensure that new and existing beneficiaries are aware of the procedures for choosing and changing primary healthcare providers, the criteria for assessment of care, and the scope of benefit coverage. These activities are reportedly being carried out by HMOs:
"We intermittently organize forums. The HMOs are able to use that opportunity to tell their enrolees their rights and privileges, which is their benefit package." (IDI 1, NHIS respondent) However, beneficiary respondents indicated there was limited awareness of the scope of benefits and entitlements among the majority of beneficiaries. The same group suggested that using the media would achieve wider coverage rather than the current, infrequent forums. Limited awareness of rights has often led to providers making unnecessary financial demands from beneficiaries: "Many people don't even know their rights. You get to the hospital and you are expected to pay just 10% for the drugs. Sometimes, if you get to the hospital and you are not aware of that, the people [hospital] will just extort you; take money from you that they are not supposed to. We want a situation where we know…if government has a registry of the services that we are entitled to receive, we want to have copies so that everybody knows how much is required from us [copayment]. We want to know. And on the issue of drugs, giving us cheap drugs and all that, we also want it addressed." (FGD 2, FSSHIP beneficiary).
## Mechanisms to receive and address complaints
Complaint and redress policy in the tax-funded system stipulates complaint lodgement procedures and appropriate actions in the event that grievances are not satisfactorily resolved. The policy further specifies that providers should pay utmost respect to patients' opinions and rights within the health system: "We are trying to inculcate some level of discipline and decorum in the facilities' staff, so that they handle the patients with care." (IDI 9, MoH respondent)
In terms of seeking resolution, many people are not aware of their rights and are reluctant to complain due to a lack of awareness of processes and little confidence to do so:
"Well, the citizens are not properly informed. In short, they don't know their rights most of the time. Someone may go to the hospital and he is not properly treated but he finds it difficult to complain. It's like a cultural issue" (IDI 4, MoH respondent).
A second redress mechanism that is stipulated in policy is the placing of opinion/complaints boxes at all public health facilities. In practice, many facilities do not have complaint boxes and, where they do exist, they are not considered useful as they are often poorly secured meaning complains do not reach health authorities:
"If you put your concerns into those boxes, someone may even take it before it gets to the authority.
[…] It has no key. The one here is like box. It is broken and someone else can have access to anything you put there." (FGD 4, beneficiaries of tax funded system) Similarly, the HMOs are required to install complaint boxes at provider facilities, periodically conduct satisfaction surveys, and establish 24-h help lines to receive and address beneficiary complaints. All HMOs attest to having functional help lines and responses from respondents also affirm that mechanisms do exist for them to express discontent either by phone, face-to-face during meetings or through scheduled visits from officers (staff of agencies/ministries that have been appointed by HMOs to serve as intermediaries between the agency and their HMO). Despite these mechanisms, beneficiaries assert that complaints, especially those relating to provider behaviour and overall service delivery, are seldom addressed and feedback rarely provided by HMOs:
"They [HMOs] encourage us to complain about bad treatment and things like that, but after that what happens? Nothing, because we have made series of complaints, but the thing stops there because when you go back to your provider, they still mete out the same treatment." (FGD 2, FSSHIP beneficiary)
The main reason reported for poor response to complaints is the limited competition for beneficiaries among HMOs and over-centralization of decisionmaking by NHIS -most decisions are taken at the national level and take time to be enacted at the community level.
## Mechanisms for purchasers to ensure the population can access their entitlements
In the tax-funded system, primary and secondary public providers are set up and managed by the government and members of the public can directly access care from these providers according to need and ability to pay stipulated fees. Assessment of secondary services is based on referral. The MoH tries to regulate user fees and drug charges so that accessing health services remains affordable to patients. The provision of free MCH services for pregnant women, and children under-5 is a further measure to ensure access for vulnerable groups. A district-based health system (DHS) was implemented to decentralize health services and resources to ensure access to services by directly linking primary and secondary care in each locality. The DHS was expected to meet the infrastructural and human resource redistribution needs by establishing, at a minimum, a primary health facility (PHC) in every ward, while also strengthening referral lines:
"There is no part of this state that is excluded from this district health system… Some communities have health facilities within them but the ones that don't were linked to a facility that is closest to them." (IDI 2, MoH respondent)
The implementation of the DHS has improved construction of health facilities, especially primary care facilities, in many previously under-served areas. However, there have not been corresponding increases in human resources, health services and the availability of drugs and medical equipment:
"We are handicapped in so many areas, not just the personnel, but also in the supply of drugs and the other equipment, so the services are not up to the standards that are expected." (IDI 1, MoH respondent)
To access health services, HMOs register beneficiaries and issue identification cards. Some beneficiaries reported protracted delays in obtaining registration cards from NHIS. Accessing referral services requires pretreatment authorization from the relevant HMO within 24 h of the request from the provider facility. Beneficiaries expressed concerns about recurring delays in authorization of referrals by HMOs, which often prevented patients from receiving referral services:
"There was a day I was there [at hospital] till evening and I didn't get the go ahead from the HMO, and they [provider] asked me to go [home]. Assuming I had an ailment that could kill me, then I'll not get help...I don't understand the reason." (FGD 3, FSSHIP beneficiary)
Other barriers to accessing necessary care include the limited availability of NHIS approved drugs, leading to frequent out-of-pocket payment for drugs; and poor treatment of scheme patients compared to fee-paying patients.
"…NHIS patients are like the riffraff, poor people, the common masses, the nobodies. And so any time you come in there under the platform of NHIS, they look down on you, you don't get the attention you require." (FGD 2, FSSHIP beneficiary).
## Existence of choice and exit strategies
While beneficiaries of the tax-funded system can decide whether to seek publicly-provided care, the mandatory nature of the FSSHIP limits the scope of beneficiaries to exit the scheme if they are federal civil servants. Furthermore, staffs from federal ministries, departments, and agencies (MDAs) are assigned to certain HMOs and beneficiaries are not allowed to switch HMOs, even if they are dissatisfied with that HMO. Respondents were of the view that absence of choice of HMOs has led to a nonchalant attitude towards performing mandated tasks for beneficiaries by HMOs:
"If we had the chance to change HMOs, they would try to find out what we need. So am suggesting that [allowing beneficiaries to change HMOs] so that the establishment can move en masse." (FGD 1, FSSHIP beneficiary)
Beneficiaries are free to choose their own primary provider and change providers after an initial six (6) month period. However, beneficiaries indicated that no criteria have been defined by NHIS to help them to gauge the performance of healthcare providers or of service quality. Beneficiaries' opinions of providers are largely based on discussions with other beneficiaries, and such judgements are sometimes made using incorrect or unsubstantiated information:
"I just changed from Hospital A to Hospital B; but I just noticed that hospital A was even better than the place I'm at now. So, I'm just on the high sea, I do not know what to do again." (FGD 2, FSSHIP beneficiary).
# Discussion
This study examined the mechanisms used to reflect the needs and preferences of beneficiaries in two healthcare financing mechanisms: the Nigerian taxfunded system and the FSSHIP. It also compared how these financing arrangements function in practice in light of an "ideal" strategic purchasing model. The study found that the policy includes various mechanisms to allow purchasers to undertake strategic purchasing actions. These mechanisms include the use of specific strategies to assess the health needs of beneficiaries, raise awareness of beneficiaries' rights and entitlements, address beneficiary concerns, and ensure access to health services. However, the strategic purchasing mechanisms operating under the taxfunded system are not effectively implemented and/or managed, and variations in how the strategic purchasing mechanisms are implemented were observed in the FSSHIP. The root causes of implementation gaps vary for the two healthcare financing systems.
In both mechanisms, deliberate design of institutional arrangements is required for mechanisms that allow people to make their opinions known (voice mechanisms) to be effective. The design should include clear channels of communication between health system users and decision-makers, with user-responsive management arrangements also implemented [bib_ref] Exit, voice, governance and userresponsiveness: the case of English primary care trusts, Pickard [/bib_ref]. In the tax-funded system, voice mechanisms relating to needs assessment, member/community engagement devices and complaint systems that enable people's needs and preferences to be reflected in purchasing decisions all operate independently and are not specifically integrated into the design and review of benefit entitlements. Consequently, benefits do not necessarily reflect the expressed needs and preferences of those who they are designed to help. Effectively engaging and reflecting the needs of beneficiaries in purchasing decisions remains a weakness of the FSSHIP. Voice mechanisms for engaging with beneficiaries and addressing complaints are supposed to be implemented by HMOs and the NHIS is supposed to oversee the implementation of these mechanisms. However, individual HMOs have not implemented these mechanisms according to the FSSHIP guidelines. Consequently, beneficiary engagement mechanisms are rarely robust or sufficiently inclusive to achieve intended outcomes.
A major cause of the implementation gaps in voice mechanisms under the tax system is weak information systems for the generation and transmission of quality, accurate and timely information for use in planning and decision-making. Ineffective community engagement processes and unclear lines of communication in the processes have widened the communication gap between purchasers and citizens and reduced the ability of voice mechanisms to achieve their intended purpose. Under the FSSHIP, implementation gaps are mainly associated with poor beneficiary engagement processes. The poor processes are due to a lack of coordination and inadequate supervision of the HMO practices by NHIS, which has resulted in HMOs neglecting their roles and responsibilities to FSSHIP members.
Raising community and beneficiary awareness of entitlements appears to be undervalued in the tax system. Voice mechanisms function on the existence of informed users with some level of power to influence, however the tax system does not include this group. The patient charter has received limited attention from MoH and has not been implemented in such a way that it achieves its intended purpose; beneficiaries have limited awareness of rights and lack of power to influence purchasing decisions. Under the FSSHIP, systems for increasing member awareness of benefit entitlements are specified in the policy design. The contract between HMOs and NHIS specifies that HMOs must implement such systems however, implementation of these tools has not occurred as intended because poor engagement mechanisms also weaken information exchange resulting in low levels of awareness of entitlements by program beneficiaries [bib_ref] Knowledge and attitude of civil servants in Osun state, southwestern Nigeria towards..., Olugbenga-Bello [/bib_ref].
Community accountability mechanisms are intended as useful tools for enhancing purchaser accountability and responsiveness while also empowering citizens. The tax-funded system uses the FHCs as an accountability mechanism to communities. In practice, the blurred lines of communication and feedback between the FHCs and the MoH weaken mutual accountability. The literature indicates that properly functioning committees exist in Nigeria and elsewhere to manage facility resources, identify population needs, improve community participation, address complaints and manage overall quality management [bib_ref] Health facility committees and facility management-exploring the nature and depth of their..., Goodman [/bib_ref]. The results from this study corroborate the literature but also indicate the main constraints to FHC performance are issues associated with incentives, clarity of roles, and proper communication channels. While a contractual agreement exists between the HMOs, healthcare providers and NHIS in the FSSHIP, the roles of citizens appear to be downplayed due to the absence of explicit contracts and mechanisms through which beneficiaries can directly hold HMOs accountable.
Under the tax-funded system, the implementation of DHS contributed to the proliferation of healthcare facilities in all geographic localities, but shortages and geographic imbalances in the distribution of human resources in the public sector constitute a significant barrier to access. Consequently, many people are still unable to access necessary healthcare services, especially in under-served areas. Under the FSSHIP, the selfserving behaviour of some HMOs can and do constrain access to care, particularly for secondary services which require authorization by HMOs prior to use. The often deliberate efforts by some HMOs to control the volume of service utilization can have significant negative impacts on beneficiaries when they are denied their entitled healthcare or made to contribute out-of-pocket payments [bib_ref] National health insurance scheme: how protected are households in Oyo state, Nigeria..., Ilesanmi [/bib_ref].
Purchasing mechanisms can include strategies that allow beneficiaries the right of exit and choice. When faced with a loss of customers, increased competition among providers and purchasers can bring improvements in service quality [bib_ref] Exit, voice, governance and userresponsiveness: the case of English primary care trusts, Pickard [/bib_ref]. While there are no explicit exit mechanisms for beneficiaries in the tax-funded system, the health system generally allows exit when users decide to use non-public sector facilities and services. Although the FSSHIP allows choice of primary care providers, a lack of information on the quality of healthcare from different providers limits the ability of members to make rational decisions on the choice of healthcare providers. Conversely, the absence of choice of HMOs for FSSHIP members has created little incentive for competition among HMOs and has weakened the ability of FSSHIP members to compel purchasers to be responsive to their health needs.
For purchasing to support progress towards the achievement of health system and universal coverage goals, purchasers must recognise the need to actively engage the citizens in determining the right combination of services and align resource allocation to health needs. Improved mutual accountability between purchasers and populations they cover is necessary for increased citizen voice and purchaser responsiveness. While the introduction of strategic purchasing measures remains a challenge, some experiences can be drawn from other country practices. Thailand's Universal Coverage Scheme (UCS) provides some lessons on how strategic purchasing can evolve over time given the right prerequisites and enabling environment [bib_ref] Achieving universal health coverage goals in Thailand: the vital role of strategic..., Tangcharoensathien [/bib_ref].
A limitation of this study is that the two financing mechanisms examined target different groups of the population which may affect the comparison of the relationship between purchasers and beneficiaries between mechanisms. The findings should be interpreted in this context. The two financing mechanisms were selected because they represent the two largest healthcare financing arrangements operating in the state, providing the opportunity for governments to learn important lessons for health system reform and universal health coverage.
# Conclusions
Both the tax-funded system and the FSSHIP include mechanisms that allow purchasers to reflect people's needs, choices, and preferences in purchasing decisions. The mechanisms include quantitative and qualitative assessment of health needs, mechanisms to raise awareness on benefit entitlements and mechanisms that allow choice. However, purchasers are not using the mechanisms effectively to engage and hold themselves accountable to people they represent. The MoH, HMOs and NHIS appear to give little value to their responsibilities to the people, which is the root cause of the implementation gaps in the voice mechanisms described in policy. Under the FSSHIP, the NHIS's lack of stewardship over the HMOs, who undertake purchasing administration, has resulted in variations in how HMOs have implemented voice and accountability mechanisms. In addition, under FSSHIP, the inability of members to choose HMOs does not encourage HMOs to be responsive to the needs and preferences of members.
Public purchasers should re-evaluate and strengthen their awareness of their roles and responsibilities to the people they represent, and be more responsive and accountable to the people. Civil society organisations (CSOs) have a role to play in empowering people to make their opinions known and facilitate purchaser engagement with the people. Furthermore, under the tax-funded system, as the current definition of services provided in the public sector is based on monitoring information from health facilities, information systems must be strengthened and the accuracy and timeliness of information improved.
The NHIS should be mindful of the consequences of being unable to meet FSSHIP members' needs and preferencesthe scheme currently suffers from low uptake rates and does not receive sufficient resources from member contributions to cover operating costs. Alternative strategies for improving beneficiaries' understanding of entitlements are necessary to improve participation. The NHIS should also provide stronger stewardship to HMOs and encourage HMOs to engage with and become accountable to their members. Furthermore, the NHIS should introduce mechanisms that allow FSSHIP members to choose HMOs, facilitating HMO responsiveness to members.
|
Hypoxia Mimetic Agents for Ischemic Stroke
Every year stroke claims more than 6 million lives worldwide. The majority of them are ischemic stroke. Small molecule-based therapeutics for ischemic stroke has attracted a lot of attention, but none has been shown to be clinically useful so far. Hypoxiainducible factor-1 (HIF-1) plays a crucial role in the transcriptional adaptation of cells to hypoxia. Small molecule-based hypoxia-mimetic agents either stabilize HIF-1α via HIFprolyl hydroxylases (PHDs) inhibition or through other mechanisms. In both the cases, these agents have been shown to confer ischemic neuroprotection in vitro and in vivo. The agents which act via PHD inhibition are mainly classified into iron chelators, iron competitors, and 2 oxoglutarate (2OG) analogs. This review discusses HIF structure and key players in the HIF-1 degradation pathway as well as the genes, proteins and chemical molecules that are connected to HIF-1 and how they affect cell survival following ischemic injury. Furthermore, this review gives a summary of studies that used PHD inhibitors and other HIF-1α stabilizers as hypoxia-mimetic agents for the treatment of ischemic injury.
# Introduction
HIF-1 is a transcription factor that acts as a master regulator of O 2 homeostasis [bib_ref] Hypoxia-inducible factor 1: master regulator of O2 homeostasis, Semenza [/bib_ref]. It modulates the expression of an array of genes , and many of them are involved in the cellular adaptation to hypoxia [bib_ref] Hypoxia-inducible factor 1: oxygen homeostasis and disease pathophysiology, Semenza [/bib_ref]. Vascular endothelial growth factor (VEGF), Erythropoietin (EPO), glucose transporter 1 (GLUT1), heme oxygenase 1 (HO-1), and endothelial nitric oxide (eNOS) synthase are some of the genes that are directly involved in HIF-1 mediated cellular response to hypoxia [bib_ref] Oxygen sensing by HIF hydroxylases, Schofield [/bib_ref]. As HIF-1 stabilization is an endogenous cellular mechanism to reduce hypoxia-induced injury, it is hypothesized that pre and post-hypoxic stabilization of HIF-1 by external means can enhance protection [bib_ref] Hypoxia inducible factor 1 as a therapeutic target in ischemic stroke, Shi [/bib_ref]. Among these foreign interventions, the small molecule-based HIF-1 stabilization is widely studied and has shown promising results [bib_ref] Hypoxia-inducible factor prolyl 4-hydroxylase inhibition, Siddiq [/bib_ref]. Most of the small-molecules used for this purpose either mimic or compete with the co-factors or co-substrates of HIF-1 degradation pathway. The small molecules that are used to stabilize HIF-1α are widely known as hypoxia-mimetic agents. Some of the familiar hypoxia-mimetic agents are desferrioxamine (DFO) and cobalt chloride (CoCl 2 ). Apart from these small molecules, there are newly found mRNAs and other compounds which stabilize HIF-1α. In this review, we discuss the small chemical molecules and other biological compounds that have conferred protection against ischemic stroke injury via HIF-1 pathway stabilization.
## Hif structure and pathway
Hypoxia-inducible factor belongs to the PAS family (PER-ARNT-SIM), and functionally active HIF consists of two basic helixloop-helix (bHLH) protein subunits, alpha and beta/ARNT (aryl hydrocarbon receptor nuclear translocator). Apart from bHLH, N-terminal half of HIFs composed of Per-ARNT-Sim (PAS) homology domains. The bHLH and PAS regions are involved in the heterodimerization and DNA binding, respectively. The terminal-transactivation domains (TADs) are present in the C-terminal region of HIF, and they control the transcriptional activity of the protein [fig_ref] FIGURE 2 |: Domain structure of human HIF subunits and isoforms [/fig_ref] [bib_ref] Regulation of myogenesis by environmental hypoxia, Beaudry [/bib_ref].
The alpha subunit of HIF has three isoforms, HIF-1α, HIF-2α, and HIF-3α. Likewise, the beta subunit has three isoforms, ARNT, ARNT2, and ARNT3 [bib_ref] HIF-1: mediator of physiological and pathophysiological responses to hypoxia, Semenza [/bib_ref]. The HIF-1α/β dimer controls the transcription of genes by binding to a core DNA motif (G/ACGTG) in hypoxia-response elements (HREs). The HIF-1β subunit is constitutively produced in the nucleus whereas the stability of HIF-1α depends on the availability of oxygen [bib_ref] Hypoxia-Inducible Factor-1 (HIF-1): a potential target for intervention in ocular neovascular diseases, Vadlapatla [/bib_ref]. Under hypoxic condition, HIF-1α is stable and transported to the nucleus. However, under the normoxic condition, HIF mediated transcription is inhibited by the HIF prolyl hydroxylases (PHDs) and asparaginyl hydroxylase (FIH). PHD2 hydroxylates the two proline residues (402 and 564) on the oxygen-dependent degradation domain (ODDD) of HIF-1α and makes it susceptible to ubiquitin-mediated proteolysis via von Hippel-Lindau tumour suppressor (pVHL) binding. On the contrary, FIH hydroxylases asparagine residue (803) of HIF-1α and prevents binding of the transcriptional coactivators (p300/CBP) to HIF-1α [fig_ref] FIGURE 3 |: Regulation of HIF-1 during hypoxia and normoxia [/fig_ref] [bib_ref] Hypoxia-inducible factor prolyl hydroxylases as targets for neuroprotection by "antioxidant" metal chelators:..., Speer [/bib_ref]. Hydroxylation of HIF α subunit by PHDs is an evolutionarily conserved mechanism of sensing the oxygen concentration inside the cell. Thus, stabilization of HIF-1α by an exogenous technique may activate transcriptional and posttranscriptional responses that will help the cell to overcome hypoxic damage [bib_ref] HIF-1α pathway: role, regulation and intervention for cancer therapy, Masoud [/bib_ref].
HIF-1α degradation can be restrained by inhibiting the activity of PHDs and FIH. PHDs and FIH belong to a class of enzymes known as 2-oxoglutarate dependent dioxygenases and are the most prominent known family of non-heme oxidizing enzymes [bib_ref] Structural and mechanistic studies on 2-oxoglutarate-dependent oxygenases and related enzymes, Schofield [/bib_ref]. There are three PHD isoforms, PHD1, PHD2 and PHD3. Even though these isoforms have a similar C-terminal catalytic domain, N-terminal sequences differ significantly [bib_ref] Selective inhibition of Hypoxia-Inducible Factor (HIF) prolylhydroxylase 1 mediates neuroprotection against normoxic..., Siddiq [/bib_ref]. mRNA expression analysis revealed that the expression patterns of PHD isoforms are tissuespecific and all the isoforms play a unique role in the regulation of HIF-1α and HIF-2α [bib_ref] Differential function of the Prolyl Hydroxylases PHD1, PHD2, and PHD3 in the..., Appelhoff [/bib_ref]. Both PHDs and FIH require oxygen, iron (Fe 2+ ) and 2OG for the hydroxylation reaction [bib_ref] Hypoxia inducible factor-1: its potential role in cerebral ischemia, Singh [/bib_ref].
Hypoxic microenvironments are important in the developing embryo, and they control cellular differentiation. These hypoxic microenvironments prompt the cells to activate HIFs which in turn regulate the development of the blood, vasculature, placenta, nervous system, and other organs [bib_ref] The role of oxygen availability in embryonic development and stem cell function, Simon [/bib_ref]. Thus, HIFs have a crucial role in embryonic development and survival. Validating the role of HIF during embryonic development, studies using HIF-1α deficient mouse embryos stopped developing by day 9 and died by day 10.5 due to extensive cell death and other abnormalities [bib_ref] Hypoxia-inducible factor 1 and cardiovascular disease, Semenza [/bib_ref].
## The relevance of the hif-1 pathway in ischemic stroke
Cerebral stroke is a significant cause of morbidity and mortality worldwide [bib_ref] Temporal and geographic trends in the global stroke epidemic, Kim [/bib_ref]. More than 80% of strokes are ischemic while the remaining is hemorrhagic. Ischemic stroke occurs when a vessel supplying blood to the brain is obstructed due to a thrombus, embolus, or other blockage [bib_ref] Prolyl hydroxylase domain inhibitors: a route to HIF activation and neuroprotection, Harten [/bib_ref]. Prevention of neuronal death and improving the recovery following ischemic injury are the central focuses of developing stroke therapeutics. Apart from tissue plasminogen activator (tPA), no drugs are able to confer neuroprotection following ischemic stroke in clinical studies [bib_ref] Hypoxia inducible factor 1 as a therapeutic target in ischemic stroke, Shi [/bib_ref].
A large number of preclinical studies that used PHD inhibitors were able to show that the activation of HIF-1 can be a potentially attractive way to achieve significant neuroprotection following ischemic injury [bib_ref] Hypoxia-inducible factor prolyl hydroxylase inhibition: robust new target or another big bust..., Karuppagounder [/bib_ref]. Apart from stroke, PHD inhibition persistently extended neuroprotection in diverse neurological disease models [bib_ref] Hypoxia-inducible factor prolyl hydroxylases as targets for neuroprotection by "antioxidant" metal chelators:..., Speer [/bib_ref]. At the same time, a few studies observed adverse effects on brain cells upon treatment with PHD inhibitors mostly because of the activation of pro-death proteins like BNIP3 and NIX [bib_ref] Prodeath or prosurvival: two facets of hypoxia inducible factor-1 in perinatal brain..., Chen [/bib_ref]. As hypoxia induces HIF-1α expression, most hypoxia-mimetic agents upregulate HIF-1α and orchestrate its downstream gene expression [bib_ref] Hypoxia-inducible factor prolyl hydroxylase inhibition: robust new target or another big bust..., Karuppagounder [/bib_ref]. Neuroprotection via HIF-1α stabilization following ischemic stroke is still a controversial topic due to its link to a vast variety of genes that mediate both adaptive and pathological processes [bib_ref] The good, the bad, and the cell type-specific roles of hypoxia inducible..., Vangeison [/bib_ref]. Unlike PHD inhibitors, not many FIH inhibitors are discovered, and the possibility of attaining neuroprotection via FIH inhibition has yet to be fully explored.
The altered expression of HIF-1 and its downstream genes during hypoxia have been very well studied in both in vitro and in vivo models [bib_ref] Hypoxia-inducible factor 1: oxygen homeostasis and disease pathophysiology, Semenza [/bib_ref]. Several processes like angiogenesis and erythropoiesis are connected to HIF-1 stabilization. Apart from this, recent studies were able to deduce more novel phenomena and genes linked to the HIF-1 pathway [bib_ref] NCX1 Is a novel target gene for hypoxia-inducible factor-1 in ischemic brain..., Valsecchi [/bib_ref] [bib_ref] Targeted genes and interacting proteins of hypoxia inducible factor-1, Liu [/bib_ref]. In a transient global ischemia model, HIF-1 expression was enhanced as early as 1 h in the recovery phase, and this enhanced expression lasted for 7 days [bib_ref] Activation of hypoxia-inducible factor-1 in the rat cerebral cortex after transient global..., Chavez [/bib_ref]. Besides, expression of HIF-1 regulated genes, EPO and GLUT1 were elevated during the reperfusion period. The above study also found that insulin-like growth factor-1 (IGF-1) was able to induce HIF-1 expression in vitro and in vivo.
HIF-2α has 48% structural similarity with HIF-1α, and it also activates HRE-dependent gene transcription [bib_ref] Differential roles of Hypoxia-Inducible Factor 1 (HIF-1 ) and HIF-2 in hypoxic..., Hu [/bib_ref]. However, HIF-1α and HIF-2α have discrete transcriptional targets and perform non-redundant roles. Their expression patterns are tissue specific and the part played by these isoforms in tumor progression is contrasting [bib_ref] HIF-1 and HIF-2: working alone or together in hypoxia?, Ratcliffe [/bib_ref]. A study observed that HIF-1 mediated genes conferred brain FIGURE 1 | Network showing HIF-1 connected to its target genes (adapted from: http://www.grnpedia.org/trrust/). Colors inside the shape represent the type of interaction. hypoxia-induced tolerance rather than HIF-2 regulated genes [bib_ref] Brain genomic response following hypoxia and re-oxygenation in the neonatal rat, Bernaudin [/bib_ref]. They came to this conclusion as because HIF-2α levels did not increase significantly in neuronal cells following hypoxia. Further, this study showed that adrenomedullin, propyl 4-hydroxylase α, metallothionein 1 (MT-1), mitogen-activated protein kinase phosphatase 1(MKP-1), CUGBP Elav-Like Family Member (CELF), 12-lipoxygenase and carbonic anhydrase 1 (CAR-1) are regulated by hypoxia. HIF-1α was also found to bind to the caspase-3 gene promoter region specifically. Moreover, HIF-1α and procaspase-3 showed similar expression pattern following photothrombotic cerebral ischemia [bib_ref] Evidence of HIF-1 functional binding activity to caspase-3 promoter after photothrombotic cerebral..., Hoecke [/bib_ref].
Hypoxia-inducible factor responsive genes perform a significant role in the ischemic preconditioning (IPC) induced neuroprotection [bib_ref] Hypoxic preconditioning induces changes in HIF-1 target genes in neonatal rat brain, Jones [/bib_ref] [bib_ref] NCX1 Is a novel target gene for hypoxia-inducible factor-1 in ischemic brain..., Valsecchi [/bib_ref]. Up-regulation of cytochrome P450 2C11 by HIF-1α has a role in the tolerance induced by hypoxic preconditioning (HPC) in astrocytes [bib_ref] Hypoxic preconditioning and tolerance via Hypoxia Inducible Factor (HIF) 1α-linked induction of..., Liu [/bib_ref]. 2-methoxyestradiol (2ME2) is a metabolite of estrogen and an inhibitor of HIF-1α . A study by Zhou and co-workers showed that 2ME2 aggravated the CA1 hippocampal cell death after global ischemia in rats . Glucose at a concentration of 25 mM up-regulated the expression of HIF-1α in rat primary cortical neurons exposed to hypoxia. Surprisingly, the study showed that along with hypoxia, a reducing environment is required for the stabilization of HIF-1α in neuronal cells [bib_ref] Glucose upregulates HIF-1α expression in primary cortical neurons in response to hypoxia..., Guo [/bib_ref]. The above lab also found out that BCL2 interacting protein 3 (BNIP3) and BCL2 interacting protein 3 like (NIX/BNIP3L) play a role in the HIF-1α mediated neuroprotection during ischemia-reperfusion [bib_ref] Role of HIF-1a in regulating autophagic cell survival during cerebral ischemia reperfusion..., Guo [/bib_ref].
Sodium-calcium exchanger-1 (NCX1) helps in the maintenance of Na + and Ca 2+ homeostasis [bib_ref] Function and regulation of the Na -Ca2 exchanger NCX3 splice variants in..., Michel [/bib_ref]. In a recent study, it was demonstrated that NCX1 gene is controlled by HIF-1 and contributes to the HIF-1 mediated neuroprotection during brain preconditioning [bib_ref] NCX1 Is a novel target gene for hypoxia-inducible factor-1 in ischemic brain..., Valsecchi [/bib_ref]. The mammalian target of rapamycin (mTOR) belongs to a lipid kinases family, and it is involved in cell growth, survival, and proliferation [bib_ref] The role of mammalian target of rapamycin (mTOR) in insulin signaling, Yoon [/bib_ref]. mTOR facilitated neuroprotection under ischemic conditions through the activation of HIF-1α, and by the regulation of VEGF expression and neuronal apoptosis in developing rat brain .
A few studies showed that deletion of PHD2 in neuronal cells has beneficial effects like improvement in histological and functional outcomes and neovascularisation through the promotion of HIF-VEGF axis . Supporting this observation, knockdown of HIF-1α before oxidative stress aggravated the impairment of learning and memory in rat model [bib_ref] Hypoxia inducible factor 1α promotes endogenous adaptive response in rat model of..., Yang [/bib_ref].
The effect of HIF-1α stabilization is not much studied in glial cells as in neurons. In a study by [bib_ref] The good, the bad, and the cell type-specific roles of hypoxia inducible..., Vangeison [/bib_ref] explored the impact of knocking down HIF-1α specifically in astrocytes in neuron/astrocyte co-cultures exposed to hypoxia. Selective deletion of HIF-1α in astrocytes significantly protected neurons from hypoxia-induced neuronal death in co-cultures. Myeloid-specific knockout (KO) of HIF-1α resulted in faster behavioral recovery of mice subjected to MCAO [bib_ref] Hypoxiainducible factor-1α regulates microglial functions affecting neuronal survival in the acute phase..., Bok [/bib_ref]. The HIF-1α KO mice had fewer infiltrating microglia and apoptotic neurons in the hippocampus following MCAO compared to its wild-type counterpart. The above group also observed that HIF-1α governs microglial phagocytosis and production of ROS and TNF-α under ischemic conditions. Sirtuin 1 (Sirt 1) has many vital roles in response to metabolic stress, especially cerebral ischemia [bib_ref] Sirt1 in cerebral ischemia, Koronowski [/bib_ref]. A study by [bib_ref] Sirt1 regulates glial progenitor proliferation and regeneration in white matter after neonatal..., Jablonska [/bib_ref] observed that the presence of HIF-1α plays an essential role in hypoxia-induced Sirt1 expression in oligodendrocyte progenitor cell (OPC).
## The pro and anti-apoptotic face of hif-1α stabilization in ischemic stroke
HIF-1 has both pro and anti-apoptotic features. It improved the expression of genes like EPO and VEGF that are linked to several pathways related to neuroprotection. However, it also participates in the pro-apoptotic process by stabilizing the tumor suppressor protein p53 during severe hypoxia [bib_ref] The role and regulation of hypoxia-inducible factor-1α expression in brain development and..., Fan [/bib_ref]. Further, it promoted cell necrosis in collaboration with calcium and calpain. It was found to increase blood-brain barrier (BBB) permeability and thus heighten brain oedema.
In agreement with the pro-apoptotic role of HIF-1, a few studies found that HIF-1deficiency can confer neuroprotection against acute hypoxia-induced cell death in mice [bib_ref] Brain-specific knock-out of hypoxia-inducible factor-1 reduces rather than increases hypoxic-ischemic damage, Helton [/bib_ref]. A similar trend was observed in a study by [bib_ref] Early inhibition of HIF-1α with small interfering RNA reduces ischemic-reperfused brain injury..., Chen [/bib_ref] that used HIF-1α siRNA. The authors showed that HIF-1α siRNA treatment protected neurons from ischemic injury in vivo through the inhibition of VEGF and apoptosis-related proteins like p53 and Caspase-3 along with HIF-1α. The c-glycosylated flavone, vitexin (5, 7, 4-trihydroxyflavone-8-glucoside) is a HIF-1α inhibitor [bib_ref] Vitexin reduces hypoxia-ischemia neonatal brain injury by the inhibition of HIF-1alpha in..., Min [/bib_ref]. Inhibition of HIF-1α in the early stages of neonatal cerebral ischemia with vitexin conferred neuroprotection in vivo.
Further, [bib_ref] HIF-1 alpha inhibition ameliorates neonatal brain damage after hypoxic-ischemic injury, Chen [/bib_ref] observed that inhibition of HIF-1α by 2ME2 5min after the hypoxia-ischemia protected the neuronal cells in the neonatal rat. There was no significant reduction in infarct volume when the treatment with the 2ME2 was delayed for 3hrs. HIF-1 inhibition also protected BBB and reduced brain oedema. While, the treatment with HIF-1α stabilizer DMOG, increased BBB permeability and brain oedema. Additionally, [bib_ref] Neuronal HIF-1α and HIF-2α deficiency improves neuronal survival and sensorimotor function in..., Barteczek [/bib_ref] showed that HIF-1 and HIF-2 are not needed for cell survival under hypoxic conditions. This study also showed that HIF-1/2 deficiency might protect neurons from early neuronal cell death and neurological impairment caused by ischemic stroke.
Additionally, [bib_ref] Prosurvival and prodeath effects of hypoxia-inducible factor-1α stabilization in a murine hippocampal..., Aminova [/bib_ref] showed that the prodeath or anti-death role of HIF-1α in a cell depends on the type of death stimulus. HIF-1α overexpressing HT22 cells showed increased sensitivity to glutamate toxicity, but they were more resistant to cell death induced by camptothecin or tunicamycin or thapsigargin. When the level of p53 is low, hypoxia-induced HIF-1α triggers the transcriptional activation of adaptive genes. However, during severe and sustained hypoxia, p53 levels increase along with HIF-1α stabilization and these proteins together activate pathological genes such as BAX [bib_ref] Hypoxia-inducible factor-1α mediates hypoxia-induced delayed neuronal death that involves p53, Halterman [/bib_ref].
Tricyclodecan-9-ylxanthogenate (D609) down-regulates HIF-1α similar to 2ME2. Treatment with D609 or 2ME2 reduced infarct volume and improved neuroscore in rats following MCAO [bib_ref] Multiple effects of 2ME2 and D609 on the cortical expression of HIF-1α..., Chen [/bib_ref]. These HIF-1α inhibitors decreased the expression of VEGF, BNIP3 and cleaved caspase 3. According to a study carried out by [bib_ref] VEGF increases permeability of the blood-brain barrier via a nitric oxide synthase/cGMP-dependent..., Mayhan [/bib_ref] , VEGF promoted the BBB permeability via a nitric oxide synthase/cGMPdependent pathway. Since VEGF is directly controlled by HIF-1, the negative effect of HIF-1α in ischemic stroke may be linked to VEGF. Information from the above studies suggests that the positive or negative effects of HIF-1 depend on cell type and the severity of hypoxia. The nature of ischemic insult also plays a role in determining whether HIF-1α will advocate pro-death or pro-survival pathways [bib_ref] Hypoxia inducible factor 1 as a therapeutic target in ischemic stroke, Shi [/bib_ref].
## Biphasic expression pattern of hif-1α following ischemic injury
Biphasic expression pattern of HIF-1α was observed in some of the studies following in vitro and in vivo models of cerebral ischemia. A study by [bib_ref] MiR-335 regulates Hif-1α to reduce cell death in both mouse cell line..., Liu [/bib_ref] monitored the changes in the HIF-1α gene and protein expression at different time points in a middle cerebral artery occlusion (MCAO) model using an embolus (eMCAO). HIF-1α mRNA level gradually increased following ischemia and peaked at 6 h. After that, it slowly decreased and came to a low level at 24 h. A similar pattern of expression was observed from 24 to 168 h reaching the peak at 72 h. An exactly same trend was seen in the case of HIF-1α protein expression.
According to a study by [bib_ref] Selective inhibition of early-but not late-expressed HIF-1α is neuroprotective in rats after..., Yeh [/bib_ref] , inhibition of HIF-1α in the early time point (0.5 h after ischemic stress) abated brain injury and oedema. While inhibiting HIF-1α at a later stage (8 h after ischemic injury) increased brain damage and decreased VEGF expression. Theoretically, HIF-1α induction at the early time points of ischemia activates apoptotic pathways, and at late time points, it promotes cell survival pathways. Comparable expression pattern of HIF-1α was observed in other study carried out by [bib_ref] Neuron-specific inactivation of the hypoxia inducible factor 1 increases brain injury in..., Baranova [/bib_ref] in the MCAO model.
In neonatal rat brain, upon MCAO, HIF-1α protein expression was elevated immediately (0 h, without reperfusion) and peaked at 8 h, and then deteriorated at 24 h after reperfusion [bib_ref] Regulation of hypoxia-inducible factor 1α and induction of vascular endothelial growth factor..., Mu [/bib_ref]. VEGF protein expression was also similar to that of HIF-1α. This variation shows the biphasic nature of HIF-1 expression following cerebral ischemia in neonatal brain. When compared to the expression pattern in neuronal cells, similar alterations in expression of HIF-1α mRNA and protein were observed in microglia cells following exposure to hypoxic conditions . These studies have enhanced our understanding of the dual and contrasting roles of HIF-1α expression in neurons following ischemic stroke. To a certain extent, this biphasic expression pattern of HIF-1α will decide the time window of administration of drugs that act via HIF-1α stabilization for maximum protection.
## Hypoxia mimetic agents in ischemic stroke
The term "hypoxia mimetic agent" has been widely used to indicate biological or chemical molecules which are used to up-regulate HIF-1α expression. They do it either via PHDs inhibition or some other mechanisms. Here, in this review, HIF-1α stabilizers are broadly classified according to the involvement of PHDs in the stabilization process.
## Hif-1α up-regulation via inhibition of phd pathway
Most of the HIF-1α up-regulators act via PHD inhibition. Currently, the following interventions are used stabilize HIF-1α and mimic hypoxic conditions via PHD inhibition: (a) reduce Fe 2+ availability using iron chelators; (b) introduce metal ions that will compete with Fe 2+ ; (c) use 2OG analogs. HIF-1α stabilizers that do not come under any of the above categories are discussed under a separate subheading.
## Iron chelators and competitors
Iron chelators are small molecules that bind very tightly to iron and reduce the amount of free divalent iron available for PHD mediated HIF hydroxylation reaction [bib_ref] Role of the hypoxia inducible factors HIF in iron metabolism, Peyssonnaux [/bib_ref]. Deferoxamine mesylate (DFO) is a widely used iron chelator that removes excess iron from the body [bib_ref] Desferrioxamine mesylate for managing transfusional iron overload in people with transfusion-dependent thalassaemia, Fisher [/bib_ref]. Previously, it was thought that iron chelators prevent oxidative injury by obstructing hydroxyl radical formation but later studies revealed that iron chelators inhibit the action of PHDs and upregulate HIF-1 expression [bib_ref] Hypoxia-inducible factor prolyl 4-hydroxylase inhibition, Siddiq [/bib_ref]. Alternatively, iron competitors (mostly divalent metal ions) that compete with Fe 2+ in the PHD mediated HIF-1 hydroxylation process can be used. Although there are several iron competitors like Ni 2+ and Mn2 + , Co 2+ in the form of cobalt chloride (CoCl 2 ) is the most widely used iron competitor for PHD inhibition [bib_ref] Regulating cellular oxygen sensing by hydroxylation, Fandrey [/bib_ref]. A recent study byreported that Zn 2+ could selectively inhibit PHD3 over PHD2.
A study by [bib_ref] Protection from oxidative stress-induced apoptosis in cortical neuronal cultures by iron chelators..., Zaman [/bib_ref] was one of the first studies to explore HIF-1 mediated neuroprotection by iron chelators following hypoxic stress. In this study, iron chelators, DFO and mimosine (MIM) protected embryonic cortical neurons from glutathione depletion and oxidative stress-induced cell death at a concentration of 100 µM. Neuroprotective capabilities of these agents were also comparable to their ability to augment DNA binding of HIF-1. These compounds also upregulated mRNA and protein expression of HIF-1 controlled genes. Likewise, cell death induced by depletion of glutathione was decreased significantly by treatment with CoCl 2 at a concentration > 200 µM , a competitive inhibitor of iron. CoCl 2 also elevated the HIF-1α protein expression.
Subsequently, several studies explored the ability of iron chelators for HIF-1 mediated neuroprotection against hypoxic injury. [bib_ref] Hypoxia-inducible factor prolyl 4-hydroxylase inhibition, Siddiq [/bib_ref] investigated the mechanism by which DFO offered protection in neuronal cells. In cortical neurons, HIF-1α was stabilized upon treatment with 100 µM DFO and protected cells from oxidative glutamate toxicity. Similarly, [bib_ref] Desferrioxamine induces delayed tolerance against cerebral ischemia in vivo and in vitro, Prass [/bib_ref] demonstrated the neuroprotective effect of DFO against OGD injury in purified cortical neurons. DFO at a concentration of 150 µM/L induced 47% reduction in cell death in primary cortical neurons subjected to OGD.
Compound A is a novel proprietary (Fibrogen Inc., United States) small molecule iron chelator. It stabilized HIF-1α in cortical neurons and increased the expression of HIF-1 targeted genes at a concentration of 40 µM [bib_ref] Hypoxia-inducible factor prolyl 4-hydroxylase inhibition, Siddiq [/bib_ref]. It also conferred protection against oxidative glutamate toxicity. In addition to its protective effect in in vitro studies, treatment with 100 mg/kg of compound A, stabilized HIF-1 in rat brain within 3h. It also reduced infarct volume by 67%, compared to control littermates. In hippocampal neurons from newborn mice, pretreatment with 10 mM/L DFO decreased OGD induced cell death by 45% compared to the control group. The protection conferred by DFO diminished upon cell's transfection with anti-HIF-1α. This finding suggests that the protection by DFO is through HIF-1α induction [bib_ref] A role for hypoxia-inducible factor-1α in desferoxamine neuroprotection, Hamrick [/bib_ref].
Prolyl hydroxylases isoforms differ in their expression patterns, tissue distribution, subcellular localization, and their ability to hydroxylate HIF-1α [bib_ref] Selective inhibition of Hypoxia-Inducible Factor (HIF) prolylhydroxylase 1 mediates neuroprotection against normoxic..., Siddiq [/bib_ref]. Based on these characteristics, Siddiq and co-workers hypothesized that the neuroprotection conferred by PHD inhibitors are PHD isoform-specific and independent of HIF-1. In this regard, they investigated the role of PHD and HIF isoforms in giving neuroprotection against normoxic oxidative death by the treatment of PHD inhibitors in primary rat neurons. When compared to a previous in vitro study carried out by [bib_ref] Protection from oxidative stress-induced apoptosis in cortical neuronal cultures by iron chelators..., Zaman [/bib_ref] , they observed something distinctive in this study. They noted that the prevention of normoxic oxidative neuronal death by DFO is through the inhibition of PHD1 and HIF-1α doesn't have any role in this mechanism. They also noted that its HIF-2α, not HIF-1α controls the sensitivity to normoxic oxidative neuronal death.
In comparison to adult CNS, the neonatal brain is highly sensitive to the availability of essential substrates. Treatment with 60 mg/kg of CoCl 2 and 200 mg/kg of DFO increased the level of HIF-1α protein in rat pups compared to the vehicleinjected controls [bib_ref] Role of hypoxia-inducible factor-1 in hypoxiainduced ischemic tolerance in neonatal rat brain, Bergeron [/bib_ref]. Compared with that in vehicle-injected controls, preconditioning with DFO and CoCl 2 24h before hypoxia-ischemia gave 56 and 75% protection, respectively.
Desferrioxamine promoted the binding of HIF-1 to DNA and transcription of EPO in vivo. There was a 20-fold increase in DNA binding of HIF-1 upon DFO application. The tolerance induced by DFO was time and dosage-dependent [bib_ref] Desferrioxamine induces delayed tolerance against cerebral ischemia in vivo and in vitro, Prass [/bib_ref]. Expression of GLUT1 protein was elevated in rat brain by the treatment of 300 mg/kg of DFO. Brain lesion area and whole brain cell loss reduced upon treatment with DFO. Notably, DFO treatment didn't affect striatal lesion but provided a 20% reduction in cortical injury compared with the vehicle group. Likewise, there was a 62% decrease in thalamic shrinkage by the treatment of DFO in comparison with vehicle-treated animals. Neurological score and sensorimotor performances also improved upon treatment with DFO [bib_ref] Delayed administration of deferoxamine reduces brain damage and promotes functional recovery after..., Freret [/bib_ref]. Deferasirox (DFR) is an orally administrated iron chelator and is in clinical use with high tolerance [bib_ref] Synthetic and natural iron chelators: therapeutic potential and clinical use, Hatcher [/bib_ref]. As previously observed in the study by [bib_ref] Selective inhibition of Hypoxia-Inducible Factor (HIF) prolylhydroxylase 1 mediates neuroprotection against normoxic..., Siddiq [/bib_ref] [bib_ref] Prophylactic neuroprotection against stroke: low-dose, prolonged treatment with deferoxamine or deferasirox establishes..., Zhao [/bib_ref] also showed that neuroprotection conferred by DFO and DFR following MCAO does not require HIF-1α in mice. They also observed that both DFO and DFR do not induce expression of HIF-1 target genes.
2, 2 -dipyridyl (DP) is a liposoluble iron chelator, and it is used to regulate HIF-1α expression up. Experiments show that DP treatment reduced infarct volume and maximal infarct area and thus significantly decreased histological damage [bib_ref] Cytoprotective efficacy and mechanisms of the liposoluble iron chelator 2, 2'-dipyridyl in..., Demougeot [/bib_ref]. There was a 42% reduction of infarct volume and a 25% decrease in the maximal infarct area upon treatment with DP. It protected both endothelial cells and neurons from ischemic injury. Besides, DP treatment in rats reduced ischemia-induced reactive oxygen species (ROS) production and suppressed the transformation of penumbra into infarct.
In chemical photothrombosis stroke models, DP treatment hindered the progression of infarct core and the surrounding regions [bib_ref] Apoptotic cell death progression after photothrombotic focal cerebral ischaemia: effects of the..., Hoecke [/bib_ref]. The density of apoptotic bodies reduced upon treatment with DP in both infarct core and surrounding pale region. Compared to the vehicletreated animals, DP treatment decreased and limited the DNA fragmentation only to P1 region. Upon treatment with DP, there was a significant reduction in procaspase-9 cleavage in the P1 and P2 regions. Similarly, procaspase-3 cleavage was decreased in these regions. There was also a drop in cleaved caspase-3 in the entire infarct region upon treatment with DP in vivo.
Neuroprotection achieved by DP post-treatment was lower when compared with DP pretreatment [bib_ref] Neuron-specific inactivation of the hypoxia inducible factor 1 increases brain injury in..., Baranova [/bib_ref]. Further, the protective effect produced through the administration of DP before or after MCAO was significantly attenuated but not completely abolished in neuron-specific HIF-1α-deficient mice. This shows the involvement of an alternative mechanism in giving neuroprotection upon DP treatment. Similarly, another study [bib_ref] Hypoxic preconditioning induces changes in HIF-1 target genes in neonatal rat brain, Jones [/bib_ref] observed that there was no change in expression of genes modulated by HIF-1 upon preconditioning with CoCl 2 suggesting a surrogate process may be involved in the induction of tolerance by CoCl 2 in the brain of newborn rats.
## 2og analogs
Numerous hydroxybenzenes, hydroxybenzoic acids and analogous compounds have similar structural features as both 2-oxoglutarate and ascorbate. They can compete with these molecules and inhibit prolyl 4-hydroxylase enzymes [bib_ref] Partial identity of the 2-oxoglutarate and ascorbate binding sites of prolyl 4-hydroxylase, Majamaa [/bib_ref]. 10 µM dihydroxybenzoic acid (DHB) stabilized HIF-1α and increased the transcription of HIF-1 dependent genes. DHB conferred protection to cortical neurons exposed to oxidative glutamate toxicity [bib_ref] Hypoxia-inducible factor prolyl 4-hydroxylase inhibition, Siddiq [/bib_ref].
N-oxalylglycine can inhibit most of the 2OG-dependent oxygenases and other enzymes [bib_ref] The broad spectrum 2-oxoglutarate oxygenase inhibitor N-oxalylglycine is present in rhubarb and..., Al-Qahtani [/bib_ref]. Dimethyloxalylglycine (DMOG) is a readily cell permeable ester of N-oxalylglycine. DMOG has a similar structure to 2OG. It competes with PHD co-substrate, 2OG and prevents HIF-1α degradation [bib_ref] Selective inhibition of Hypoxia-Inducible Factor (HIF) prolylhydroxylase 1 mediates neuroprotection against normoxic..., Siddiq [/bib_ref] [bib_ref] Neuroprotection by dimethyloxalylglycine following permanent and transient focal cerebral ischemia in rats, Nagel [/bib_ref]. Twenty four hour treatment with DMOG increased the protein expression of HIF-1α and mRNA levels of VEGF in cortical neurons growing under normoxic condition. Preconditioning with DMOG for 24 h before OGD significantly reduced OGD-induced cell death in cortical neuron cell. Similarly, post-treatment with DMOG also reduced cell death. This protection abrogated by the pretreatment of HIF-1α-shRNA showed the need of HIF-1α for achieving neuroprotection using PHD inhibitors [bib_ref] Inhibition of prolyl hydroxylases by dimethyloxaloylglycine after stroke reduces ischemic brain injury..., Ogle [/bib_ref].
In vivo stabilization of HIF-1 was observed within 6 h of administration of DHB. There was a significant reduction of infarct volume upon DHB treatment compared to control [bib_ref] Hypoxia-inducible factor prolyl 4-hydroxylase inhibition, Siddiq [/bib_ref]. Studies in mice showed that intraperitoneal injection of DMOG increased the level of HIF-1α protein up to threefold in a time-dependent manner in mouse brain cortex. It also elevated the expression of HIF-1 controlled genes that regulate erythropoietin and pyruvate dehydrogenase kinase-1. Further, post-OGD treatment with DMOG reduced the infarct volume in an animal stroke model. In addition to this, it stimulated pro-apoptotic caspase-3 protein, curtailed behavioral deficits and decreased the loss of local blood flow in the middle cerebral artery territory. These attributes nullified upon inhibition of HIF-1α by Digoxin [bib_ref] Inhibition of prolyl hydroxylases by dimethyloxaloylglycine after stroke reduces ischemic brain injury..., Ogle [/bib_ref].
In another study by [bib_ref] Neuroprotection by dimethyloxalylglycine following permanent and transient focal cerebral ischemia in rats, Nagel [/bib_ref] , DMOG reduced ischemic injury and improved behavior and neuroscore after both permanent and transient MCAO in adult male Wistar rats. Regional cerebral blood flow also improved. There was also a reduction in the amount of BBB breakdown. A consistent increase in the expression of both mRNA and protein levels of VEGF and endothelial nitric oxide synthase observed after DMOG treatment. As seen in the case of DFO, a study by Siddiq and co-workers suggested that the neuroprotection exerted by DMOG and DHF does not require HIF-1α stabilization and instead inhibit the PHD1 isoform [bib_ref] Hypoxia-inducible factor prolyl 4-hydroxylase inhibition, Siddiq [/bib_ref].
## Phd inhibitors with an alternative mechanism of action
This section also includes HIF-1α stabilizers, but their exact mechanism of PHD inhibition is not either mentioned in the source paper or different from the above classification. These molecules may act by binding directly to the active site of the PHDs without mimicking 2OG or chelating the iron atom. A study [bib_ref] Inhibition of HIF prolyl-4-hydroxylases by FG-4497 reduces brain tissue injury and edema..., Reischl [/bib_ref] focusing on reducing neuronal death following ischemic stroke via maintaining BBB integrity used a novel PHD inhibitor, FG-4497. HIF-1α stabilized upon treatment with FG-4497 in both primary murine astrocytes and the murine cerebrovascular endothelial cell line. FG-4497's capability to inhibit HIF-1α proteolysis was higher than DMOG. It also up-regulated the transcription of HIF-1 targeted genes VEGF and EPO. In mouse hippocampal culture, FG-4497 significantly enhanced the survival of cells following OGD.
Intraperitoneal injection of FG-4497 elevated the amount of HIF-1α in cerebral tissue and significantly reduced infarct size compared to vehicle-treated mice [bib_ref] Inhibition of HIF prolyl-4-hydroxylases by FG-4497 reduces brain tissue injury and edema..., Reischl [/bib_ref]. FG-4497 pre-treatment reduces infarct volume in transient MCAO models. Post-treatment with FG-4497 also showed neuroprotection following permanent cerebral ischemia.
2-(1-chloro-4-hydroxyisoquinoline-3-carboxamido) acetic acid (IOX3) is a novel inhibitor of PHDs [bib_ref] HIF prolyl hydroxylase inhibition prior to transient focal cerebral ischaemia is neuroprotective..., Chen [/bib_ref]. It has a structure similar to FG2216, which promoted the expression of EPO in vivo. Administration of IOX3 24h before a MCAO improved neuroscores and reduced infarct volume in mice. It also elevated the level of HIF-1α protein and EPO expression. However, injection of IOX3 soon before the MCAO showed no significant neuroprotection. IOX3 also prevented BBB disruption when treated 24 h before MCAO. GSK360A is an orally active PHD inhibitor that has reduced myocardial infarct size in the rat and murine heart [bib_ref] HIF-1 reduces ischaemia-reperfusion injury in the heart by targeting the mitochondrial permeability..., Ong [/bib_ref]. Treatment with GSK360A increased the plasma EPO and VEGF mRNA and protein levels [bib_ref] The prolyl 4-hydroxylase inhibitor GSK360A decreases post-stroke brain injury and sensory, motor,..., Zhou [/bib_ref]. When compared to the vehicle-treated group, pretreatment with GSK360A reduced post-stroke surgery neurological deficits, cognitive dysfunction and infarct volume 4 weeks after MCAO. Another novel PHD inhibitor, TM6008 decreased cell death and up-regulated HIF-1α levels in vitro. It also induced the protein expression of HIF-1 downstream genes, HO-1, EPO, and glucose transporter 3 (GLUT3) [bib_ref] A novel prolyl hydroxylase inhibitor protects against cell death after hypoxia, Kontani [/bib_ref].
Lately, molecular docking and simulation studies carried out in our lab found folic acid (FA) to be an ideal compound that binds to PHD2, FIH and pVHL alike at their HIF-1α binding region [bib_ref] Folic acid exerts postischemic neuroprotection in vitro through HIF-1α stabilization, Davis [/bib_ref]. In vitro experiments showed that postischemic treatment with FA up-regulated HIF-1 expression and its downstream genes. Further, FA conferred neuroprotection following OGD and promoted angiogenesis.
## Hif-1α up-regulation without the inhibition of phds
Non-PHD HIF-1α stabilizers relay on a mechanism other than the classical PHD-HIF pathway to up-regulates HIF-1α. Compared to the conventional PHD inhibitors, these molecules are newly discovered. HIF-1α stabilization through IPC is not discussed in this review as it is out of the scope of the title theme.
A study by [bib_ref] MiR-335 regulates Hif-1α to reduce cell death in both mouse cell line..., Liu [/bib_ref] showed that miR-335 is a direct regulator of HIF-1α, and has shown an inverse expression profile both in vivo and in vitro ischemic models. Treatment with miR-335 decreased infarct volume immediately after MCAO, and miR-335 inhibitor was found to be beneficial at 24 h after MCAO in a rat model. Thus, regulation of HIF-1α expression by miR-335 at various time points manipulated the genes required for adaption to hypoxia and ultimately resulted in the reduction of infarct volume.
Preconditioning with 1.5% isoflurane protected neurons from oxygen-glucose deprivation (OGD) injury by the induction of HIF-1α protein, and this activation is linked to the Erk1/2 pathway [bib_ref] Isoflurane preconditioning activates HIF-1α, iNOS and Erk1/2 and protects against oxygen-glucose deprivation..., Li [/bib_ref]. Further, a recent study by the above group showed the neuroprotection provided by isoflurane postconditioning via HIF-1α stabilization [bib_ref] Induction of inducible nitric oxide synthase by isoflurane post-conditioning via hypoxia inducible..., Li [/bib_ref]. Postischemic treatment with isoflurane reduced the infarct volume and improved neurological scores. Both post and pre-ischemic treatment with isoflurane treatment up-regulated iNOS mRNA level.
N-acetylcysteine (NAC) has various therapeutic properties, and it is an antioxidant [bib_ref] A minireview on N -acetylcysteine: an old drug with new approaches, Dhouib [/bib_ref]. It is also protected the brain from ischemic injury in preclinical studies [bib_ref] Hypoxia-inducible factor 1 contributes to N-acetylcysteine's protection in stroke. Free Radic, Zhang [/bib_ref]. A study by Zhang and colleagues observed that pretreatment with 150 mg/kg of NAC increased the protein levels of HIF-1α, EPO and GLUT3 in the ipsilateral hemispheres of rodents. The group also found out that HIF-1 is responsible for NAC mediated neuroprotection following MCAO. NAC treatment up-regulated heat shock protein 90 (Hsp90) expression and in turn interaction of Hsp90 to HIF-1.
A study by [bib_ref] Relative contribution of rolyl Hydroxylasedependent and -independent degradation of HIF-1alpha by proteasomal..., Badawi [/bib_ref] reported that 20S and 26S proteasomal pathways were involved in HIF-1α degradation in ischemic neurons. For the in vitro studies, MG-132, a proteasome inhibitor was used and it stabilized HIF-1α protein better than DMOG. A combination of MG-132 and DMOG had a better effect on HIF-1α stabilization than them individually. Further, they observed that treatment with epoxomicin, another proteasome inhibitor reduced infarct size and brain oedema following oxidative stress in vivo by the stabilization of HIF-1α. In vivo also a combination of epoxomicin and DMOG yielded better protection. The study also showed that proteasome inhibitors are more effective than PHD inhibitors in providing HIF-1α stabilization and neuroprotection. [bib_ref] The novel proteasome inhibitor BSc2118 protects against cerebral ischaemia through HIF1A accumulation..., Doeppner [/bib_ref] tested a novel proteasome inhibitor BSc2118 for its neuroprotective activity following ischemic stroke. 12h pre-ischemic injection of BSc2118 conferred significant reduction of infarct volumes on day four. BSc2118 yielded long-term neuroprotection for as long as 3 months with the post-stroke treatment. Apart from reducing the infarct volume, BSc2118 was able to decrease brain oedema, and BBB break down. Further, it promoted both angiogenesis and neurogenesis. Using the knock out studies, they concluded that all these positive effects of BSc2118 were mediated by HIF-1α.
Tilorone is a low molecular weight antiviral agent [bib_ref] Efficacy of tilorone dihydrochloride against ebola virus infection, Ekins [/bib_ref]. In a study by [bib_ref] Small molecule activation of adaptive gene expression, Ratan [/bib_ref] tilorone at a concentration of 100 mg/kg increased the stabilization of HIF-1α and elevated the expression of its downstream genes in vitro. The stabilization of HIF-1α by tilorone was independent of iron chelation and HIF-PHD inhibition in vitro. Tilorone also significantly reduced infarct volume in vivo following MCAO.
Huang-Lian-Jie-Du-Tang (HLJDT) is a traditional Chinese medicine with properties like heat-clearing and detoxification [bib_ref] Preconditioning with the traditional Chinese medicine Huang-Lian-Jie-Du-Tang initiates HIF-1α-dependent neuroprotection against cerebral..., Zhang [/bib_ref]. In cerebral cortical neurons, pretreatment with HLJDT increased HIF-1α, EPO and VEGF levels and protected the cells against OGD. It decreased ischemia-induced apoptosis and promoted proliferation in neuronal cells. HLJDT also significantly reduced cerebral infarction; cerebral water content and improved neurological deficient score in MCAO rat models. All the hypoxia mimetic agents discussed in this review either act by inhibiting PHDs or other mechanisms and they ultimately regulate the expression of HIF-1. Tables 1, 2 gives a summary of the studies that used hypoxia-mimetic agents to obtain neuroprotection following ischemic injury.
# Conclusion
Hypoxia causes the cell to accumulate HIF-1 in the nucleus and up-regulate specific HIF-1 targeted genes so that cells can overcome the adverse condition. All the hypoxia mimetic agents are designed to act in such a way to boost this endogenous mechanism. One way to do this is through the inhibition of PHDs. These PHD inhibitors come under either the iron chelator and competitive inhibitor category or the 2OG analog category. Other molecules rely on mechanisms other than PHD inhibition to stabilize HIF-1α. Apart from a few newly discovered molecules, widely used hypoxia-mimetic agents are DFO, DHB, CoCl 2 , DP, and DMOG. These agents up-regulated HIF-1α levels and regulated its downstream genes.
Further, these agents conferred significant neuroprotection both in vitro and in vivo experiments. As the progression of ischemic brain injury is multifactorial, each of the HIF-1 regulated genes influences various cellular processes of infarct advancement at different time points. In most of the cases, treatment with hypoxia-mimetic agents several hours before or after the induction of hypoxia bestowed greater protection compared to treatment with these agents soon before or after oxidative stress. The neuroprotection conferred by these agents also depends on the cell type and magnitude of the ischemic injury.
However, there are studies which did not show neuroprotection via HIF-1 up-regulation and even demonstrated its pro-death features. Therefore, the extent of neuroprotection provided by HIF-1 in the case of ischemic injury is still a debatable topic. Nevertheless, in light of studies discussed in this review [fig_ref] FIGURE 2 |: Domain structure of human HIF subunits and isoforms [/fig_ref] ; we can conclude that HIF-1's beneficial characteristics appear to outweigh its adverse effects. Many of the harmful effects of HIF-1α can be eliminated by carefully choosing the time of administration of HIF-1α stabilizers. Type of hypoxia mimetic agent used, nature of ischemic injury and kind of cell targeted by the treatment are the other factors which affect the neuroprotection bestowed by HIF-1α stabilizers.
One drawback of using the conventional mimetic agent is that several other enzymes use Fe 2+ and 2OG as co-factor or cosubstrate. Therefore the use of iron chelators, competitive iron inhibitors and 2OG analogs can create undesired interference of different pathways. In this regard, the introductions of novel drugs that can specifically inhibit PHDs or stabilize HIF-1α through other mechanisms need more consideration. Specificity of PHD inhibitors can be achieved by the development of molecules that can bind to the active sites in PHDs. These agents could be analogous to the parts of HIF-1α that interact with PHDs. The possibility of stabilizing HIF-1α via FIH inhibition has not received much attention and warrants further exploration.
# Author contributions
GR and AM conceived the idea. CD wrote the manuscript. SJ, GR, AM, and O-NB helped to draft the manuscript.
[fig] FIGURE 2 |: Domain structure of human HIF subunits and isoforms. [/fig]
[fig] FIGURE 3 |: Regulation of HIF-1 during hypoxia and normoxia. [/fig]
[fig] FUNDING: AM is funded by MRC grant (reference no: MR/R005923/1). [/fig]
[table] TABLE 1 |: Details about the studies that used non-PHD HIF-1α stabilizers as hypoxia-mimetic agents. [/table]
|
Screw-assisted 3D printing with granulated materials: a systematic review
This paper presents a systematic review on extrusion additive manufacturing (EAM), with focus on the technological development of screw-assisted systems that can be fed directly with granulated materials. Screw-assisted EAM has gained importance as an enabling technology to expand the range of 3D printing materials, reduce costs associated with feedstock fabrication, and increase the material deposition rate compared to traditional fused filament fabrication (FFF). Many experimental printheads and commercial systems that use some screw-processing mechanism can be found in the literature, but the design diversity and lack of standard terminology make it difficult to determine the most suitable solutions for a given material or application field. Besides, the few previous reviews have offered only a glimpse into the topic, without an in-depth analysis about the design of the extruders and associated capabilities. A systematic procedure was devised to identify the screw-assisted EAM systems that can print directly from granulated materials, resulting in 61 articles describing different pieces of equipment that were categorized as experimental printheads and commercial systems, for small-and large-scale applications. After describing their main characteristics, the most significant extruder modifications were discussed with reference to the materials processed and performance requirements. In the end, a general workflow for the development of 3D printers based on screw extrusion was proposed. This review intends to provide information about the state-of-the-art screw-assisted EAM and help the academy to identify further research opportunities in the field.
# Introduction
According to the standard terminology, the extrusion additive manufacturing (EAM) technique is characterized by the selective dispensing of softened materials through a nozzle, which generates continuous strands that are usually deposited layer upon layer to print a 3D part. EAM systems, from which the fused filament fabrication (FFF) equipment derived from the Stratasys' Fused Deposition Modeling (FDM®) patentare the most widespread, have benefited from simple constructive solutions that are cheap to implement and easy to operate [bib_ref] Additive manufacturing of metallic based on extrusion process: a review, Nurhudan [/bib_ref] , favoring its extensive adoption for the fabrication of prototypes or end-user products in multiple application fields [bib_ref] Additive manufacturing of metallic and ceramic components by the material extrusion of..., Gonzalez-Gutierrez [/bib_ref] [bib_ref] From materials to devices using fused deposition modeling: A state-of-art review, Zhang [/bib_ref].
Although nowadays the filament-based 3D printers can be made more robust, scalable, and compatible with a range of polymers, the quest for novel materials and applications, higher deposition speed, and reduction of costs associated with feedstock have motivated the development of alternative EAM machines based on piston and screw mechanisms [bib_ref] Additive manufacturing of metallic based on extrusion process: a review, Nurhudan [/bib_ref] [bib_ref] Additive manufacturing of metallic and ceramic components by the material extrusion of..., Gonzalez-Gutierrez [/bib_ref] [bib_ref] From materials to devices using fused deposition modeling: A state-of-art review, Zhang [/bib_ref]. In fact, the utilization of a screw mechanism with continuous feeding of granulated materials, similarly to industrial polymer extruders, is particularly interesting to allow a more precise control over the extrusion process, associated with improved capacity to generate pressure, homogenize the molten material, and allow some level of mixing.
Previous reviews on EAM with focus on screw-assisted printheads were not found. Gonzalez-Gutierrez et al. [bib_ref] Additive manufacturing of metallic and ceramic components by the material extrusion of..., Gonzalez-Gutierrez [/bib_ref] and Rane, Strano [bib_ref] A comprehensive review of extrusionbased additive manufacturing processes for rapid production of..., Rane [/bib_ref] described the filament-, piston-, and screwbased extrusion mechanisms used to 3D print metallic and ceramic components from solid materials in the form of filaments or granules [bib_ref] Additive manufacturing of metallic and ceramic components by the material extrusion of..., Gonzalez-Gutierrez [/bib_ref] , as well as pastes or suspensions [bib_ref] A comprehensive review of extrusionbased additive manufacturing processes for rapid production of..., Rane [/bib_ref]. However, few paragraphs were dedicated to discuss the screw-assisted systems, which were mostly used due to the viscosity of the highly filled materials processed. A review on 3D printing of engineering thermoplastics, which usually require high processing temperatures, has also described the use of screw-assisted systems [bib_ref] Current understanding and challenges in high temperature additive manufacturing of engineering thermoplastic..., Das [/bib_ref]. However, most of the described machine modifications focused on the temperature of the build environment temperature, which should be controlled to help softening the material, reduce thermal gradients in the part and avoid damage of sensitive components. Zhang et al. [bib_ref] From materials to devices using fused deposition modeling: A state-of-art review, Zhang [/bib_ref] mentioned the screw-assisted printheads as important improvements of EAM, which could favor 3D printing with innovative materials (e.g., micro-or nano-reinforced polymers), and lead to faster and more precise deposition.
As shown, the reviews available are usually oriented by the material type or application field, and usually mention few examples of screw-assisted 3D printing equipment, overlooking important aspects of the screw extrusion technology. To the best of the authors' knowledge, the literature lacks a review on the design of the extruders used in EAM, with focus on the evolution of this technology, the main printhead modifications and their motivations, as well as on the typical development stages taken to carry out research in that field.
In that context, the present systematic review aims to investigate the technological development of screw-assisted EAM, evidencing the different solutions proposed for the extrusion units of the 3D printheads both from experimental and commercially available machines. Although a variety of materials in different states can be processed in such equipment, the scope of this review was narrowed to the systems that could be fed with granulated materials, so that the role of the screw not only as a conveying mechanism but also as an active processing component could be noticed.
This review paper consists of seven sections. Section 2 begins with the methodology used to carry out the literature review. Section 3 presents the search output and a short analysis of the publications, evidencing the main trends in the research with screw-assisted systems. In Section 4, the technical evolution of the equipment is discussed, providing a general overview of the different designs proposed for experimental systems as well as the main features of commercially available equipment. Section 5 summarizes the main modifications of the extrusion units with reference to the materials processed and performance requirements. Section 6 discusses the general development workflow for screw-assisted printheads, with insights relative to the design, operationalization, and assessment of the systems. Finally, the main conclusions of this systematic review are drawn in Section 7.
# Research method
A review protocol was devised according to the guidelines presented by Siddaway et al. [bib_ref] How to do a systematic review: a best practice guide for conducting..., Siddaway [/bib_ref]. Three electronic databases (ScienceDirect, Scopus, and WebofKnowledge) were used, with the keywords divided into two strings, as shown in [fig_ref] Table 1: Search strings used according to the electronic database [/fig_ref]. The survey covered the period from 2000 to 2020 and considered only documents labeled as research articles.
For each string, the search results from the three databases were exported to the Endnote® online reference manager, to eliminate the duplicates and perform the screening process. This was done by reading the articles' title, abstract, and keywords considering the following inclusion criteria: (1) the 3D printing process should be based on material extrusion; (2) material conveying and/or melting and/or pressurization should be performed by a screw mechanism; (3) solid particles should be directly fed to the 3D printer.
The pre-selected articles were read thoroughly, to make sure that only functional equipment with which actual processing/deposition experiments could have been performed were considered, to better investigate their contribution to the field of screw-assisted EAM. Finally, the articles were analyzed regarding the (i) type of equipment (i.e., experimental or commercial), (ii) application scale (i.e., small-scale or large-scale 3D printing), (iii) the main modifications of the extrusion units with relation to the materials processed and performance requirements, and (iv) the research presented in each publication (i.e., system development, functionalization, performance assessment, and characterization of 3D printed parts). [fig_ref] Figure 1: Flowchart of the systematic search, showing the resulting number of articles after... [/fig_ref] presents a flowchart of the search process, showing the resulting number of papers after each step. From the 251 articles found for the two search strings, 45 duplicates were excluded and 53 met the three inclusion criteria. After full reading, 37 articles found in the systematic search and 17 crossed references were qualified to compose the repository. Besides, 7 articles previously known by the authors that described screw-assisted EAM systems still uncovered by the search were also added to the repository, totalizing 61 articles. Since this study aims to structure the knowledge about screw-assisted 3D printers that use granulated materials, the equipment explored in the different publications are listed in [fig_ref] Table 2: Selected information about the screw-assisted EAM systems, grouped according to the application... [/fig_ref]. Experimental machines are referenced by the name of the first author, while commercial systems are denoted by the models' name. All publications related to each table entry are indicated by the numbers in brackets. [fig_ref] Table 2: Selected information about the screw-assisted EAM systems, grouped according to the application... [/fig_ref] also shows the countries that contributed to the research with each system, as well as the data used to classify the equipment into small-and large-scale systems. These include equipment details (nozzle diameter, maximum printing volume, and screw diameter) and the values used for the main process parameters (screw rotation speed, and deposition speed).
Typically, the nozzle diameter used in small-scale systems was smaller than 1 mm, the deposition speed ranged from 1 to 30 mm/s, and the printing volume reached up to 300 × 300 × 300 mm. For large-scale systems, the nozzle diameter ranged from 0.8 to 10.1 mm, with deposition speed usually greater than 20 mm/s up to 279 mm/s, and printing volume starting at 800 × 600 × 600 mm 3 . Although not always reported, in most small-scale printheads, the screw diameter was not much larger than 15 mm, while screw diameters up to 25 mm were found in large-scale systems.
The number of publications throughout the years is summarized in according to the type of equipment and scale of application. The first years can be seen as a pioneering period, which was followed by a continuous increase in the number of publications from 2014 to 2018, reflecting the crescent interest in the topic. The reduced numbers from 2019 to 2020 might relate to the COVID-19 pandemic, which has limited the conduction of experimental research worldwide as mentioned by Billah et al. [bib_ref] Thermomechanical characterization of short carbon fiber and short glass fiber-reinforced ABS used..., Billah [/bib_ref].
The publications from the pioneering period described only experimental small-scale systems [bib_ref] Precision extruding deposition and characterization of cellular poly-ε-caprolactone tissue scaffolds, Wang [/bib_ref] [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref] [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] [bib_ref] Mechanical and in vitro evaluations of composite PLDLLA/TCP scaffolds for bone engineering, Lam [/bib_ref]. The first commercial system intended for small-scale applications was described in two publications from 2016 [bib_ref] Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration, Wang [/bib_ref] [bib_ref] Enhancing the hydrophilicity and cell attachment of 3D printed PCL/graphene scaffolds for..., Wang [/bib_ref] , the same year when an experimental equipment for potential large-scale application was introduced [bib_ref] Additive manufacturing of poly(propylene) by means of melt extrusion, Hertle [/bib_ref]. With regard to the commercial large-scale systems, the first equipment was described a year later, in two articles from 2017 [bib_ref] Structure and mechanical behavior of Big Area Additive Manufacturing (BAAM) materials, Duty [/bib_ref] [bib_ref] Thermal analysis of additive manufacturing of large-scale thermoplastic polymer composites, Compton [/bib_ref]. As will be shown in Section 4, many new experimental systems for both smalland large-scale applications appeared in the period from 2014 to 2018 , which can be related to the high number of publications in the same period. Although the scientific production in the last 2 years decreased, a crescent share of the research was carried out with commercial systems , indicating the consolidation of screw-assisted EAM in the market.
With regard to the geographic distribution, researchers from 21 countries could be identified. The USA respond for most publications [bib_ref] Fabrication of PCL/β-TCP scaffolds by 3D mini-screw extrusion printing, Dávila [/bib_ref] , followed by the UK [bib_ref] Mechanical and in vitro evaluations of composite PLDLLA/TCP scaffolds for bone engineering, Lam [/bib_ref] , and India [bib_ref] Current understanding and challenges in high temperature additive manufacturing of engineering thermoplastic..., Das [/bib_ref]. The authors from the USA account for most publications using large-scale AM systems [bib_ref] Structure and mechanical behavior of Big Area Additive Manufacturing (BAAM) materials, Duty [/bib_ref] [bib_ref] Thermal analysis of additive manufacturing of large-scale thermoplastic polymer composites, Compton [/bib_ref] [bib_ref] Rheological survey of carbon fiberreinforced high-temperature thermoplastics for big area additive manufacturing..., Ajinjeru [/bib_ref] [bib_ref] Designing for Big Area Additive Manufacturing, Roschli [/bib_ref] [bib_ref] Thermomechanical characterization of short carbon fiber and short glass fiber-reinforced ABS used..., Billah [/bib_ref] [bib_ref] Compressive deformation analysis of large area pellet-fed material extrusion 3D printed parts..., Trejo [/bib_ref] [bib_ref] Mechanical characterization of high-temperature carbon fiber-polyphenylene sulfide composites for large area extrusion..., Yeole [/bib_ref] and recycling AM [bib_ref] Fused particle fabrication 3-D printing: recycled materials' optimization and mechanical properties, Woern [/bib_ref] [bib_ref] Mechanical properties and applications of recycled polycarbonate particle material extrusion-based additive manufacturing, Reich [/bib_ref] [bib_ref] Technical pathways for distributed recycling of polymer composites for distributed manufacturing: windshield..., Dertinger [/bib_ref] [bib_ref] Towards distributed recycling with additive manufacturing of PET flake feedstocks, Little [/bib_ref]. With relation to the UK, most research was conducted on additive bio-manufacturing (bio-AM) [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] [bib_ref] A plasma-assisted bioextrusion system for tissue engineering, Liu [/bib_ref] [bib_ref] Process-driven microstructure control in melt-extrusion-based 3D printing for tailorable mechanical properties in..., Liu [/bib_ref] [bib_ref] Structural evolution of PCL during melt extrusion 3D printing, Liu [/bib_ref] [bib_ref] User interface tool for a novel plasma-assisted bio-additive extrusion system, Liu [/bib_ref] [bib_ref] Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration, Wang [/bib_ref] [bib_ref] Enhancing the hydrophilicity and cell attachment of 3D printed PCL/graphene scaffolds for..., Wang [/bib_ref] [bib_ref] Polymer-ceramic composite scaffolds: the effect of hydroxyapatite and β-tri-calcium phosphate, Huang [/bib_ref] [bib_ref] Aligned multi-walled carbon nanotubes with nanohydroxyapatite in a 3D printed polycaprolactone scaffold..., Huang [/bib_ref] [bib_ref] 3D Printing of polycaprolactone-polyaniline electroactive scaffolds for bone tissue engineering, Wibowo [/bib_ref] , defined by Ferrari et al.as the "use of 3D printing for medical purposes or non-therapeutic 'human enhancement', whether they involve the production of biological material or not." In the case of India, most publications can be attributed to the same research group that conducted many experiments on a custom-made deposition tool [bib_ref] The effect of process parameters on tensile behavior of 3D printed flexible..., Kumar [/bib_ref] [bib_ref] Investigation on the effects of process parameters in CNC assisted pellet based..., Kumar [/bib_ref] [bib_ref] Extrusion-based additive manufacturing process for producing flexible parts, Kumar [/bib_ref] [bib_ref] Experimental investigations on suitability of polypropylene (PP) and ethylene vinyl acetate (EVA)..., Kumar [/bib_ref] [bib_ref] 3D printing of flexible parts using eva material, Kumar [/bib_ref] [bib_ref] Investigations on the melt flow behaviour of aluminium filled ABS polymer composite..., Kumar [/bib_ref] [bib_ref] Analysing the influence of raster angle, layer thickness and infill rate on..., Kumar [/bib_ref].
The co-occurrence of keywords was analyzed with aid of the VOSviewer software, after correcting spelling differences and merging synonyms. Also, abbreviated terms were replaced by their corresponding full equivalents. [fig_ref] Figure 3: Network map of the co-occurrence of keywords [/fig_ref] shows the resulting network, in which the size of the circles is proportional to the frequency of each keyword, while the width of the links indicate how often two keywords were used together. The keywords that appeared together in the publications were positioned close to each other in the map.
The most frequent keywords were "additive manufacturing" and "3D printing," both occurring 29 times, followed by "fused deposition modeling," occurring 18 times. The clusters in red, yellow, and orange, in which the most frequent keywords were respectively "scaffold" (occurring 8 times), "big area additive manufacturing" (occurring 5 times), and "recycling" (occurring 5 times), relate to the three main topics of interest of the USA and the UK (i.e., bio-AM, large-scale AM, and recycling AM). Other prominent clusters revolve around the terms "screw extrusion" (10 times), "extrusion" (9 times), "pellet" (8 times), and "composites" (7 times).ranks the five most influential articles, with information about the corresponding author, journals' name, research domain, year of publication, and number of citations according to the Scopus database. Again, bio-AM and large- As one of the main researchers working on large-scale AM, Prof. C. E. Duty (University of Tennessee Knoxville, USA) was cited more than 70 times within the repository, and his first article describing the BAAM® machine [bib_ref] Structure and mechanical behavior of Big Area Additive Manufacturing (BAAM) materials, Duty [/bib_ref] appeared in the references of other 8 publications from the repository. The paper by A. Bellini [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref] was a reference for other 6 publications within the repository, while other publications with her contribution investigating the phenomena of the EAM process have been cited 12 times in the other papers. Although the same equipment developed by A. Bellini was used in the paper authored by F. Wang [bib_ref] Precision extruding deposition and characterization of cellular poly-ε-caprolactone tissue scaffolds, Wang [/bib_ref] , he was cited few times within the repository. Prof. P. Bártolo (University of Manchester, UK) is the corresponding author of most publications in the field of bio-AM, and his best cited article [bib_ref] Enhancing the hydrophilicity and cell attachment of 3D printed PCL/graphene scaffolds for..., Wang [/bib_ref] appears on the reference of other 4 papers within the repository. In sum, his publications were cited 36 times by the other papers from the repository. The paper by A. Goyanes [bib_ref] Direct powder extrusion 3D printing: fabrication of drug products using a novel..., Goyanes [/bib_ref] was not cited by the other publications; however, other works in the field of pharmaceutical AM with his contribution were found in the references.
## Technological evolution of screw-assisted eam
The technological evolution of screw-assisted EAM is shown in the in a timeline that organizes the various experimental printheads and commercial systems according to the year when they were first described in the literature, with the same notation as used in [fig_ref] Table 2: Selected information about the screw-assisted EAM systems, grouped according to the application... [/fig_ref].
As indicated by the timeline, the first screw-assisted EAM systems appeared in the years from 2002 to 2008 [bib_ref] Precision extruding deposition and characterization of cellular poly-ε-caprolactone tissue scaffolds, Wang [/bib_ref] [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref] [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] [bib_ref] Mechanical and in vitro evaluations of composite PLDLLA/TCP scaffolds for bone engineering, Lam [/bib_ref] , coinciding with the decade when low-cost filament-fed desktop machines, and alternative syringe-based printheads were developed [bib_ref] Fab@Home: the personal desktop fabricator kit, Malone [/bib_ref]. Although the first systems were mostly used in bio-AM applications [bib_ref] Precision extruding deposition and characterization of cellular poly-ε-caprolactone tissue scaffolds, Wang [/bib_ref] [bib_ref] Mechanical and in vitro evaluations of composite PLDLLA/TCP scaffolds for bone engineering, Lam [/bib_ref] , the concept was further developed from 2014 to 2019 to different purposes [bib_ref] Study of Moineau-based pumps for the volumetric extrusion of pellets, Canessa [/bib_ref] [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] [bib_ref] Direct coupling of fixed screw extruders using flexible heated hoses for FDM..., Khondoker [/bib_ref] ]. An expressive equipment variety for small-scale applications could be observed in the period, including four commercial systems (3D Discovery® [bib_ref] Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration, Wang [/bib_ref] [bib_ref] Enhancing the hydrophilicity and cell attachment of 3D printed PCL/graphene scaffolds for..., Wang [/bib_ref] , PAM® [bib_ref] Additive manufacturing of fire-retardant ethylene-vinyl acetate, Geoffroy [/bib_ref] , M3DIMAKER® [bib_ref] Direct powder extrusion 3D printing: fabrication of drug products using a novel..., Goyanes [/bib_ref] , ExAM 255® [bib_ref] Fabrication and properties of extrusion-based 3D-printed hardmetal and cermet components, Lengauer [/bib_ref]. Despite the description of the first screw-assisted EAM based on a do-ityourself (DIY) printhead kit [bib_ref] Mechanical properties of direct waste printing of polylactic acid with universal pellets..., Alexandre [/bib_ref] , no significant equipment improvements were introduced in 2020.
In turn, large-scale 3D printing started timidly only by 2016 with experimental setups adapting benchtop-sized industrial extruders to different positioning systems. Systems based on robotic arms have been considered in 2017 [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] [bib_ref] Robotic AM System for Plastic Materials: Tuning and On-line Adjustment of Process..., Magnoni [/bib_ref] , due to their higher degree-of-freedom and potentially bigger printing volume. The most innovative system incorporated a first plasticating stage followed by an auger screw for controlled deposition [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref]. Besides, the first commercial system (BAAM® [bib_ref] Structure and mechanical behavior of Big Area Additive Manufacturing (BAAM) materials, Duty [/bib_ref] was introduced in the period. New large-scale systems were last introduced in 2018, with two commercial machines (Gigabot X® [bib_ref] Fused particle fabrication 3-D printing: recycled materials' optimization and mechanical properties, Woern [/bib_ref] , Super Discovery® [bib_ref] Large-format polymeric pellet-based additive manufacturing for the naval industry, Nieto [/bib_ref] and no significant modifications in terms of extruder design. [fig_ref] Figure 5: Schematic illustrations of the pioneering screw-assisted printheads, developed by a Bellini[10], b... [/fig_ref] shows the general design of the pioneering printheads developed by Bellini, and Reddy et al. [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] , which were capable of extruding directly from granulated material, and consisted of miniaturized vertical screw extruders replacing the usual filament-fed deposition tool.
## Experimental small-scale systems
In the case of Bellini, an auger screw (i.e., screw with constant pitch and channel depth) was confined by a heated barrel, whose extremity could accommodate deposition nozzles with varied diameter. Besides, the printhead assembly moved in the three directions thanks to a custom-made desktop positioning system, while the deposition surface was kept fixed [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref]. The system's performance was initially assessed with relation to the processing temperature, nozzle geometry, and deposition velocity using ceramic materials with different granulometry. According to the authors, some level of agglomeration as well as air entrapment was observed within the deposited structures [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref]. The equipment was also used to fabricate scaffolds from biopolymers [bib_ref] Precision extruding deposition and characterization of cellular poly-ε-caprolactone tissue scaffolds, Wang [/bib_ref].
To overcome the feeding problems described by Bellini et al. [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref] , the printhead developed by Reddy et al. [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] included a separated granulate feeding unity and a screw with variable channel depth and pitch (i.e., with compression profile). A support similar to a breaker plate to avoid the deflection of the screw due to uneven radial pressure was also included. The resulting printhead was kept fixed due to its increased weight, while the deposition surface moved in the three directions. Test specimens were fabricated directly from polymer pellets under different conditions.
Another early development in screw-assisted EAM was found in Lam et al. [bib_ref] Mechanical and in vitro evaluations of composite PLDLLA/TCP scaffolds for bone engineering, Lam [/bib_ref] , which developed a printhead to process and deposit biocomposite scaffolds. Multi-material mixing followed by three-dimensional deposition using a single piece of equipment was reported for the first time; however, no details about the screw design and disposition of the components were provided.
The basic design of the printheads developed by Belliniand Reddy et al. [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] was further explored and improved by Silveira et al. [bib_ref] Design development and functional validation of an interchangeable head based on mini..., Silveira [/bib_ref] , Jackson et al. [bib_ref] 3D printing of asphalt and its effect on mechanical properties, Jackson [/bib_ref] , Singamneni et al. [bib_ref] Extrusion 3D printing of polybutyrate-adipate-terephthalate-polymer composites in the pellet form, Singamneni [/bib_ref] [bib_ref] Biopolymer alternatives in pellet form for 3D printing by extrusion. 3D, Singamneni [/bib_ref] , Tseng et al. [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] , Zhou et al. [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] , Leng et al. [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] , Alexandre et al. [bib_ref] Mechanical properties of direct waste printing of polylactic acid with universal pellets..., Alexandre [/bib_ref] , and Wang et al.. Except for the printhead developed by Jackson et al. [bib_ref] 3D printing of asphalt and its effect on mechanical properties, Jackson [/bib_ref] , all equipment adopted a screw with a compression profile. Silveira et al. [bib_ref] Design development and functional validation of an interchangeable head based on mini..., Silveira [/bib_ref] , Jackson et al. [bib_ref] 3D printing of asphalt and its effect on mechanical properties, Jackson [/bib_ref] , and Alexandre et al. [bib_ref] Mechanical properties of direct waste printing of polylactic acid with universal pellets..., Alexandre [/bib_ref] integrated their deposition tools into low-cost desktop 3D printers. In contrast, Tseng et al. [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] , Singamneni et al. [bib_ref] Extrusion 3D printing of polybutyrate-adipate-terephthalate-polymer composites in the pellet form, Singamneni [/bib_ref] , and Wang et al.constructed their own Cartesian positioning platforms.
The smallest screw diameter (7 mm) was found in the equipment developed by Silveira et al. [bib_ref] Design development and functional validation of an interchangeable head based on mini..., Silveira [/bib_ref] , which could reuse waste powders from a Selective Laser Sintering (SLS) machine, and also formulated biocomposites to print tissue engineering scaffolds [bib_ref] Fabrication of PCL/β-TCP scaffolds by 3D mini-screw extrusion printing, Dávila [/bib_ref]. Biocomposites were also explored by Singamneni et al. [bib_ref] Extrusion 3D printing of polybutyrate-adipate-terephthalate-polymer composites in the pellet form, Singamneni [/bib_ref] [bib_ref] Biopolymer alternatives in pellet form for 3D printing by extrusion. 3D, Singamneni [/bib_ref] using their equipment, but in this case, the materials were previously mixed and cut into pellets. Nozzles with increased diameter were used to avoid material clogging. Tseng et al. [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] designed their printhead to process pellets of highly viscous polymers, and added infrared (IR) heaters near the deposition nozzle to prevent delamination and warping when printing. The equipment designed by Jackson et al. [bib_ref] 3D printing of asphalt and its effect on mechanical properties, Jackson [/bib_ref] was very simple and both the auger screw and housing could be fabricated in acrylic resin by a stereolithography process, since the intended processing temperature for the petroleum-based pellets was relatively low. Alexandre et al. [bib_ref] Mechanical properties of direct waste printing of polylactic acid with universal pellets..., Alexandre [/bib_ref] replaced the FFF printhead of a lowcost 3D printer with a do-it-yourself screw-extruder kit, marketed by a recently founded company. The equipment was tested with virgin and recycled polymer pellets, as well as shredded plastic waste. The printhead developed by Wang et al.to process biopolymer pellets has no significant innovations, and the authors did not provide a detailed description of its dimensions. The main novelty is that the authors used a mandrel rotating around the x-axis as the deposition surface. Tests were performed to calibrate the thickness of the extruded fibers in function of the rotation speed of screw and mandrel, as well as the printhead travel speed.
Modifications to the basic design of the screw-assisted printheads are shown in . Instead of an ad hoc extrusion screw, Kumar et al. [bib_ref] The effect of process parameters on tensile behavior of 3D printed flexible..., Kumar [/bib_ref] appropriated from a drill bit attached to the spindle head of a computer numerical control [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] (CNC) milling machine, devising a straightforward solution to the complex problem of screw design and fabrication. Zhou et al. [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] used an auger screw, but included four feeding ports at different heights along the extruder barrel to allow multimaterial feeding and better control of the process residence time. In turn, Leng et al. [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] designed a conical screw to achieve improved plastification and better material homogenization over a shorter length.
Many articles describing the feasibility of the deposition tool developed by Kumar et al. [bib_ref] The effect of process parameters on tensile behavior of 3D printed flexible..., Kumar [/bib_ref] were published. Various materials, including neat polymers and composites in the form of pellets, were processed [bib_ref] Investigation on the effects of process parameters in CNC assisted pellet based..., Kumar [/bib_ref] [bib_ref] Extrusion-based additive manufacturing process for producing flexible parts, Kumar [/bib_ref] [bib_ref] Experimental investigations on suitability of polypropylene (PP) and ethylene vinyl acetate (EVA)..., Kumar [/bib_ref] [bib_ref] 3D printing of flexible parts using eva material, Kumar [/bib_ref] [bib_ref] Investigations on the melt flow behaviour of aluminium filled ABS polymer composite..., Kumar [/bib_ref] [bib_ref] Analysing the influence of raster angle, layer thickness and infill rate on..., Kumar [/bib_ref]. The idea of using a drill bit was also explored by Whyman et al. [bib_ref] Design and development of an extrusion system for 3D printing biopolymer pellets, Whyman [/bib_ref] , which included a separate pellet feeder to prevent the drill from stalling. The printhead was integrated into a custom-built Cartesian positioning platform. Besides biopolymer pellets, polymer blends previously prepared and cut as pellets were also tested [bib_ref] Acrylonitrile butadiene styrene and polypropylene blend with enhanced thermal and mechanical properties..., Harris [/bib_ref]. Boyle et al. [bib_ref] 3D printing using powder melt extrusion, Boyle [/bib_ref] used a drill bit to adapt and construct the open-source Rich Rap Universal Pellet Extruder (RRUPE), which was intended to recycle defective polymers parts. Although 3D printed parts could be produced, Whyman et al. [bib_ref] Design and development of an extrusion system for 3D printing biopolymer pellets, Whyman [/bib_ref] and Boyle et al. [bib_ref] 3D printing using powder melt extrusion, Boyle [/bib_ref] reported material clogging in the drill and inconsistent feeding, respectively, during the calibration experiments. A similar printhead using a drill bit was developed by Kim, Lee [bib_ref] Printability and physical properties of iron slag powder composites using material extrusion-based..., Kim [/bib_ref] , which was then integrated into a low-cost 3D printer. Pellets made of ceramic materials and polymer binders were used to fabricate test specimens and porous "green parts," which is a term used to describe parts that require further processing (e.g., debinding and sintering) to achieve their final properties [bib_ref] A comprehensive review of extrusionbased additive manufacturing processes for rapid production of..., Rane [/bib_ref].
The printhead developed by Zhou et al. [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] was integrated into a 3D Touch® (BitsFromBytes, Bristol, UK) printer and tested with polymer pellets to determine the adequate screw rotation speed with relation to the printing parameters. The auger screw was fabricated employing a powder bed fusion AM process, followed by manual polishing. Using the most inferior feeding port, heat-sensitive polymer granules could be extruded with no significant thermal degradation, and a benchmarking model was 3D printed to assess the dimensional accuracy provided by the equipment. Multi-material printing was also explored, by adding fluorescent particles to the melt.
In the case of Leng et al. [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] , the theory behind the advantages of the conical screw with relation to material plastication was exposed, but the dimensions of the printhead components were not provided. A Cartesian positioning platform was constructed, and test specimens could be fabricated from polymer pellets under different conditions while keeping fixed the screw rotation speed. The capacity to 3D print geometries of different levels of complexity was also explored. shows alternative printhead designs found by the systematic search, which involve more than one mechanism. Annoni et al.appropriated from a small injection molding machine, composed of an inclined plasticating screw and a vertical piston-assisted injection unity. Canessa et al. [bib_ref] Study of Moineau-based pumps for the volumetric extrusion of pellets, Canessa [/bib_ref] added a progressive cavity pump (a.k.a. Moineau pump) to the extremity of an auger screw, to achieve improved control of material flow. Liu et al. [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] integrated an auxiliary liquefying chamber to their screw-assisted printhead, which was capable of pumping already molten material to the screw mechanism. Finally, Khondoker, Sameoto [bib_ref] Direct coupling of fixed screw extruders using flexible heated hoses for FDM..., Khondoker [/bib_ref] developed a two-stage system, which consisted of a fixed horizontal plasticating unity that was connected to a typical FFF deposition head.
Annoni et al.explored direct EAM of metals and ceramics from injection molding feedstocks. The pellets had expressive contents of solid particles and, therefore, required a deposition tool capable of generating high levels of Schematic illustrations of the most significant modifications to the basic design of vertical screw-assisted printheads proposed by a Kumar et al. [bib_ref] The effect of process parameters on tensile behavior of 3D printed flexible..., Kumar [/bib_ref] , b Zhou et al. [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] , and c Leng et al. [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] extrusion pressure. Due to the weight of the injection molding unity, the equipment was fixed to a custom-made delta positioning system. The deposition surface was capable of move in 3 directions, as well as rotating around 2 axes. Although extrudability tests were performed to investigate the influence of the binder percentage and geometry of the nozzle, no actual 3D part was built.
Canessa et al. [bib_ref] Study of Moineau-based pumps for the volumetric extrusion of pellets, Canessa [/bib_ref] did not provide much information about the design of the auger screw, since their focus was on the design of the Moineau pump to achieve better volumetric control of the extrusion flow rate. The concept would circumvent the need for retraction strategies, allowing intermittent deposition even with continuous material feeding. The printhead was integrated into a RepRap 3D printer and was validated with different materials.
The printhead developed by Liu et al. [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] was a multitechnological deposition tool that presented two pistondriven dispensing unities, a screw-assisted extruder, and a plasma jetting unity. No information about the screw design was provided. Each deposition unity was indexed by rotational movement, and the head assembly was integrated into a custom-made Cartesian positioning system [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref]. The equipment was used to fabricate scaffolds from neat polymer pellets [bib_ref] A plasma-assisted bioextrusion system for tissue engineering, Liu [/bib_ref] [bib_ref] Process-driven microstructure control in melt-extrusion-based 3D printing for tailorable mechanical properties in..., Liu [/bib_ref] [bib_ref] Structural evolution of PCL during melt extrusion 3D printing, Liu [/bib_ref] [bib_ref] User interface tool for a novel plasma-assisted bio-additive extrusion system, Liu [/bib_ref] and biocomposites [bib_ref] A plasma-assisted bioextrusion system for tissue engineering, Liu [/bib_ref] [bib_ref] User interface tool for a novel plasma-assisted bio-additive extrusion system, Liu [/bib_ref] , which were previously prepared by manual mixing processes.
In contrast to the previous screw-assisted printheads that integrate into the same assembly the processing and deposition unities, the alternative design proposed by Khondoker, Sameoto [bib_ref] Direct coupling of fixed screw extruders using flexible heated hoses for FDM..., Khondoker [/bib_ref] decouples the inherent limitations imposed by the weight and inertia of the screw extruder from the speed and resolution requirements of the deposition head. The filament extruder consisted of a DIY kit (Filastruder®, Snellville, USA), which used a drill bit as plasticating screw. The deposition head was connected to the pellet extruder using a heated hose and was adapted to a low-cost 3D printer. Printing feasibility was demonstrated with pellets of elastomeric materials. A similar system was described by Liu et al. [bib_ref] High-pressure interfacial impregnation by micro-screw in-situ extrusion for 3D printed continuous carbon..., Liu [/bib_ref] , which also connected a horizontal pellet extruder to a filament-fed deposition head. However, in this case, the extruder was custom-made and used a screw with a compression profile.
An impregnation mold was coupled to the die of the extruder, to generate a pre-impregnated filament that could be fed to the deposition head. The printhead was mounted on a custommade Cartesian platform and could move on the vertical direction, while the deposition surface moved on the x-y plane. The equipment allowed 3D printing with continuous fiberreinforced composites.
## Experimental large-scale systems
Most experimental screw-assisted printheads that were specifically developed (or could be applied) for large-scale 3D printing share the same general configuration of the vertical extruders presented in the previous subsection. Some systems appropriated from commercially available extruders [bib_ref] Additive manufacturing of poly(propylene) by means of melt extrusion, Hertle [/bib_ref] [bib_ref] Production of polymer-metal hybrids using extrusion-based additive manufacturing and electrochemically treated aluminum, Hertle [/bib_ref] [bib_ref] Robotic AM System for Plastic Materials: Tuning and On-line Adjustment of Process..., Magnoni [/bib_ref] , while others presented custom-made deposition heads with auger screw [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] , or typical extrusion screw with a compression profile [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] [bib_ref] Characterization of a granulate-based strand deposition process in the FLM-method for definition..., Schmidt [/bib_ref]. The use of robotic arms in experimental systems [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] [bib_ref] Robotic AM System for Plastic Materials: Tuning and On-line Adjustment of Process..., Magnoni [/bib_ref] [bib_ref] Characterization of a granulate-based strand deposition process in the FLM-method for definition..., Schmidt [/bib_ref] was more frequent than Cartesian platforms [bib_ref] Additive manufacturing of poly(propylene) by means of melt extrusion, Hertle [/bib_ref] [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref].
Hertle et al. [bib_ref] Additive manufacturing of poly(propylene) by means of melt extrusion, Hertle [/bib_ref] adapted a screw-driven welding extruder to a gantry structure, with the deposition surface moving in two directions. The equipment could be fed with polymer pellets and was initially used to assess material adhesion in shear test specimens. In a later publication, a similar welding extruder was mounted on a six-axes robot arm, and material deposition was performed onto electrochemically treated aluminum sheets to assess once again the shear strength of test specimens [bib_ref] Production of polymer-metal hybrids using extrusion-based additive manufacturing and electrochemically treated aluminum, Hertle [/bib_ref].
Magnoni et al. [bib_ref] Robotic AM System for Plastic Materials: Tuning and On-line Adjustment of Process..., Magnoni [/bib_ref] attached a benchtop-sized screw extruder to a robotic arm that could move on six axes to perform large-scale 3D printing directly from polymer pellets. The focus of their work was determining the influence of the process parameters on the resulting height and width of the deposited material. An online control routine was implemented to correct the positioning of the robotic arm based on data acquired during material deposition. In a subsequent work, an online re-slicing algorithm was developed to compensate for the variation of stature during 3D printing, preserving the original part size [bib_ref] Process parameters tuning and online re-slicing for robotized additive manufacturing of big..., Rebaioli [/bib_ref]. Brooks et al. [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] have also described a Schematic illustrations of the alternative small-scale screw-assisted printhead designs developed by a Annoni et al., b Canessa et al. [bib_ref] Study of Moineau-based pumps for the volumetric extrusion of pellets, Canessa [/bib_ref] , c Liu et al. [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] , and d Khondoker, Sameoto [bib_ref] Direct coupling of fixed screw extruders using flexible heated hoses for FDM..., Khondoker [/bib_ref] screw-assisted deposition tool that could be used for largescale 3D printing, but it was explored in the manufacture of thin-shelled parts without the need for supports. Instead of moving the printhead, a convex deposition surface was attached to the robotic arm so that it could be moved with six degrees-of-freedom. Pellets of polymer filled with vegetal fiber were processed. The focus of their paper was on the algorithm developed to generate the printing trajectory of 2D geometries projected to the deposition surface. Similarly, the AM system described by Schmidt et al. [bib_ref] Characterization of a granulate-based strand deposition process in the FLM-method for definition..., Schmidt [/bib_ref] had a fixed printhead, with the deposition surface attached to a robotic arm that could move on six axes. The custom-made printhead included a screw with a compression profile. Calibration experiments were performed in function of the material type and, since the material flow rate was known under different conditions, 3D printing tests were conducted at a range of speeds. Test specimens were cut from single-walled cylinders built under a continuous extrusion mode.
In contrast to the aforementioned examples, presents the alternative solution for a screw-assisted large-scale 3D printer, proposed by Liu et al. [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref]. The extrusion head was composed of two stages: in the first stage, a screw with a compression profile was responsible for conveying and melting the polymer pellets, while an auger screw controlled the extrusion rate in the second stage. Both screws could be driven independently, and the vertical screw could rotate backward to retract the material during the 3D printing process.
The extruder assembly was integrated into a milling machine and could be moved on the y-z plane while the deposition surface moved on the x-axis. Tests were performed with composite pellets to understand the influence of the pressure generated in the first stage and the rotation speed of the auger screw on the resulting flow rate. Specimens were cut from 3D printed plates. The capacity to print more complex parts was also demonstrated; however, melt flow instabilities were reported [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref].
## Commercial small-scale systems
As shown in , four commercial small-scale EAM systems with screw-assisted printheads were found in the systematic search. The 3D Discovery® machine, marketed by RegenHU (Villaz-St-Pierre, Switzerland), is a multitechnological biofabrication platform that incorporates up to six printheads, including a piston unity, and a screw-assisted extruder head. The Pellet Additive Manufacturing® (PAM) system is marketed by Pollen AM (Ivry-sur-Seine, France) and presents four pellet-fed mini screw extruders. The M3DIMAKER® pharmaceutical 3D printer, marketed by FabRx (London, UK), can print from filaments or powders, in which case a miniaturized screw extruder is used. The ExAM 255® 3D Printer, marketed by AIM3D GmbH (Rostock, Germany), presents two mini screw extruders that can print from highly filled pellets.
The printheads from the 3D Discovery® are mounted on a Cartesian positioning system and move on the x-z plane while the deposition surface moves on the y-axis. The screwassisted unity is very similar to the experimental equipment described by Liu et al. [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] , in which the pellets are first melted in a heated chamber and then air-pumped to the rotating screw. The 3D Discovery® was used to fabricate scaffolds from different biocomposites [bib_ref] Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration, Wang [/bib_ref] [bib_ref] Enhancing the hydrophilicity and cell attachment of 3D printed PCL/graphene scaffolds for..., Wang [/bib_ref] [bib_ref] Polymer-ceramic composite scaffolds: the effect of hydroxyapatite and β-tri-calcium phosphate, Huang [/bib_ref] [bib_ref] Aligned multi-walled carbon nanotubes with nanohydroxyapatite in a 3D printed polycaprolactone scaffold..., Huang [/bib_ref] [bib_ref] 3D Printing of polycaprolactone-polyaniline electroactive scaffolds for bone tissue engineering, Wibowo [/bib_ref]. In all cases, the biocomposites were previously formulated by manual melt mixing and then cut in pellets. According to the authors, the screw-assisted processing was essential to further distribute the solid particles in the molten polymer [bib_ref] Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration, Wang [/bib_ref] [bib_ref] Enhancing the hydrophilicity and cell attachment of 3D printed PCL/graphene scaffolds for..., Wang [/bib_ref] [bib_ref] Polymer-ceramic composite scaffolds: the effect of hydroxyapatite and β-tri-calcium phosphate, Huang [/bib_ref] [bib_ref] Aligned multi-walled carbon nanotubes with nanohydroxyapatite in a 3D printed polycaprolactone scaffold..., Huang [/bib_ref] [bib_ref] 3D Printing of polycaprolactone-polyaniline electroactive scaffolds for bone tissue engineering, Wibowo [/bib_ref].
The PAM® Series P machine has the extruders fixed to the structure, while the deposition surface is mounted on a delta positioning system. Geoffroy et al. [bib_ref] Additive manufacturing of fire-retardant ethylene-vinyl acetate, Geoffroy [/bib_ref] used the PAM® machine to directly print pellets from neat and flame-retarded polymers. Polymer compounding and pelletizing were previously carried with the aid of a twin-screw extruder. The dispersion degree of the additives was characterized by x-ray mapping, and the flame-retardancy performance achieved with the 3D printing process was compared to the results obtained from thermocompression. The screw-assisted 3D printing process was found to influence the resulting size of the additive particles, which affected the fire behavior.
The M3DIMAKER® presents a Cartesian positioning system and the printhead moves in three directions, while the deposition surface is fixed. Goyanes et al. [bib_ref] Direct powder extrusion 3D printing: fabrication of drug products using a novel..., Goyanes [/bib_ref] used the M3DIMAKER® machine to produce printlets (3D printed tablets) from drug-loaded polymer powders. The extrusion temperature and other process parameters were kept fixed, to evaluate the performance of the printlets produced with different grades of the polymer matrix. The equipment allowed to fabricate medicines in a single-step process from small amounts of material, which can prevent thermal degradation of the drugs, expedite pre-clinical tests, and enable the production of dose-personalized medicines using a much broader range of excipients.
Similarly, the Exam255® uses a Cartesian positioning system with the printheads moving in the x-y plane, while the deposition surface moves on the z-axis. However, the system is encapsulated. Lengauer et al. [bib_ref] Fabrication and properties of extrusion-based 3D-printed hardmetal and cermet components, Lengauer [/bib_ref] used the ExAM 255® to Schematic illustration of the alternative design for large-scale screw-assisted 3D printing developed by Liu et al. [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] directly 3D print hardmetal combined with polymeric binders used in the metal injection molding industry. After 3D printing, the green parts were subject to thermal debinding and sintering. The printing quality achieved with the screwassisted system was compared to the results from a low-cost desktop FFF 3D printer, using self-fabricated hardmetal filaments. As discussed by the authors, the smaller nozzle diameter generated higher extrusion pressure that was beneficial to the quality of the printed parts.
## Commercial large-scale systems
Three commercially available screw-assisted EAM systems for large-scale applications were found in the systematic search and are shown in [fig_ref] Figure 1: Flowchart of the systematic search, showing the resulting number of articles after... [/fig_ref] : the BAAM® (Big Area Additive Manufacturing) 3D printer, marketed by Cincinnati Inc. , the Gigabot X® 3D printer, marketed by re:3D (Houston, USA), and the Super Discovery 3D Printer®, market by CNC Barcenas (Valdepeñas, Spain). Benchtop-sized screw extruders that are mounted on gantry structures compose all three systems. However, the printhead from the BAAM® system moves on the x-axis, and the deposition surface moves on the y-z plane, while the printheads from the Super Discovery 3D Printer® and Gigabot X® move on the x-y plane and the deposition surface on the z-axis.
Duty et al. [bib_ref] Structure and mechanical behavior of Big Area Additive Manufacturing (BAAM) materials, Duty [/bib_ref] reported the development of the BAAM® technology. The deposition head originally appropriated from the screw of a welding extruder, which was later replaced by a longer screw version to avoid intra-bead porosity and increase the material output. Besides, a reciprocating z-tamping attachment was added to the printhead to further reduce porosity and improve interlayer adhesion. Process feasibility was first demonstrated using neat polymer pellets but, due to the expressive distortion and warping, fiber-reinforced materials became preferred. The BAAM® machine was explored in several publications, using diverse materials, including composites based on high-temperature thermoplastics [bib_ref] Rheological survey of carbon fiberreinforced high-temperature thermoplastics for big area additive manufacturing..., Ajinjeru [/bib_ref] [bib_ref] Thermomechanical characterization of short carbon fiber and short glass fiber-reinforced ABS used..., Billah [/bib_ref] [bib_ref] Compressive deformation analysis of large area pellet-fed material extrusion 3D printed parts..., Trejo [/bib_ref] [bib_ref] Mechanical characterization of high-temperature carbon fiber-polyphenylene sulfide composites for large area extrusion..., Yeole [/bib_ref]. Due to the significant difference between the large-and small-scale EAM, the BAAM® process was investigated with the aid of thermomechanical modeling [bib_ref] Thermal analysis of additive manufacturing of large-scale thermoplastic polymer composites, Compton [/bib_ref] , leading to the development of specific design guidelines for large-scale 3D printing by Roschli et al. [bib_ref] Designing for Big Area Additive Manufacturing, Roschli [/bib_ref].
Woern et al. [bib_ref] Fused particle fabrication 3-D printing: recycled materials' optimization and mechanical properties, Woern [/bib_ref] first tested the Gigabot X® 3D printer with a variety of particulate materials, including virgin pellets, and recycled polymers. An experimental matrix was proposed to determine the optimum printing parameters for each material, in function of the temperature and deposition speed. The The four commercial small-scale 3D printer with screw-assisted printheads found in the systematic search: a 3D Discovery®, b PAM® Series P, c M3DIMAKER® [bib_ref] FabRx's pharmaceutical 3D printer for personalised medicines, Fabrx [/bib_ref] , and d Exam255®Fig [fig_ref] Figure 1: Flowchart of the systematic search, showing the resulting number of articles after... [/fig_ref] The large-scale screw-assisted systems found in the systematic search: a the BAAM® machine, b the Gigabot X® machine (adapted from [bib_ref] Towards distributed recycling with additive manufacturing of PET flake feedstocks, Little [/bib_ref] , and c the Super Discovery® 3D printeradequate deposition speed was mostly dependent on the shape of the material particles (e.g., pellets, shreds, or flakes). Tensile test specimens were fabricated, and the properties from the recycled materials were found to be comparable to the virgin polymers. Later publications described recycling tests with other materials [bib_ref] Mechanical properties and applications of recycled polycarbonate particle material extrusion-based additive manufacturing, Reich [/bib_ref] [bib_ref] Technical pathways for distributed recycling of polymer composites for distributed manufacturing: windshield..., Dertinger [/bib_ref] [bib_ref] Towards distributed recycling with additive manufacturing of PET flake feedstocks, Little [/bib_ref] , obtained from ground plastic parts. Besides calibration experiments and the fabrication of test specimens, complex parts were 3D printed to demonstrate technical and economic feasibility [bib_ref] Mechanical properties and applications of recycled polycarbonate particle material extrusion-based additive manufacturing, Reich [/bib_ref] [bib_ref] Towards distributed recycling with additive manufacturing of PET flake feedstocks, Little [/bib_ref]. However, the 3D printing tests with composite particles made of hard and flexible polymers performed by Dertinger et al. [bib_ref] Technical pathways for distributed recycling of polymer composites for distributed manufacturing: windshield..., Dertinger [/bib_ref] were unsuccessful due to feeding difficulties. Later, Little et al. [bib_ref] Towards distributed recycling with additive manufacturing of PET flake feedstocks, Little [/bib_ref] added a Crammer feeder to avoid melt flow inconsistencies due to the irregularity of shape and size of the ground particles.
The Super Discovery 3D Printer® was tested by Moreno Nieto [bib_ref] Large-format polymeric pellet-based additive manufacturing for the naval industry, Nieto [/bib_ref] , to demonstrate its application feasibility to the naval industry. The printhead was equipped with an articulated arm, where different tools such as a video surveillance device could be fitted to remotely supervise the 3D printing process. Pellets of different polymers were processed, in some cases with manual addition of reinforcing fibers. Expressive thermal distortion was observed with some of the neat polymers, while poor printability with the fiber-reinforced composites was reported. Although the negative results with the composites might indicate that the Super Discovery 3D Printer® is somewhat less developed than the BAAM® machine, it should be noted that in this study the fibers were manually mixed with the polymer pellets, and not properly compounded, which could have caused nozzle clogging due to non-uniform distribution.
## Extruder modifications
The publications analyzed in the previous section have shown a significant design variety for the screw-assisted EAM systems. While many articles in which small-scale equipment were described intended mainly to demonstrate the process feasibility, others have focused on the properties of the 3D printed parts. In either case, adjusting the design of the printhead with relation to specific material properties or processing requirements was not on the scope. Anyway, the referred articles described experiments with different materials, such as ceramics (PZT/ECG8 [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref] , Fe 2 O 3 +Al 2 O 3 /binders, and Fe 2 O 3 +ZrO 2 +Y 2 O 3 /binders [bib_ref] Printability and physical properties of iron slag powder composites using material extrusion-based..., Kim [/bib_ref] , neat and reinforced thermoplastics (polycaprolactone (PCL) [bib_ref] Precision extruding deposition and characterization of cellular poly-ε-caprolactone tissue scaffolds, Wang [/bib_ref] , poly(L-lactideco-D,L-lactide)/tricalcium phosphate (PLDLLA/TCP), ethylene vinyl acetate (EVA) , acrylonitrile butadiene styrene (ABS) [bib_ref] Experimental investigations on suitability of polypropylene (PP) and ethylene vinyl acetate (EVA)..., Kumar [/bib_ref] [bib_ref] Investigations on the melt flow behaviour of aluminium filled ABS polymer composite..., Kumar [/bib_ref] , polypropylene (PP) [bib_ref] Experimental investigations on suitability of polypropylene (PP) and ethylene vinyl acetate (EVA)..., Kumar [/bib_ref] , ABS/ aluminum [bib_ref] Investigations on the melt flow behaviour of aluminium filled ABS polymer composite..., Kumar [/bib_ref] , bitumen [bib_ref] 3D printing of asphalt and its effect on mechanical properties, Jackson [/bib_ref] , hardmetals (WC-Co [bib_ref] Fabrication and properties of extrusion-based 3D-printed hardmetal and cermet components, Lengauer [/bib_ref] , and medicines (hydroxypropylcellulose/itraconazole [bib_ref] Direct powder extrusion 3D printing: fabrication of drug products using a novel..., Goyanes [/bib_ref].
With respect to the large-scale systems, the same trend could be observed. Even though extruder scaling-up was necessary, most publications described systems that appropriated from commercial welding [bib_ref] Additive manufacturing of poly(propylene) by means of melt extrusion, Hertle [/bib_ref] [bib_ref] Production of polymer-metal hybrids using extrusion-based additive manufacturing and electrochemically treated aluminum, Hertle [/bib_ref] and laboratory extruders [bib_ref] Robotic AM System for Plastic Materials: Tuning and On-line Adjustment of Process..., Magnoni [/bib_ref] [bib_ref] Process parameters tuning and online re-slicing for robotized additive manufacturing of big..., Rebaioli [/bib_ref] to demonstrate the feasibility of the equipment. Neat pellets of PP [bib_ref] Additive manufacturing of poly(propylene) by means of melt extrusion, Hertle [/bib_ref] [bib_ref] Production of polymer-metal hybrids using extrusion-based additive manufacturing and electrochemically treated aluminum, Hertle [/bib_ref] , polylactic acid (PLA) [bib_ref] Robotic AM System for Plastic Materials: Tuning and On-line Adjustment of Process..., Magnoni [/bib_ref] [bib_ref] Process parameters tuning and online re-slicing for robotized additive manufacturing of big..., Rebaioli [/bib_ref] , and ABS [bib_ref] Process parameters tuning and online re-slicing for robotized additive manufacturing of big..., Rebaioli [/bib_ref] were processed using the extruders as acquired. Moreno Nieto et al. [bib_ref] Large-format polymeric pellet-based additive manufacturing for the naval industry, Nieto [/bib_ref] have also processed neat ABS, PLA, and polyethylene terephthalate glycol (PETG)-modified, as well as flame-retarded ABS, carbon fiber (CF)-, and glass fiber (GF)-reinforced ABS without alterations on the machine.
Few publications described modifications on the screwassisted printheads. Some were motivated by specific material properties and others could be related to improvements on the performance of the equipment, independent from the materials processed. These will be described in the following subsections.
# Material-related modifications
Most material-related modifications on the printhead addressed the state of the granules (i.e., average size and size distribution of the particles). To avoid premature melting due to the small size of the particles when processing neat ABS, high-impact polystyrene (HIPS), PLA, and PP in the powder form, Boyle at al. [bib_ref] 3D printing using powder melt extrusion, Boyle [/bib_ref] modified the geometry of the hopper and improved the thermal insulation of the extruder. Little et al. [bib_ref] Towards distributed recycling with additive manufacturing of PET flake feedstocks, Little [/bib_ref] adjusted the geometry of the hopper and used a separate feeder to circumvent feeding problems due to the irregular size of the particles when recycling shredded polyethylene terephthalate (PET) bottles. Reich et al. [bib_ref] Mechanical properties and applications of recycled polycarbonate particle material extrusion-based additive manufacturing, Reich [/bib_ref] used additional cooling fans near to the hopper to avoid premature melting and also improved the thermal insulation of the extruder in order to recycle ground polycarbonate (PC) parts, which required a relatively high temperature to be extruded.
A very specific example of system modification related to the state of the feedstock refers to the equipment developed by Liu et al. [bib_ref] High-pressure interfacial impregnation by micro-screw in-situ extrusion for 3D printed continuous carbon..., Liu [/bib_ref] , which intended to impregnate continuous carbon fibers with molten polyamide (PA12). Besides using a screw with compression profile, an impregnation mold was especially designed so that the fiber bundles could be continuously impregnated with the polymer. The resulting preextruded composite filament was then pulled into the actual deposition head.
The feedstock composition led Annoni et al.to use an injection machine as printhead, in order to demonstrate the EAM feasibility of highly filled materials, such as ZrO 2 + Y 2 O 3 , and AISI 630 steel combined with polymer binders. The equipment presented a screw-extruder followed by a piston mechanism, which was important due to the high viscosity of the ceramic and metal injection molding materials. Besides, the internal geometry of the deposition nozzle (i.e., orifice diameter, length of the land) was also adjusted to achieve stable extrusion flow. Singamneni et al. [bib_ref] Extrusion 3D printing of polybutyrate-adipate-terephthalate-polymer composites in the pellet form, Singamneni [/bib_ref] also had to adjust the diameter of the nozzle when processing composites of polybutylene-adipate-terephthalate (PBAT) reinforced with wood flour, to avoid clogging caused by the vegetal fibers.
Two publications described modifications related to the thermal properties of the materials. Zhou et al. [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] designed a barrel with multiple feeding ports and used the lowest one to avoid thermomechanical degradation polyvinyl alcohol (PVOH). Besides, an auger screw was fabricated to subject the polymer to minimal shearing and pressure levels. Tseng et al. [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] completely adjusted the design of their printhead to process pellets of polyether-ether-ketone (PEEK). Three independent heaters were used to achieve the required temperature profile in the extruder, and a screw with the L/D ratio and compression profile prescribed by the material supplier was fabricated. Moreover, the screw flights were designed and checked to ensure the safety of the equipment with regard to the high melt viscosity of PEEK.
# Performance-related modifications
The adjustments on the screw geometry of the experimental systems addressed performance requirements, and not specific material properties. Reddy et al. [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] designed a screw with compression profile to eliminate the air entrapped between the ABS pellets, and used a separate feeder to avoid material aggregation. The printhead described by Silveira et al. [bib_ref] Design development and functional validation of an interchangeable head based on mini..., Silveira [/bib_ref] had a screw with reduced diameter and compression profile, making it capable to process small amounts of material and mix different components [bib_ref] Fabrication of PCL/β-TCP scaffolds by 3D mini-screw extrusion printing, Dávila [/bib_ref]. A scraper was used to avoid powder bridging at the hopper. Despite the different properties, PA12 [bib_ref] Design development and functional validation of an interchangeable head based on mini..., Silveira [/bib_ref] and PCL reinforced with tricalcium phosphate [bib_ref] Fabrication of PCL/β-TCP scaffolds by 3D mini-screw extrusion printing, Dávila [/bib_ref] were successfully processed.
The extrusion screw from the system developed by Schmidt et al. [bib_ref] Characterization of a granulate-based strand deposition process in the FLM-method for definition..., Schmidt [/bib_ref] followed the conventional design guidelines from polymer extrusion theory to achieve a certain mass output, avoid material agglomeration at the feeding zone, and thermal degradation of the melt. Pellets of ABS and PLA were processed. In turn, Leng et al. [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] designed a conical screw to achieve higher pressure and improved plastication, but in a shorter length. Thermoplastic polyurethane (TPU) was processed because it is biocompatible and difficult to print in FFF systems.
Duty et al. [bib_ref] Structure and mechanical behavior of Big Area Additive Manufacturing (BAAM) materials, Duty [/bib_ref] reported the substitution of the original screw from the BAAM® machine for another with larger pitch at the feed zone and higher compression ratio. The screw was replaced to allow higher throughput with ABS/CF composites. No other modification was described in the following publications, in which high-temperature materials such as polyphenylsulfone (PPSU), polyethersulfone (PES), and polyphenylenesulfide (PPS) reinforced with carbon fibers were processed [bib_ref] Rheological survey of carbon fiberreinforced high-temperature thermoplastics for big area additive manufacturing..., Ajinjeru [/bib_ref].
Instead of modifying the screw geometry, Brooks et al. [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] prevented premature melting and agglomeration of pellets of biopolyesters reinforced with New Zealand flax fibers by positioning a heatsink with fan near to the hopper. Similarly, Whyman et al. [bib_ref] Design and development of an extrusion system for 3D printing biopolymer pellets, Whyman [/bib_ref] used a separate feeder, so that the printhead is operated under starve-fed conditions, preventing the drill bit from stalling when processing PLA pellets. The low shearing imposed by the drill bit was beneficial when processing pellets of ABS/PP blend, to avoid morphology alterations [bib_ref] Acrylonitrile butadiene styrene and polypropylene blend with enhanced thermal and mechanical properties..., Harris [/bib_ref].
The bio-AM system developed by Liu et al. [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] included a pre-melting chamber, which could be fed with polymer pellets and delivered molten material directly to the auger screw. Although the printhead design was not explained, this probably avoided the auger from stalling due to its low plasticating capacity and torque limitations. Besides, the whole screw length was available to convey and eventually mix different materials. Neat PCL pellets [bib_ref] A plasma-assisted bioextrusion system for tissue engineering, Liu [/bib_ref] [bib_ref] Process-driven microstructure control in melt-extrusion-based 3D printing for tailorable mechanical properties in..., Liu [/bib_ref] [bib_ref] Structural evolution of PCL during melt extrusion 3D printing, Liu [/bib_ref] and composites based on PCL and carbon nanotubes (CNT) were processed [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] [bib_ref] User interface tool for a novel plasma-assisted bio-additive extrusion system, Liu [/bib_ref].
Some printhead modifications were performed to improve the control over the flow of the extruded material. While Canessa et al. [bib_ref] Study of Moineau-based pumps for the volumetric extrusion of pellets, Canessa [/bib_ref] coupled a progressive cavity pump to the extremity of an auger, Liu et al. [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] used a second independent metering auger. The first authors demonstrated the feasibility of their concept with different materials, such as chocolate, wax, and PLA, while the latter processed only ABS reinforced with glass fiber.
Lastly, Khondoker, Sameoto [bib_ref] Direct coupling of fixed screw extruders using flexible heated hoses for FDM..., Khondoker [/bib_ref] observed that most screw-assisted printheads resulted in bulky pieces of equipment, which were difficult to move with precision at the satisfactory speed in order to perform three-dimensional deposition. By using a heated rose between a miniature filament extruder and a typical FFF printhead, the authors were able to develop a low-cost system to successfully 3D print thermoplastic elastomers, including pellets of styrene-ethylenebutylene-styrene (SEBS) and of a shape memory polymer.
## General development workflow
The research presented in each publication from the repository allowed to elaborate a general development workflow that summarizes the common stages that the authors went through when investigating on the topic of screw-assisted EAM. The workflow is presented in [fig_ref] Figure 1: Flowchart of the systematic search, showing the resulting number of articles after... [/fig_ref] , highlighting the development stages and the main decision aspects.
Many publications describing experimental systems showed that the development workflow starts by the definition of the application field and system scale [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref] [bib_ref] Design development and functional validation of an interchangeable head based on mini..., Silveira [/bib_ref] [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] [bib_ref] 3D printing of asphalt and its effect on mechanical properties, Jackson [/bib_ref] [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] [bib_ref] Direct coupling of fixed screw extruders using flexible heated hoses for FDM..., Khondoker [/bib_ref] [bib_ref] High-pressure interfacial impregnation by micro-screw in-situ extrusion for 3D printed continuous carbon..., Liu [/bib_ref] [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref]. Also, it was important to investigate the properties and characteristics of the feedstock materials (e.g., viscosity [bib_ref] Experimental investigations on suitability of polypropylene (PP) and ethylene vinyl acetate (EVA)..., Kumar [/bib_ref] [bib_ref] Investigations on the melt flow behaviour of aluminium filled ABS polymer composite..., Kumar [/bib_ref] [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] [bib_ref] Production of polymer-metal hybrids using extrusion-based additive manufacturing and electrochemically treated aluminum, Hertle [/bib_ref] [bib_ref] Rheological survey of carbon fiberreinforced high-temperature thermoplastics for big area additive manufacturing..., Ajinjeru [/bib_ref] , thermal behavior [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] [bib_ref] Direct powder extrusion 3D printing: fabrication of drug products using a novel..., Goyanes [/bib_ref] [bib_ref] Thermomechanical characterization of short carbon fiber and short glass fiber-reinforced ABS used..., Billah [/bib_ref] [bib_ref] Technical pathways for distributed recycling of polymer composites for distributed manufacturing: windshield..., Dertinger [/bib_ref] , or particle size distribution [bib_ref] Mechanical properties of direct waste printing of polylactic acid with universal pellets..., Alexandre [/bib_ref] [bib_ref] Fused particle fabrication 3-D printing: recycled materials' optimization and mechanical properties, Woern [/bib_ref] [bib_ref] Mechanical properties and applications of recycled polycarbonate particle material extrusion-based additive manufacturing, Reich [/bib_ref] [bib_ref] Technical pathways for distributed recycling of polymer composites for distributed manufacturing: windshield..., Dertinger [/bib_ref] [bib_ref] Towards distributed recycling with additive manufacturing of PET flake feedstocks, Little [/bib_ref] , since the material properties can direct and impose limitations on the development of the equipment.
Next, when designing the printhead, the authors had to choose which driving elements would be used, define the geometry of the main components of the extruder, and decide about any peripherals. Stepper motors were often used , probably due the good relation between the torque provided and dimensions, low cost, and for its easiness of control. Although it can be seen as the core of the printhead, screw design was strictly considered by few authors [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] [bib_ref] Design development and functional validation of an interchangeable head based on mini..., Silveira [/bib_ref] [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] [bib_ref] Characterization of a granulate-based strand deposition process in the FLM-method for definition..., Schmidt [/bib_ref] mostly to meet specific performance requirements. Peripherals such as independent feeding systems [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] [bib_ref] Design and development of an extrusion system for 3D printing biopolymer pellets, Whyman [/bib_ref] [bib_ref] Technical pathways for distributed recycling of polymer composites for distributed manufacturing: windshield..., Dertinger [/bib_ref] , pre-melting chamber [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] [bib_ref] Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration, Wang [/bib_ref] , additional cooling elements [bib_ref] Acrylonitrile butadiene styrene and polypropylene blend with enhanced thermal and mechanical properties..., Harris [/bib_ref] [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] [bib_ref] Mechanical properties and applications of recycled polycarbonate particle material extrusion-based additive manufacturing, Reich [/bib_ref] , additional heaters [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] , and other mechanisms to improve the state of the deposited layer [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] [bib_ref] Additive manufacturing of poly(propylene) by means of melt extrusion, Hertle [/bib_ref] [bib_ref] Production of polymer-metal hybrids using extrusion-based additive manufacturing and electrochemically treated aluminum, Hertle [/bib_ref] [bib_ref] Structure and mechanical behavior of Big Area Additive Manufacturing (BAAM) materials, Duty [/bib_ref] have also been proved important to achieve consistent results.
The printhead should be integrated into an adequate positioning system to comply with its planned application, which goes through defining the number of degrees-of-freedom, and structure type. Some authors used robotic arms with multiple degrees of freedom or CNC systems [bib_ref] The effect of process parameters on tensile behavior of 3D printed flexible..., Kumar [/bib_ref] [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref]. When the design of the printhead limited its movements, the positioning system enabled 3D printing through a moveable deposition surface [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] [bib_ref] Additive manufacturing of fire-retardant ethylene-vinyl acetate, Geoffroy [/bib_ref] [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] [bib_ref] Characterization of a granulate-based strand deposition process in the FLM-method for definition..., Schmidt [/bib_ref]. 3D printing platforms were adapted from pre-existing systems [bib_ref] Fabrication of PCL/β-TCP scaffolds by 3D mini-screw extrusion printing, Dávila [/bib_ref] [bib_ref] Study of Moineau-based pumps for the volumetric extrusion of pellets, Canessa [/bib_ref] [bib_ref] 3D printing of asphalt and its effect on mechanical properties, Jackson [/bib_ref] or were custom-made to meet specific requirements [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref] [bib_ref] Design, fabrication and initial evaluation of a novel hybrid system for tissue..., Liu [/bib_ref] [bib_ref] Biopolymer alternatives in pellet form for 3D printing by extrusion. 3D, Singamneni [/bib_ref] [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] [bib_ref] Thermal analysis of additive manufacturing of large-scale thermoplastic polymer composites, Compton [/bib_ref].
Screw-assisted EAM was performed by continuous [bib_ref] Fabrication of PCL/β-TCP scaffolds by 3D mini-screw extrusion printing, Dávila [/bib_ref] [bib_ref] User interface tool for a novel plasma-assisted bio-additive extrusion system, Liu [/bib_ref] [bib_ref] The effect of process parameters on tensile behavior of 3D printed flexible..., Kumar [/bib_ref] [bib_ref] Biopolymer alternatives in pellet form for 3D printing by extrusion. 3D, Singamneni [/bib_ref] [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] or intermittent deposition [bib_ref] Study of Moineau-based pumps for the volumetric extrusion of pellets, Canessa [/bib_ref] [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] [bib_ref] Large-format polymeric pellet-based additive manufacturing for the naval industry, Nieto [/bib_ref]. The problem of unwanted extrusion, referred to as "over deposition", "bleeding" [bib_ref] 3D printing of asphalt and its effect on mechanical properties, Jackson [/bib_ref] , "leakage", or "salivation" [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] , was reported and could be circumvented with a continuous deposition approach, control of the volumetric flow [bib_ref] Study of Moineau-based pumps for the volumetric extrusion of pellets, Canessa [/bib_ref] , or material retraction [bib_ref] Robot-assisted 3D printing of biopolymer thin shells, Brooks [/bib_ref] [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] [bib_ref] Large-format polymeric pellet-based additive manufacturing for the naval industry, Nieto [/bib_ref]. A software is necessary to control the screw rotation speed, temperature, and machine movements. Proprietary [bib_ref] The development of a conical screw-based extrusion deposition system and its application..., Leng [/bib_ref] [bib_ref] Fabrication and properties of extrusion-based 3D-printed hardmetal and cermet components, Lengauer [/bib_ref] [bib_ref] Towards distributed recycling with additive manufacturing of PET flake feedstocks, Little [/bib_ref] [bib_ref] Large-format polymeric pellet-based additive manufacturing for the naval industry, Nieto [/bib_ref] or open-source options that can include or not slicing tools were used [bib_ref] Design and development of an extrusion system for 3D printing biopolymer pellets, Whyman [/bib_ref] [bib_ref] Direct coupling of fixed screw extruders using flexible heated hoses for FDM..., Khondoker [/bib_ref] [bib_ref] Mechanical properties of direct waste printing of polylactic acid with universal pellets..., Alexandre [/bib_ref] [bib_ref] Process parameters tuning and online re-slicing for robotized additive manufacturing of big..., Rebaioli [/bib_ref] [bib_ref] Technical pathways for distributed recycling of polymer composites for distributed manufacturing: windshield..., Dertinger [/bib_ref].
Finally, when the machine is ready, experiments for performance assessment can be performed. This involves initial tests for the system's calibration, usually to correlate the screw rotation speed with the material output at different temperatures [bib_ref] Investigation on the effects of process parameters in CNC assisted pellet based..., Kumar [/bib_ref] [bib_ref] Investigations on the melt flow behaviour of aluminium filled ABS polymer composite..., Kumar [/bib_ref] [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] [bib_ref] Direct coupling of fixed screw extruders using flexible heated hoses for FDM..., Khondoker [/bib_ref] [bib_ref] A large-scale double-stage-screw 3D printer for fused deposition of plastic pellets, Liu [/bib_ref] [bib_ref] Characterization of a granulate-based strand deposition process in the FLM-method for definition..., Schmidt [/bib_ref]. Knowing the volumetric output was critical to estimate a range of values for the deposition speed. Functional validation involves actual 3D printing and is often followed by a quality assessment. This often involves mechanical testing , microscopic analysis [bib_ref] Precision extruding deposition and characterization of cellular poly-ε-caprolactone tissue scaffolds, Wang [/bib_ref] [bib_ref] New developments in fused deposition modeling of ceramics, Bellini [/bib_ref] [bib_ref] Mechanical and in vitro evaluations of composite PLDLLA/TCP scaffolds for bone engineering, Lam [/bib_ref] [bib_ref] Fabrication of PCL/β-TCP scaffolds by 3D mini-screw extrusion printing, Dávila [/bib_ref] [bib_ref] Structural evolution of PCL during melt extrusion 3D printing, Liu [/bib_ref] [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] [bib_ref] Design and development of an extrusion system for 3D printing biopolymer pellets, Whyman [/bib_ref] [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] [bib_ref] 3D printing using powder melt extrusion, Boyle [/bib_ref] [bib_ref] Printability and physical properties of iron slag powder composites using material extrusion-based..., Kim [/bib_ref] [bib_ref] High-pressure interfacial impregnation by micro-screw in-situ extrusion for 3D printed continuous carbon..., Liu [/bib_ref] [bib_ref] Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration, Wang [/bib_ref] [bib_ref] Polymer-ceramic composite scaffolds: the effect of hydroxyapatite and β-tri-calcium phosphate, Huang [/bib_ref] [bib_ref] Direct powder extrusion 3D printing: fabrication of drug products using a novel..., Goyanes [/bib_ref] [bib_ref] Structure and mechanical behavior of Big Area Additive Manufacturing (BAAM) materials, Duty [/bib_ref] , and surface analysis [bib_ref] Fused deposition modelling using direct extrusion, Reddy [/bib_ref] [bib_ref] Screw extrusion-based additive manufacturing of PEEK, Tseng [/bib_ref] [bib_ref] 3D printing using powder melt extrusion, Boyle [/bib_ref]. Depending on the intended application, other tests (e.g., biological [bib_ref] Precision extruding deposition and characterization of cellular poly-ε-caprolactone tissue scaffolds, Wang [/bib_ref] [bib_ref] Mechanical and in vitro evaluations of composite PLDLLA/TCP scaffolds for bone engineering, Lam [/bib_ref] [bib_ref] Polymer-ceramic composite scaffolds: the effect of hydroxyapatite and β-tri-calcium phosphate, Huang [/bib_ref] [bib_ref] Aligned multi-walled carbon nanotubes with nanohydroxyapatite in a 3D printed polycaprolactone scaffold..., Huang [/bib_ref] [bib_ref] 3D Printing of polycaprolactone-polyaniline electroactive scaffolds for bone tissue engineering, Wibowo [/bib_ref] , physic-chemical [bib_ref] Fabrication of PCL/β-TCP scaffolds by 3D mini-screw extrusion printing, Dávila [/bib_ref] [bib_ref] Process-driven microstructure control in melt-extrusion-based 3D printing for tailorable mechanical properties in..., Liu [/bib_ref] [bib_ref] Additive manufacturing of heat-sensitive polymer melt using a pellet-fed material extrusion, Zhou [/bib_ref] [bib_ref] Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration, Wang [/bib_ref] [bib_ref] Polymer-ceramic composite scaffolds: the effect of hydroxyapatite and β-tri-calcium phosphate, Huang [/bib_ref] [bib_ref] 3D Printing of polycaprolactone-polyaniline electroactive scaffolds for bone tissue engineering, Wibowo [/bib_ref] [bib_ref] Additive manufacturing of fire-retardant ethylene-vinyl acetate, Geoffroy [/bib_ref] [bib_ref] Direct powder extrusion 3D printing: fabrication of drug products using a novel..., Goyanes [/bib_ref] might be needed. Comparison with the performance achieved by other machines and/or technologies may also be valuable [bib_ref] Mechanical properties of direct waste printing of polylactic acid with universal pellets..., Alexandre [/bib_ref] [bib_ref] Fabrication and properties of extrusion-based 3D-printed hardmetal and cermet components, Lengauer [/bib_ref] [bib_ref] Mechanical properties and applications of recycled polycarbonate particle material extrusion-based additive manufacturing, Reich [/bib_ref] [bib_ref] Technical pathways for distributed recycling of polymer composites for distributed manufacturing: windshield..., Dertinger [/bib_ref].
# Conclusions
A systematic review on screw-assisted EAM has been performed, identifying 61 publications on the field in the period from 2000 to 2020. Although only research papers in English were considered, a variety of experimental printheads and commercial systems based on screw extrusion for small-and large-scale applications could be found. This way, even if the repository does not cover the entire literature, the results bring together the main contributions to the technological development of screw-assisted EAM, and can be useful for future research on the field.
The publications were analyzed with respect to the annual production, geographic distribution, and co-occurrence of keywords, followed by the identification of the most influential articles in general, and most cited authors within the repository. The machines have been organized and described over time, to evidence the continuous improvements and evolution of the technique, and the printhead modifications have been summarized and correlated to the materials processed and performance requirements. In the end, a brief discussion on the common stages that have been undertaken for the development of the described equipment was made. Both for experimental and commercial systems, the most exploited constructive solution consists of vertical screw extruders, to which polymer-based granules are fed and then extruded through the deposition nozzle, constituting a single-stage process. For small-scale systems, the printheads often require a miniaturized screw mechanism, and its fabrication may represent most of the development costs. In turn, large-scale systems can benefit from benchtop-sized screw extruders, reducing costs and expediting equipment development. The aggregation state of the granules seems to be more critical to a successful extrusion process than physical properties of feedstock, and modifications on the design of the screw were made mostly to address specific performance requirements.
Although a systematic design approach has the potential to structure the development stages and bring together the different fields of knowledge (e.g., mechanisms, materials science, and manufacturing) that are necessary to achieve greater reliability, repeatability, and robustness, very few projects were carried out in a systematic way, impacting on the quality of the resulting 3D printed parts.
Finally, from the expressive variety of materials processed, it can be concluded that the screw-assisted printheads are important not only to expand the applicability of EAM but can also constitute experimental platforms for the formulation of novel material systems. In fact, in the academic context, the development of such equipment can help to democratize access to the technique, otherwise restricted to few research groups capable to afford the often expensive commercial systems.
Author contribution Joaquim M. J. Netto: conceptualization, methodology, investigation, writing-original draft, visualization. Henrique. T. Idogava: data curation, investigation, writing-original draft. Luiz E. Frezzatto Santos: data curation, investigation, writing-original draft. Zilda C. Silveira: conceptualization, methodology, writing-review and editing, supervision. Pedro Romio: data curation, investigation, writingoriginal draft. J. L. Alves: conceptualization, methodology, writingreview and editing, supervision Funding This study was financed by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior-Brasil (CAPES)-Finance Code 001 and by CNPq (grant no. 142348/2018-0).
Data availability Not applicable.
# Declarations
Ethics approval The work contains no libelous or unlawful statements, does not infringe on the rights of others, or contains material or instructions that might cause harm or injury.
Consent to participate The authors consent to participate.
## Consent for publication
The authors consent to publish.
## Competing interests
The authors declare no competing interests.
[fig] Figure 1: Flowchart of the systematic search, showing the resulting number of articles after each step [/fig]
[fig] Figure 3: Network map of the co-occurrence of keywords [/fig]
[fig] Figure 5: Schematic illustrations of the pioneering screw-assisted printheads, developed by a Bellini[10], b Reddy et al. [/fig]
[table] Table 1: Search strings used according to the electronic database [/table]
[table] Table 2: Selected information about the screw-assisted EAM systems, grouped according to the application scale [/table]
|
The Opportunities and Challenges Associated with the Implementation of Fourth Industrial Revolution Technologies to Manage Health and Safety
# Introduction
Health and safety (H&S) refer to the wellbeing and safety of humans from hazards. It includes programs, guidelines, and procedures that protect the safety, welfare, and health of any person engaged in work or employment, aiming to provide the ultimate safe working environment and reduce the risk of accidents and fatalities at work. Furthermore, it aims at protecting the health of customers and the public, including anyone who might be affected by the worksite environment. The H&S within the construction industry in South Africa is governed by legislation. The construction H&S legislation was introduced during the first world war, during a time when a high number of fatalities and injuries occurring. The authors Ibem and Laryeaexplain that the legislation derived from the 1918 Factories Act (Act no. , which set the standard for the South Africa industry. The Act was later improved and developed to the 1941 Factories Machinery and Building Work Act (Act no. and others, namely the 1983 Machinery and Occupational Safety Act (Act no. 6 of 1983) which 10 years later became the 1993 Occupational H&S challenges associated with implementing the 4iR technologies in the construction industry to manage H&S in South Africa.
Therefore, this study aimed at examining the opportunities and challenges within the construction sector in South Africa towards 4iR technologies implementation. The study commenced by identifying technologies that are available for use or adoption, primarily driven by the 4iR. Secondly, the current opportunities available for implementing 4iR technologies in the South African construction industry, and thereafter the challenges associated with implementing 4iR technologies, were identified. Based on the identified opportunities and challenges, the study proposes strategies to implement 4iR technologies in the South African construction industry. The approach may be applicable to other developing nations with similar characteristics.
# Materials and methods
The study employed a multi-pronged approach. Firstly, a systematic literature review was conducted to scope, plan, identify, screen, and assess the current body of knowledge on the subject. Materials for the review of literature on the factors of H&S non-compliance, strategies, opportunities, and challenges in the construction industry for 4iR technologies were sourced from journals listed in the Web of Science, Scopus, and Google scholar databases, including the International Journal of Environmental Research and Public Health and Safety Science in view of their safety-focused research areas. The multiple search sources were used for a comprehensive search in order to identify relevant publications as adopted by Qi, Razkenari, Costin, Kibert and Fuand Akram, Thaheem, Nasir, Ali and Khan. The formerargued that collating existing research articles from various perspectives can help understand the state-of-art on an issue. The literature search strategy included identifying keywords. In the current study, the keywords were: Innovations 4.0, Innovation, 4iR, H&S, Challenges, Opportunities, Technologies, Strategies, Awareness and South Africa. The review focused on papers published in the last ten years (2010 to 2020). According to Manda and Dhaou, this period coincides with implementing South Africa's digital transformation plan. The article searches in the various databases on the subject yielded 102 papers. The exclusion and inclusion criteria methods were used to select, screen, and identify the most suitable literature carefully. In addition, the articles were checked for duplication. After the qualification criteria and the duplicates were removed, 54 articles were retained and reviewed for the study.
Secondly, a questionnaire was used to collect data from respondents using a 5-point Likert scale as it was deemed suitable for this kind of study. The questions presented in the questionnaire were derived from the literature review on the opportunities and challenges associated with the implementation of 4iR technologies to manage H&S in the South African construction industry. Closed-ended questions were used to limit answers. The questionnaire was validated through a face validity scientific test to ensure the intended questions were maintained. The questionnaire was set up as follows: the first section, section A, explored the participants' background information. The second section contained questions about the opportunities to manage H&S using 4iR technologies. The third section included the challenges associated with implementing 4iR technologies. The respondents were required to indicate the extent to which they agreed with statements regarding the challenges and opportunities (identified from the literature review) for implementing 4iR technologies on a 5-point scale where 1 = Strongly disagree (SD); 2 = Disagree (D); 3 = Neutral (N); 4 = Agree (A) and 5 = Strongly agree (SA).
The convenience sampling was used to easily reach the targeted sample of construction safety personnel, including Safety Officers, Foremen, Site Engineers, Site Agent, Construction Manager, and other professionals. The respondents were selected on the basis that they manage H&S daily and are aware of 4iR in the industry. The safety personnel were sampled from construction sites that were running at the time of the study in the City of Johannesburg, Gauteng province. According to Habib'sreport from the University of Witwatersrand in Johannesburg, Gauteng is the first province to introduce 4iR technologies with major clients from Johannesburg. Thus, the study selected the City of Johannesburg.
A total of 110 respondents were targeted from the population on construction sites running at the time. This sample size was observed to be sufficient for studies of this nature to yield meaningful results, as was undertaken by Smallwood and Emuze, whereby 92 participants were included. Out of 110 distributed questionnaires, 88 were returned, which is an 80% return rate for the study. A return rate of less than 40% is unacceptable and yields validity issues, whereas a study with a return rate of 60% is considered suitable. Therefore, with a return rate of 80%, the researchers deemed the data sufficient for analysis. The returned questionnaires were then screened and analysed.
Thirdly, the data collected was captured on an excel spreadsheet and later transferred to the Statistical Package for the Social Sciences (IBM SPSS Statistics) version 26 software for quantitative data analysis. Before analysing the results, missing values were checked and excluded. Further, normal probability plots, histograms, and scatter plots were used to identify outliers and normality. The results showed that the data was not normally distributed, with few outliers emerging. The outliers were kept in the data because the respondents were the targeted population and represented valuable information.
This study employed both descriptive and inferential data analyses methods. The descriptive was used to determine the relative importance of the variables empirically using mean score (MS) and standard deviation (SD) values. The exploratory factor analysis was used to summarize the most significant variables and reduce them to quickly analyse and interpret the results and produce patterns and groupings that can be easily read and comprehended.
Preliminary analysis for the Exploratory Factor Analysis (EFA) included assessing the correlation matrix and testing with the Kaiser Meyer Olkin (KMO) and Bartlett's Sphericity tests. The data should be suitable for factor analysis, and to do so, the sample should be verified. Smaller samples result in the items' correlation coefficient being less reliable. In addition, the suitability of the strength of the inter-correlation among the variables should be verified. The factor analysis is questionable when no correlation is above 0.3. When Bartlett's test of Sphericity is less than 0.05 and KMO have a minimum 0.6 index, the factor analysis is appropriate. Once the factor assessment had been conducted, factor extraction followed in order to determine the factors that could be utilized to indicate the inter-relationships between the items; lastly, the factor rotation was conducted. The Varimax type of rotation method was used to rotate and combine the items in groups, and relative importance was represented as orthogonal for ease of interpretation. The Kaiser's criterion and scree tests were used to choose factors to be retained, with an eigenvalue greater than one. On the other hand, to interpret the EFA output, a cut-off value of 0.4 was used for the factor loadings. This was conducted in order to ensure significant outcomes and to avoid non-significance issues.
# Results
## Results from secondary data
This section presents results from the systematic literature review on challenges on H&S regulations compliance, 4iR in construction, available 4iR technologies in the construction industry, opportunities, and challenges in the construction industry to manage H&S, implementing 4iR technologies and the strategies to adopt 4iR in the South African construction industry.
## H&s regulations compliance challenges
The hazardous nature of the construction site often causes incidents and managing this issue has been a problem for years. However, there is the latest Acts that governs safety management on site such as the Occupational H&S Amendment Act, No. 181 of 1993 Labour Relations Act and Construction . Some factors affect the performance of H&S in the construction industry.
Al-Bayatidiscovered that the size of a contracting firm determines the level of compliance that occurs on construction projects; the smaller the company the poor the H&S performance. Windapo and Oladapodisclosed that subcontractors, lower management and supervisors in small firms do not implement H&S regulations. Additionally, the small contracting firms have a high rate of poor H&S due to a lack of funds. Consequently, the majority of fatalities on construction sites are caused by non-compliance with H&S regulations, which is not seen as an important factor mostly in small businesses.
Hon, Hinze and Chanopined that the accidents and injuries encountered on-site are due to ignorance by both the employees and employers. Furthermore, non-compliance results from negligent attitude, poor knowledge, and lack of understanding of the legislation and the profit focus attitude. The Department of Labour charges employers a huge amount for the penalties for non-compliance to enforce the implementation of the H&S regulations. However, this does not yield any change. H&S challenges cannot be ignored as this will lead to non-improvement of poor safety performance. Schwabrevealed that the 4iR technologies could help solve the safety problems on site. Thus, this study proposes the implementation of 4iR technologies.
## The 4ir in construction
The 4iR is South Africa's hope in boosting the current declining economy. More investment in the automation of construction industry is the one most important factor to boost the economy. However, the challenges of investing in the 4ir technologies are not covered in the paper. Through the reviewed literature, it was observed that there are benefits that this revolution brings to the construction industry. According to Choi, Ahn and Seo, the technologies will benefit the industry by resulting in the best form of accident prevention by protecting workers in hazardous areas through the provision of real-time data collection for safety reporting and incidents prevention. Furthermore, the technologies help provide greater visibility, better reporting, accountability, better communication, and improved workflows. These technologies will improve productivity, save on project time and cost, reduce workplace hazards, and push construction into the future by enhancing safety zones and mobility. In addition, the authors emphasized that the technologies could shape and improve H&S in the construction industry. Other benefits of 4iR include fast transaction, reduced cost and easy usage.
Nevertheless, the construction companies are afraid of the political landscape, concerned about potential job losses and their impact on labour forces and infrastructure challenges and that smart technologies cannot be adopted in an unstable environment. According to Olojede, Agbola and Samuel, South African construction companies have long-standing resistance to change and rather focus on traditional methods, poor productivity, and tight competition in the industry. Therefore, they are afraid that digitisation might affect sustainability of the industry. The article covered the challenges of embracing the 4ir in the construction industry. The Government should play a major role in promoting technologies through policies, standards, and procurements to make it easy for small and medium enterprises to adopt the technologies.
## Available 4ir technologies in the construction industry
The issue of poor H&S implementation cannot be ignored in the construction industry. To fight this, Lee, Shariatfar, Rashidi and Leeopined that the implementation of 4iR technologies needs to be in place to enhance the management of H&S at the workplace. The 4iR comprises technologies that work concurrently from the pre-planning stage to real-time construction site management of works, humans, and machines. The following technologies presented inbelow were identified from the literature review, particularly in the construction industry.
## Technologies description/function source
## Smart devices
Are tools attached on humans and plants, detecting possible hazards, monitoring their movements, computing the data and sounding alarms when nearing dangerous zones or objects.
Niu et al.,Geographical Information System (GIS)
Collects the geographic distribution of onsite works using spatial relations, creating a protocol that results in ease of H&S management. Virtual Reality (VR)
Is applied as an automated H&S training providing visualization of real-time detection of hazards, and enhancing knowledge on safety management.
Zhou, Whyte and SacksFour-Dimensional Computer Aided Design (4D CAD)
The information about the project activities is inserted in this technology. The information is then analysed, detect any possible risks and generate a safety management plan at the design stage.
Zhou, Whyte and Sacks; Zhang, Sulankivi, Kiviniemi, Romo, Eastman TeizerGlobal Navigation Satellite System (GNSS) Provides real-time monitoring of data of a large population from geosynchronous satellites, ensuring easy control and management of workplace safety.
Fenais, Ariaratnam, Ayer and SmilovskyGlobal positing system (GPS)
A positioning tool that uses wireless to track works and detect collision. It works as a security safeguard machinery in a robotic construction.
Li, Cheng and ChenWoodhead, Stephenson and Morreyopined that the use of IOT and RFID together will monitor and control H&S on construction sites. Meanwhile, Nnaji and Karakhanhave in real life shown that a multiuser friendly tool (virtual reality) can train the workforce to safely erect and dismantle a tower crane. The tool operates in a virtual form, providing steps to set up and operate the tower crane. Likewise, Li, Yan and Liuhave practised drones' operation in various construction projects. The drone is a vehicle used to perform various tasks, inspect works, monitor safety of employees, and identify hazards. Furthermore, tools such as smart sensors are identified to detect and report on any dangers at the workplace. Getuli, Ventura, Capone and Ciribiniexplained that BIM had been used as a semi-automatic tool to help in checking multiple safety regulations and safety plan and detect clashes to help safety performance.
The literature review has disclosed that these technologies are within 4iR in the construction industry. From the findings, it is discovered that technologies available in the construction industry include artificial intelligence, robotics, the internet of things, 3D
printing, drones, building information modelling, smart devices, virtual reality, geographic information system (GIS), 4D computer-aided designs, radio frequency identification (RFID), ultra-wide band (UWB), global navigation satellite system (GNSS), global positing system (GPS) and sensors. The technologies such as IOT, RFID, VR, sensors, drones, and BIM can be adopted to manage H&S by the management team, train relevant stakeholders on safety measures and monitor the safety of workers on site.
## Current opportunities to manage h&s using 4ir technologies
The management of H&S entails proper training, communication, monitoring and controlling. These are made easy by 4iR technologies, which encourage safety training using virtual reality, augmented reality, inspection through automation, simulation training, and collaborative (human-robot) teams. Furthermore, technologies such as BIM can enhance controlling and monitoring the overall project from the design phase to the closeout phase.
Choi, Ahn and Seocorroborated that the technologies benefit from fending off accidents, generating greater visibility, easing reporting procedures and accountability, providing healthier communication, and ameliorating workflow. On the other hand, Zhang, Cao, and Zhaoasserted that technologies such as GPS and RFID help monitor workplace operations, transfer communication, detect harmful areas, and report on possible incoming dangers. Nnaji, Gambatese, Lee and Zhangfound that the use of tools such as speed reduction systems (SRS) decreases and monitor the speed of vehicles, intrusion prevention and warning system (IPWS) warns workers and vehicles drivers when entering an intrusion zone and human-machine interaction detection system (HMIDS) warns the worker and driver of equipment collision.
From the literature findings, the opportunities existing in the construction industry to manage H&S include decreasing fatalities, more time to solve more complex tasks, greater visibility, better reporting, accountability, and communication, improved workflow, monitoring, control, and data collection. Current opportunities also include cost savings, reduced injuries, construction gains and sustainability, improved safety inspections, and better information management.
## Challenges of 4ir technologies implementation
The implementation of 4iR in the South African construction industry will attract investments. However, implementation is faced with the challenges outlined below.
## Construction firms' level of interest and views
Implementing 4iR technologies in the construction industry is perceived to be too expensive to adopt and maintain rather than innovate. On the other hand, Japheth and Kiprotichdisclosed that the professionals in the industry show no interest in implementing the technologies and resist change to their traditional ways. This low interest in embracng 4iR results from a lack of specialized professionals, technical skills and the client not insisting and strategizing on implementing the technologies. Likewise, Bayode, van der Poll and Ramphalpointed out that the construction industry is faced with insufficient electricity, unavailability of financial resources, poor accessibility to wireless broadband and lack of skills as notable barriers. Moreover, construction firms choose to stick to the proven methods of performing works and perceive adopting new technologies as risky.
## The size of projects and availability of resources
Most companies are not implementing the 4iR technologies because the projects they are involved in are small-to-medium in size. These companies are undecided about adopting the technological assets due to the cost affordability of implementing and maintenance. On the other hand, Alade and Windapoopined that the industry lacks dynamic capabilities for adopting technologies. Studies show that lack of education and unalignment of labour supply and demand are the challenges in the industry. The narrative that digitization will cause job losses is a significant worry of many companies, while another impediment is the lack of digital skills. Gaspar, Julião and Cruzasserted that the fear of job losses is a challenge to adopting 4iR. Furthermore, Kariemopined that lack of technical capacity with the absence of policy and regulation on 4iR implementation is another barrier to adopting the technologies.
## Unavailability of funds
The South African Cooperative Governance and Traditional Affairs is keen to support the civil construction industry to move to digital operations; however, it is faced with challenges in moving to e-governance. The challenges include developing policies for affordable access to developing mobile broadband infrastructure and adequate skills to develop e-government services. Furthermore, Alade and Windapodisclosed that the most significant challenge to implementing the technologies include the high cost of obtaining innovation and the high cost of training.
The literature revealed that the challenges associated with the implementation of 4iR technologies include lack of innovation, cost of adoption, fear of change and job losses, inadequate training capabilities, lack of interest, unavailability of specialists, unskilled technical support, lack of client insistence, insufficient electricity, and unavailability of financial resources. Others include lack of access to the wireless broadband, preference to traditional methods, size of the project, lack of education, misalignment of labour supply, unavailability of funds from the client and lack of adequate skills.
## Implementation strategies
The implementation of 4iR technologies is a critical factor in managing H&S in the construction industry. However, the implementation in South Africa is challenged with issues discussed in the above section. It is of great importance to mitigate these challenges. Osunsanmi, Aigbavboa and Okeindicated a low level of awareness of these technologies and a low level of understanding of how the technologies operate to manage works in the industry. This level of awareness results from the lack of understanding of the benefits in the construction industry. This suggests that the adoption of 4iR can be improved through training, workshops, and seminars.
On the other hand, Aghimien, Aigbavboa, Aghimien, Thwala and Ndlovuconcludes that the industry has a low awareness rate of the benefits of 3D printing. Therefore, higher education institutions should improve the training on these technologies in their syllabuses. In addition, case studies should be conducted, a module in 4iR should be incorporated in the construction department and professionals should be educated on the technologies. Ignoring these strategies of overcoming the implementation challenges of these technologies will result in the industry lacking the required skillsand thus will face difficulties in the future.
The literature review revealed that the suggested strategies towards implementing 4iR technologies in the construction industry include educating relevant parties, technology enlightenments, campaigns, training programmes, skill developments, workshops, seminars, government enforcement, regulation of the technologies and H&S policies on technologies.
## Empirical findings
This section presents results on the opportunities and challenges of implementing 4iR technologies in the construction industry in South Africa.
## Opportunities for 4ir implementation in h&s management
Descriptive and inferential analysis results on the 4iR technologies opportunities to manage H&S in the construction industry are presented in this section.
Descriptive Results on the Opportunitiesshows that better information management was the most important opportunity for implementing 4iR technologies. This variable was ranked first, with MS of 4.21 and SD of 1.00; improved workflow was ranked second (MS = 4.20; SD = 0.98); improved safety inspections followed (MS = 4.16; SD = 1.00); better accountability was ranked fourth (MS = 4.14; SD = 0.91); preventing accident was ranked fifth (MS = 4.13; SD = 0.97). On the other hand, the least important variables were more time to solve more difficult issues (MS = 3.98; SD = 0.93); construction gains sustainability (MS = 3.90; SD = 1.04) and cost savings (MS = 3.47; SD = 1.30). These suggest that some opportunities are perceived more important in managing H&S than others.
## Inferential analysis results (factor analysis) on existing opportunities
An exploratory factor analysis (EFA) of the opportunities revealed the opportunities existing to manage H&S in the construction industry using 4iR technologies. Before conducting the PCA, the suitability of the data for factor analysis was evaluated. The inspection of the correlation matrix showed the presence of coefficient of over 0.30 for most of the variables which was suitable for factor analysis. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy achieved a value of 0.82, which is greater than the recommended minimum value of 0.6 and Bartlett's Test of Sphericity value was significant (0.000). Therefore, the factorability of the data was possible. The findings revealed that three components could be retained. The eigenvalues greater than one (1) Kaiser's criterion was adopted. The 3 components had eigenvalues of 7.60, 1.44 and 1.05, and contributed 50.69, 9.57 and 7.01% of their variance respectively. The three groups have a total cumulative variance of 67.27% of the total importance, which indicate their significance from the fifteen opportunities selected. The factors from the principal component analysis with a loading of 0.4 should retain at least 3 components with a minimum of total variance extracted of 50%.shows the scree plot break after the third component. The steep slope represented the larger components, while the gradually decreasing components presented the rest of the variables with eigenvalue less than 1. The three groups located on the steep slope were retained. Varimax rotation was carried out to interpret the three groups of the opportunities existing in the construction industry for 4ir technologies to manage H&S. This gave rise to the pattern matrix presented in.shows the scree plot break after the third component. The steep slope represented the larger components, while the gradually decreasing components presented the rest of the variables with eigenvalue less than 1. The three groups located on the steep slope were retained. Varimax rotation was carried out to interpret the three groups of the opportunities existing in the construction industry for 4ir technologies to manage H&S. This gave rise to the pattern matrix presented in. Findings from the factor analysis identified the following three components. The percentages represent component loadings, respectively. Component 1 has eight items allotted to it, as seen in. These eight items belong to site specific benefits as concluded by Menoof the opportunities existing in the construction industry for 4ir technologies to manage H&S. The component presents: construction gains sustainability (91.00%), decreases fatalities (79.30%), reduce injuries (70.50%), improve workflow (70.10%), better collection of data platform (68.50%), improved safety inspections (65.70%), better information management (63.20%) and better accountability (58.10%). The percentages represent components loadings, respectively. Component 2 has four items allotted to it as seen in. These four items belong to company specific benefits as concluded by Menoof the opportunities existing in the construction industry for 4ir technologies to manage H&S. The component presents: saves on cost (44.10%), create greater visibility (74.10%), more time to solve more difficult issues (66.10%) and preventing accident (56.20%). The percentages represent components loadings, respectively. Component 3 has three items allotted to it as seen in. These three items belong to project specific benefits as concluded by Menoof the opportunities existing in the construction industry for 4ir technologies to manage H&S. The component presents: better reporting (97.90%), better communication (84.50%) and improve in controlling and monitoring (54.00%). The percentages represent components loadings respectively.
## Challenges in implementing 4ir technologies
Descriptive and inferential analysis results on the challenges associated with implementing 4iR technologies in the construction industry are presented in this section.
Descriptive Results on the Implementation Challengesshows that the high cost of technologies was the most significant barrier. This aspect was ranked first, with a mean score (MS) of 3.95 and a standard deviation (SD) value of 1.20; fear of job losses was ranked second (MS = 3.92; SD = 1.15). A lack of adequate skills ranked third (MS = 3.72; SD = 1.18); the lack of cost to adopt ranked fourth (MS = 3.58; SD = 1.21); the unavailability of training capabilities was ranked fifth (MS = 3.57; SD = 1.34). On the other hand, the least important challenges were lack of access to the wireless broadband, which ranked fourteenth (MS = 3.16; SD = 1.27); lack of interest ranked fifteenth (MS = 3.07; SD = 1.30); insufficient electricity ranked sixteen (MS = 2.84; SD = 1.49). These suggest that some challenges may pose more threat to the implementation of 4iR than the others.
## Inferential analysis results (factor analysis) on implementation challenges
An exploratory factor analysis (EFA) was conducted on the challenges associated with implementing 4iR technologies in the construction industry. All nineteenvariables were retained. Principal component analysis (PCA) was utilized to analyse the 15 variables using SPSS version 26 software.
Before carrying out the PCA, the suitability of the data for factor analysis was evaluated. The inspection of the correlation matrix showed the presence of a co-efficient of over 0.30 for most of the variables, which was suitable for factor analysis. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy achieved a value of 0.80, more significant than the recommended minimum value of 0.6 and Bartlett's Test of sphericity value of 0.000 was acceptable. Therefore, the factorability of the correlation matrix is significant as shown in. The results inrevealed the challenges associated with implementing 4iR technologies in the construction industry with their respective eigenvalues. The eigenvalues greater than one Kaiser's criterion was adopted. A total of 4 challenges were retained: 6.10, 1.97, 1.72 and 1.43, with 32.08, 10.39, 9.03 and 7.51% of their variance, respectively. This means that the first group accounted for 32.08% required for challenges, the second group accounted for 10.39%, the third group accounted for 9.03%, and the fourth group accounted for 7.51%. The 4 groups have a total cumulative variance of 59.02% of the total importance, which indicate their significance from the nineteen challenges selected. The factors from the principal component analysis with a loading of 0.4 should retain at least three components with a minimum of total variance extracted of 50%.shows the scree plot break after the fourth component. The steep slope represents the larger components, while the gradually decreasing components represent the rest of the variables with eigenvalue less than 1. The four groups located on the steep slope were retained. This gave rise to the pattern matrix presented in.shows the scree plot break after the fourth component. The steep slope represents the larger components, while the gradually decreasing components represent the rest of the variables with eigenvalue less than 1. The four groups located on the steep slope were retained. This gave rise to the pattern matrix presented in. Findings from the factor analysis identified the following four components. The percentages represent component loadings, respectively. Component 1 has seven items allotted to it as seen in. These seven items belong to people-related factors that can affect 4iR implementation, as concluded by Odubiyi. The component represents: unskilled technical support (83.20%), unavailability of specialist (82.90%), unavailability of training capabilities (65.00%), lack of adequate skills (62.40%), lacks client insistence (54.00%), lacks interest (64.40%) and prefer traditional method (40.30%). Component 2 has three items allotted to it as seen in. These three items belong to management-inclined problems as concluded by Odubiyi. The component represents: technologies are too expensive (74.60%), fear of change (64.40%) and lack of innovation (52.20%). Component 3 has five items identified as cost-related problems. The variables here include: insufficient electricity (45.20%), unavailability of financial resources (80.70%), unavailability of funds from client (70.30%), lack of cost to adopt (60.10%) and size of project (60.00%).
Component 4 has four items that belong to standardization problems associated with 4iR implementation. The factors include: lack of education (79.50%), unalignment of labour supply (73.10%), fear of job losses (65.60%) and lack of access to the wireless broadband (61.20%). these relate to the enlightenment strategies.
# Discussion
This section presents the findings of 4iR technologies opportunities that exist in the construction industry to manage H&S and challenges faced by the construction sector in implementing 4iR technologies. The findings from descriptive and factor analysis obtained in the previous results section are discussed in relation to the literature reviewed earlier in this research. The findings align with the results identified in literature. However, the scaling of respondents on the challenges differs.
## Opportunities for 4ir technologies to manage h&s
The systematic literature review discovered the 4iR technologies opportunities that exist include decreasing fatalities, more time to solve more complex tasks, greater visibility, better reporting, accountability, and communication, improved workflow, monitoring, control, and data collection. Current opportunities also include cost savings, reduced injuries, construction gains and sustainability, improved safety inspections, and better information management.
The descriptive results ranked variables through MIS and standard deviation, respectively. The most significant variables were better information management, improved workflow, improved safety inspections, better accountability, and preventing accidents. Furthermore, the least essential variables include better communication, reduced injuries, more time to solve more complex issues, construction gains sustainability and saves on cost.
The factor analysis findings presented the groupings for these technologies, which are component one: site-specific benefits, including improved workflow, improved safety inspections, better information management, and better accountability as most important. Construction gains sustainability and reduces injuries were noted as the least important. Component two included company-specific benefits: preventing accidents as most essential and saves on cost and time to solve more complex issues as least important. Component three includes project-specific benefits, which include better communication as the least.
The benefits make H&S management easier and faster. Information flow and progress reporting are also easier to handle because these technologies are automated and can control themselves and suggest safer decisions. The recognition of all technologies available still needs to be raised because the opportunities of the technologies are not known by many safety personnel. Therefore, the strategies of implementation are mandatory for the management of H&S.
From the results, safety personnel perceived the most regarded opportunities for the implementation of 4iR were improved workflow, improved safety inspections, better Information management and better accountability. These are in line with the findings of Choi, Ahn, and Seo, who opined that the technologies will benefit in fend off accidents, generate greater visibility, ease reporting procedures and accountability and ameliorate workflow. This is probably because these technologies are automated and can control themselves and suggest safer decisions. On the other hand, Nnaji and Karakhafound that implementing these technologies is essential for enhancing construction sustainability using preventive tools that reduce injuries and accidents at the workplace. The authors' finding contradicts the findings of this research, which perceived construction gains and sustainability and reduced injuries as less significant opportunities.
## Component two: company-specific benefits
The findings revealed that the most significant company-specific benefit was preventing accidents. The reason behind this perception might be the ability of these technologies to detect and prevent incidents. This matched with the results found by Nnaji and Karakhan, who suggested that the preventive tools make the site a controllable place that prevents accidents 4.1.2. Component Three: Project-Specific Benefits
The finding in this group described better communication as the least important of project opportunities in employing 4iR. Education on these technologies is lacking, which implies that safety personnel do not understand the operations of the automated workplace. The results mismatch with the findings of Choi, Ahn and Seo, who suggested that the technologies could provide healthier communication.
Extant literature showed that the use of 4iR technologies can improve safety of individuals and equipment at the worksite. Cases of injuries, accidents, and fatalities can be eliminated using sensors, RFID, GPS, and smart devices. Additionally, the implementation of the 4iR technologies will encourage safety training using Virtual Reality and or Augmented Reality and conducting inspections through automation. Further, 4iR can improve H&S practices/performance by simulating trainings. These technologies do not only benefit construction projects but also the organisation in planning and managing occupational H&S from the initial stage using tools such as BIM, IoT and 4D cad. These tools are able to produce safety management plans and be able to monitor the plan during process with the help of tools such as drones, GPS and GNSS. These tools are generally used for digital data collection from a targeted area and use it within BIM, IOT and 4D cad to compare with the planned H&S measures.
## Challenges in implementing 4ir technologies
This section presents the challenges faced by the construction sector in implementing 4iR technologies. The findings from descriptive and factor analysis obtained in previous results section are discussed in relation to literature reviewed earlier in this research. The findings align with the findings identified in the literature that these exist. However, the scaling of respondents on the challenges differs.
The systematic literature review identified the challenges associated with the implementation, including too expensive technologies, fear of job losses, lack of adequate skills, lack of cost to adopt, and unavailability of training capabilities. On the other hand, the challenges with lesser risks include the unavailability of specialists, lack of education, lack of access to wireless broadband, lack of interest and insufficient electricity.
The descriptive results supported that expensive technologies, fear of job losses, lack of adequate skills, lack of cost to adopt and unavailability of training capabilities were significant challenges. Further, the least important challenges were unavailability of specialists, lack of education, lack of access to wireless broadband, lack of interest and insufficient electricity.
The factor analysis findings presented the groupings for these barriers, which are component one: people-related factors including lack of adequate skills and unavailability of training capabilities as the most important challenges and lack of interest and unavailability of specialist as the least important; and component two: management-incline problems which notes that costly technologies is the most critical challenge. Component three-costrelated problems included lack of cost to adopt as the major challenge and insufficient electricity as the least important factors; and component four: standardization problems which include fear of job losses as the most important variable and lack of education and access to wireless broadband as the least important.
The findings revealed that the unavailability of training capacities and lack of adequate skills are the main threats to implementing the 4iR technologies in the construction industry. The results are in line with Japheth and Kiprotich, who disclosed that the professionals are afraid to change the traditional way and adopt the new technologies and, most importantly, employ digital training. Likewise, Kariememphasized that poor technical capacity with no plan for improvement is another threat to adopting the new technologies. This is probably because 4iR technologies are perceived as conspiracies and not possible to implement in South Africa. Further, the findings are mirrored with the results of Moloi, Zibani, and Makhubelawho stated that developing e-government services in the industry is hindered by lack of adequate skills. The reason behind this might be that these tools are new in the country and there are no professionals or technicians acquainted with operating them.
The results also disclosed a lack of interest and the unavailability of specialists as the minor threats to implementing new technologies. However, this contrasts with the findings of Mahachiwho asserted that the construction industry lacks interest because of perceptions that it is impossible to implement these technologies. Further, the findings of Kariemwho concluded that the industry lack technical personnel to operate these technologies contrast this study's findings.
The imagination and ignorance further hinder the implementation of these new technologies from perceptions. This is because the small and medium enterprises in the industry in South Africa mostly rely on government projects based on traditional operations. The results match with the findings of Mahachiwho asserted that the construction sector refuses to innovate to develop a strategic plan to eliminate this challenge but instead conclude that the technologies are too expensive.
The challenge of lack of cost to adopt this method was one of the greatest threats to the implementation of 4iR technologies. The truth is that implementing these technologies will not be easy because it might mean that most of the traditional ways would come to an end. This finding is mirrored with Janse van Rensburgwho concluded that small and medium projects with less management resources and asserts are not ready mostly financially to embrace the technologies. This is probably because projects of these sizes do not have enough funds to cover up the employment of the technologies. From the findings, the insufficient electricity challenge is less risk to the implementation of technologies. This mismatched with Bayode, Van der Poll and Ramphalwho emphasized that the industry is constantly experiencing insufficient electricity. This probably results from the country's shortage of resource to generate adequate electricity to stakeholders.
Due to the level of unemployment is very high in South Africa, with many looking for jobs for a long time with no luck, the safety personnel of the construction industry see the adoption of technologies as the greatest threat to their jobs. This is perhaps why the fear of job loss was found to be one of the highest risks in implementing technologies in the industry, as was found in Gaspar, Julião and Cruz. Likewise, Birkel, Veile, Müller, Hartmann, and Voigtemphasized that job losses are a severe fear in the industry. Further, the challenges of lack of education and lack of access to wireless broadband were found to be less of a risk in adopting the technologies. However, when one does not have the knowledge or the skills to operate these tools, they will have difficulties working/operating the tools. Further, most technologies use wireless to operate and thus lack of education and access are serious barriers Birkel, Veile, Müller, Hartmann, and Voigt. These findings are contrary to Bayode, Van der Poll and Ramphal, who asserted that access to wireless broadband is a severe challenge to the operation of these tools considering that the tools use much of wireless to operate.
The negative H&S issues prevalent in the construction industry need to be addressed. To mitigate the effects of H&S challenges, the use of 4iR technologies is key. The use of the technologies in turn can be enabled through raising the level of awareness achieved through providing training, conducting workshops and seminars. Further, to enhance skills and education on 4iR operations, institutions of higher education should incorporate 4iR practices in the syllabus. Additionally, professionals who are already working in the industry need to be provided education and skills either through professional training or short courses. Moreover, the paper recommends client incentives to drive 4iR technology adoption.
# Conclusions
The construction industry is an essential sector contributing to the economic growth of South Africa. The industry is constantly experiencing poor H&S performance that leads to indirect H&S costs from injuries, fatalities and accidents claims. The government made efforts of establishing occupational H&S regulations and practices, however, the industry has found difficulties in practising the regulations. Therefore, this study proposes using 4iR technologies to manage H&S performance.
Implementing 4iR technologies will help reduce fatalities, injuries, and accidents, finally improving performance. Moreover, it will produce construction work that has fewer hazards. The implementation of 4iR technologies is faced with severe challenges. This study established the many issues faced by the construction industry in adopting 4iR technologies. Although the study presents findings from South Africa, the issue of 4iR adoption is faced by many developing countries. Hence, the study provides information to an international audience of Built Environment practitioners and scholars to compare the applicability of similar phenomenon and/or methodologies in their countries. The issues go beyond just low level of awareness as identified by several scholars. For example, construction companies do not implement these technologies because of the nature of contracts. Most companies rely on the national government for support and the size of their firms is also a factor. The companies rely on government contracts that have a standard approach for carrying out contracts. Therefore, there is a great need for the national governments to revisit their procedures, policies, and regulations to enable technologies to be implemented. Moreover, most companies lack the financial capability and are not sure if they can maintain the 4iR process; therefore, proper planning to accumulate financial resources should be established. The government and the construction firm should develop strategies to adopt and maintain these technologies in the industry to overcome the challenges faced by the industry to implement the technologies.
The issue of 4iR adoption is faced by many developing countries. Publishing this article in a global journal will ensure that the research is easily visible for broader audience, helping other scholars to compare the applicability of similar phenomenon and/or methodologies in other countries. Moreover, the subject of 4iR in health and safety management is still not fully comprehended on a global scale. Therefore, lessons even from local settings are useful at a global level. Data Availability Statement: Not applicable.
## Conflicts of interest:
The authors declare no conflict of interest.
|
Cytoskeletal Proteins in Myotendinous Junctions of Human Extraocular Muscles
PURPOSE. The purpose of this study was to investigate the cytoskeletal composition of myotendinous junctions (MTJs) in the human extraocular muscles (EOMs). Desmin and other major cytoskeletal proteins are enriched at the MTJs of ordinary myofibers, where they are proposed to be of particular importance for force transmission and required to maintain myofiber integrity.METHODS.EOM and limb muscle samples were analyzed with immunohistochemistry using antibodies against the intermediate filament proteins desmin, nestin, keratin 19, vimentin, and different myosin heavy chain (MyHC) isoforms. MTJs were identified by labeling with antibodies against laminin or tenascin.RESULTS.In contrast to MTJs in lumbrical muscle where desmin, nestin, and keratin 19 were always present, approximately one-third of the MTJs in the EOMs lacked either desmin and/or nestin, and all MTJs lacked keratin 19. Approximately 6% of the MTJs in the EOMs lacked all of these key cytoskeletal proteins.CONCLUSIONS.The cytoskeletal protein composition of MTJs in human EOMs differed significantly from that of MTJs in limb muscles. These differences in cytoskeletal protein composition may indicate particular adaptation to meet the functional requirements of the EOMs.
T he myotendinous junction (MTJ) is the specialized site for force transmission from the intracellular contractile proteins of the myofiber to the extracellular connective tissue proteins of the tendon. [bib_ref] Tendons and ligaments-an overview, Benjamin [/bib_ref] The myofibers terminate at MTJs with digit-like projections surrounded by a basement membrane rich in laminin, tenascin, and collagen fibers on the tendon side. [bib_ref] Macromolecular organization of basement membranes, Timpl [/bib_ref] [bib_ref] Chick myotendinous antigen. I. A monoclonal antibody as a marker for tendon..., Chiquet [/bib_ref] The muscle side of the MTJs is particularly rich in cytoskeletal intermediate filament proteins associated with force transmission from the thin filaments to the cell membrane. The cytoskeleton is particularly important for the establishment of series and parallel connections with the contractile proteins resulting in a cytoskeletal lattice through which force can be transmitted. Furthermore, intermediate filament proteins also link components of myofibrils to membrane protein complexes along the surface of the myofiber, enabling lateral transduction of force generated during muscle contraction. [bib_ref] Overview of the muscle cytoskeleton, Henderson [/bib_ref] Desmin, the first muscle-specific protein detected during development, is the major intermediate filament protein in myofibers, linking adjacent Z-discs in neighboring myofibrils and connecting the most peripheral myofibrils to the costameres in the subsarcolemmal cytoskeleton. [bib_ref] Desmin/vimentin intermediate filaments are dispensable for many aspects of myogenesis, Schultheiss [/bib_ref] Desmin also provides a network to interconnect nuclei, mitochondria, and other cytoplasmic organelles, and is considered pivotal for maintenance of mature myofiber integrity upon loading and continued use. Lack of desmin targets highly used muscles working against high loads. [bib_ref] Cardiovascular lesions and skeletal myopathy in mice lacking desmin, Li [/bib_ref] Desmin is enriched at MTJs as well as neuromuscular junctions (NMJs) in limb muscle. Nestin and vimentin are two additional important intermediate filament proteins that are coexpressed with desmin during skeletal muscle development. [bib_ref] Sejersen T. Colocalization of nestin and vimentin/desmin in skeletal muscle cells demonstrated..., Sjöberg [/bib_ref] Nestin expression is downregulated during muscle maturation and in adult muscle it is restricted to NMJs and MTJs [bib_ref] Nestin is expressed during development and in myotendinous and neuromuscular junctions in..., Carlsson [/bib_ref] [bib_ref] Specific and innervation-regulated expression of the intermediate filament protein nestin at neuromuscular..., Vaittinen [/bib_ref] as well as regenerating myofibers. [bib_ref] The expression of intermediate filament protein nestin as related to vimentin and..., Vaittinen [/bib_ref] Vimentin is widely expressed in skeletal myofibers in the early and middle stages of muscle development and becomes completely undetectable in fully developed myofibers. [bib_ref] The cytoskeletal lattice of muscle cells, Small [/bib_ref] Keratin 19, another intermediate filament protein expressed in developing muscle but downregulated as myofibers mature, is present at both the Z-discs and the M-band of the most peripheral myofibrils linking them to the sarcolemma in mature muscle. [bib_ref] Overview of the muscle cytoskeleton, Henderson [/bib_ref] [bib_ref] Absence of keratin 19 in mice causes skeletal myopathy with mitochondrial and..., Stone [/bib_ref] [bib_ref] Transient cytokeratin expression in skeletal muscle during murine embryogenesis, Kosmehl [/bib_ref] Extraocular muscles (EOMs), consisting of four rectus and two oblique muscles, are highly specialized muscles responsible for the complex eyeball movements. In order to optimize perception and maintain stereo-vision in every gaze position, it is necessary to perfectly control and coordinate the delicate eyeball movements. The complexity of actions performed by the EOMs is reflected in their distinct myosin heavy chain (MyHC) composition, structural organization, innervation patterns, and physiological properties, that separate them from other skeletal muscles. [bib_ref] Myosin heavy chain isoforms in human extraocular muscles, Kjellgren [/bib_ref] [bib_ref] Definition of the unique human extraocular muscle allotype by expression profiling, Fischer [/bib_ref] Desmin was regarded as a ubiquitous muscle protein until we reported absence of desmin from a subgroup of normal and intact myofibers in the adult human EOMs. [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] [bib_ref] Complex correlations between desmin content, myofiber types, and innervation patterns in the..., Liu [/bib_ref] We initially reported absence of desmin in a subgroup of myofibers containing MyHC slow-tonic and MyHCI (MyHCsto/I) [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] but we have recently shown that there is a complex relation between desmin content and innervation in the EOMs. [bib_ref] Complex correlations between desmin content, myofiber types, and innervation patterns in the..., Liu [/bib_ref] In the present study, we investigated the intermediate filament protein composition with respect to desmin, nestin, keratin 19, and vimentin in relation to myofiber types at MTJs in the human EOMs. We found that there are important differences between MTJs of EOMs and limb muscles.
# Material and methods
## Human muscle samples
Nine normal human samples were collected from the anterior portion of the lateral rectus, superior rectus, medial rectus, and inferior rectus muscles of 6 men with a mean age of 66 years (range = 42-83 years old) and lumbrical muscles of the hand from one man (75 years old) and one girl (13 years old) at autopsy, with the approval of the Regional Ethical Review Board in Umeå and was completed in accordance with the principles of the Declaration of Helsinki. The muscles were obtained from deceased individuals who, when alive, chose to donate their eyes and other tissues postmortem for transplantation and research purposes, according to Swedish law. All donors met the criteria for tissue donation for transplantation and none of the subjects was known to suffer from neuromuscular disease. The muscle samples were mounted on cardboard and rapidly frozen in propane chilled with liquid nitrogen, and stored at -80°C until used. Care was taken to keep the orientation of the EOM specimens in order to allow the longitudinal sections to include both the orbital and global layers. Serial longitudinal sections, 5 to 6 μm thick, were cut at -23°C in a cryostat (Reichert Jung; Leica, Heidelberg, Germany) and stored at -20°C until processed for immunohistochemistry. The smaller diameter of the myofibers in the orbital layer and the fact that the orbital layer does not extent all the way to the tendon allow the clear identification of the two layers in the longitudinal sections.
## Antibodies and immunofluorescence
The following monoclonal antibodies were used: D33 to detect desmin, 17 MAB2736 and MAB1259 to detect nestin (R&D System, Abingdon, UK), and M0725 to detect vimentin (DAKO, Glostrup, Denmark). Rabbit polyclonal antibody ab15463 was used to detect keratin 19 (Abcam, Cambridge, UK). In addition, monoclonal antibodies against distinct MyHC isoforms were also used to identify the three major myofiber types in the EOMs, as previously reported [bib_ref] Complex correlations between desmin content, myofiber types, and innervation patterns in the..., Liu [/bib_ref] Because each EOM myofiber and its MTJ can be reliably studied only on a single longitudinal section, given the very small diameter of the myofibers in the EOMs, sections were processed for double or sequential immunostaining, as previously described. [bib_ref] Complex correlations between desmin content, myofiber types, and innervation patterns in the..., Liu [/bib_ref] In brief, the tissue sections were air-dried, rehydrated in 0.01 M PBS, and then blocked with 5% normal donkey or goat serum for 15 minutes. Sections were then incubated with the first two primary antibodies, one against a cytoskeletal protein on the muscle side of the MTJs and another antibody against an extracellular matrix protein on the tendon side of the MTJs (desmin + laminin; nestin + laminin; vimentin + laminin; or keratin 19 + tenascin) at +4°C overnight. Thereafter, the sections were washed in PBS, additionally blocked with 5% normal serum for 15 minutes, and then incubated for 30 minutes at 37°C with the appropriate secondary antibodies .
Control sections were treated as above, except that the primary antibodies were omitted. No staining was observed in control sections. The sections were examined and photographed with a Spot camera (RT KE slider; Diagnostic Instruments, Inc., Sterling Heights, MI, USA) connected to a Nikon microscope (Eclipse, E800; Nikon, Tokyo, Japan).
After photographing, the third primary antibody (nestin or desmin) was applied respectively to the sections double labeled with the other antibodies (for example, the antibody against nestin was applied to a section treated with antibodies against desmin + laminin, in order to obtain a triple staining with desmin + laminin + nestin), for 60 minutes at 37°C, followed by incubation with the appropriate secondary antibody for 30 minutes at 37°C. The sections were examined and photographed again as above. Subsequently, sequential immunolabeling with primary antibodies against MyHCI (BA-D5), MyHCIIa (A4.74), and/or MyHCI + MyHCIIa + MyHCeom (N2.261) was performed on the aforementioned sections, adding one antibody at a time and taking photographs of the tissue section before adding the next MyHC antibody, followed by incubation with the appropriate secondary antibodies for 30 minutes at 37°C. Detailed information on the secondary antibodies used to detect each primary antibody is provided in .
The images were processed using the Adobe Photoshop software (Adobe System, Inc., Mountain View, CA, USA).
# Results
MTJs were identified by their typical appearance and labeling with antibodies against laminin or tenascin on the tendon side, in all muscle samples examined. In lumbrical muscles, desmin immunoreactivity was observed along the length of all myofibers and their MTJs. Higher desmin immunoreactivity was observed at the MTJs (Figs. 1A-C). Immunoreactivity with nestin was restricted to the myofiber region close to the MTJ and became stronger at the MTJ itself (Figs. 1D-F). Weak rather than strong labeling with nestin was occasionally found at MTJs (Figs. 1D-F). Immunolabeling with keratin 19 was also observed at the tip of the myofiber and it generally became stronger at the MTJs (Figs. 1G-I). Vimentin immunolabeling was not observed in myofibers of lumbrical muscle, nor at the MTJs, but it was sometimes present in a string on the tendon side of the MTJs (Figs. 1J-L).
## Mtjs in eoms
The longitudinal sections of EOMs investigated in the present study comprised approximately the most anterior one fourth of each muscle. MTJs were encountered in all muscle sections examined, in both the global and the orbital layers, with higher frequency in the global layer. The MTJs were generally found at the myofiber end facing the tendon (Figs. 2A-C), as in limb muscles. However, occasional structures with similar morphology to that of typical MTJs were also found at the opposite end of the myofiber, that is, the end facing away from the tendon [fig_ref] FIGURE 2: MTJs in the human EOMs labeled with the antibody against laminin [/fig_ref] , and were myo-endomysial junctions, as previously described. [bib_ref] Force transmission in skeletal muscle: from actomyosin to external tendons, Patel [/bib_ref] [bib_ref] Complex threedimensional patterns of myosin isoform expression: differences between and within specific..., Mcloon [/bib_ref] [bib_ref] Muscular force transmission: a unified, dual or multiple system? A review and..., Huijing [/bib_ref] A total of 493 MTJs were examined in 9 human EOM specimens. Heterogeneous labeling patterns with antibodies against intermediate filament proteins were noted, in contrast to the consistent labeling patterns described above and previously reported for the MTJs of limb muscles. [bib_ref] Nestin is expressed during development and in myotendinous and neuromuscular junctions in..., Carlsson [/bib_ref] [bib_ref] Specific and innervation-regulated expression of the intermediate filament protein nestin at neuromuscular..., Vaittinen [/bib_ref] In the EOM samples studied, approximately 85% of MTJs were present in myofibers containing MyHCIIa, and the rest were present in myofibers containing MyHCI.
## Desmin
The vast majority of the MTJs (85%) examined in muscle sections treated with the antibody against desmin were labeled with this antibody [fig_ref] FIGURE 3: MTJs in EOMs double-immunolabeled with antibodies against desmin [/fig_ref] and these MTJs were observed in myofibers containing either MyHCIIa or MyHCI, in both the global and the orbital layers. The number of labeled MTJs varied among specimens within and between subjects. The immunoreactivity for desmin at MTJs varied from strong to weak. More than half (68%) of the MTJs labeled with desmin in these muscle sections showed stronger immunoreactivity than that observed in the remaining length of the myofiber present in the longitudinal section, irrespective of whether the tip of the myofiber was strongly or weakly labeled with this antibody [fig_ref] FIGURE 3: MTJs in EOMs double-immunolabeled with antibodies against desmin [/fig_ref]. Approximately 28% of MTJs labeled with desmin showed identical levels of immunoreactivity as the remaining part of the myofiber present in the section (not shown). Decreased immunoreactivity with desmin was sporadically (approximately 4%) observed in the folds of MTJs of myofibers containing desmin along the remaining of their length (not shown).
Fifteen percent of the total number of MTJs in muscle sections treated with the antibody against desmin were unlabeled with this antibody (Figs. 3G-I) and were found in both the global (54.5%) and the orbital (45.5%) layers. Immunoreactivity with the antibody against desmin was generally absent along the remaining length of the myofibers displaying the unlabeled MTJs (Figs. 3G-I). MTJs unlabeled with the antibody against desmin were observed mostly in myofibers containing MyHCIIa (71%), but also in myofibers containing MyHCI (29%). There was no apparent relation between the MyHC isoform content of myofibers and the levels of desmin immunoreactivity at their MTJs.
## Nestin
Immunolabeling with the antibody against nestin was present in a proportion of myofibers in the human EOMs, as previously reported [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] and in contrast to myofibers in limb muscle, which are unstained with this antibody. The immunolabeling intensity with the antibody against nestin in these myofibers was generally weaker than that observed with desmin, in a given myofiber, as previously described. [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] The vast majority of MTJs (91%) in the human EOM sections treated with the antibody against nestin showed immunoreactivity with this antibody in both the global and the orbital layers, irrespective of whether the myofibers were labeled or unlabeled along the remaining of their length [fig_ref] FIGURE 4: Immunoreactivity with the antibody against nestin [/fig_ref]. The remaining proportion of MTJs in these muscle sections was unlabeled with the antibody against nestin (Figs. 4G-I). The proportion of nestin labeled MTJs varied between different specimens and between subjects. The MTJs labeled with the antibody against nestin were mostly found in myofibers containing MyHCIIa (77%), less in myofibers containing MyHCI (23%).
We asked the question whether there was a correlation between desmin and nestin at the MTJs in the EOMs. Taking into account the presence/absence of desmin or nestin along the myofibers seen in each muscle section treated with both antibodies, five distinct patterns of immunolabeling were observed
## Keratin 19
The vast majority of myofibers was unlabeled with the antibody against keratin 19 in the EOM samples examined (not shown). At MTJs, keratin 19 immunolabeling was always absent regardless of whether it was absent or occasionally present along the myofiber length (not shown). Double immunolabeling with keratin 19 and desmin showed that keratin 19 was absent at MTJs and in the remaining parts of the myofibers irrespective of whether desmin was present or absent at MTJs or along the myofibers [fig_ref] FIGURE 7: Immunolabeling with antibodies against desmin [/fig_ref].
## Vimentin
Immunolabeling with the antibody against vimentin was not observed on the muscle side of MTJs but was readily observed as strings on the tendon side of most MTJs, in both the global and orbital layers (not shown). Immunolabeling for vimentin was present in the connective tissue, blood vessels, and nerves, as described previously. [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] Vimentin labeling was absent from the myofibers, with the notable exception of extremely sporadic myofibers (less than a total of 10 myofibers in all specimens examined) that were labeled weakly with this antibody (not shown). Sequential immunolabeling with vimentin and desmin confirmed that MTJs were always unlabeled with the antibody against vimentin, irrespectively whether myofibers were labeled or unlabeled with the antibody against desmin [fig_ref] FIGURE 8: Immunolabeling with antibodies against vimentin [/fig_ref].
# Discussion
The present study was the first to thoroughly investigate the distribution of the key intermediate filament proteins desmin, nestin, keratin 19, and vimentin at MTJs in the human EOMs and to compare immunolabeling patterns with those of the lumbrical muscles. In lumbrical muscle, desmin was present both at NMJs and along the length of the myofibers, nestin, and keratin 19 were present at MTJs only, whereas vimentin was absent both at the MTJs and in myofibers, as previously reported in skeletal muscle in general. [bib_ref] Overview of the muscle cytoskeleton, Henderson [/bib_ref] [bib_ref] Nestin is expressed during development and in myotendinous and neuromuscular junctions in..., Carlsson [/bib_ref] [bib_ref] Specific and innervation-regulated expression of the intermediate filament protein nestin at neuromuscular..., Vaittinen [/bib_ref] [bib_ref] The expression of intermediate filament protein nestin as related to vimentin and..., Vaittinen [/bib_ref] In contrast, in the human EOMs: (i) a subgroup of MTJs lacked either desmin or nestin; (ii) some MTJs lacked both desmin and nestin; (iii) keratin 19 was absent from all MTJs; and (iv) the majority of the MTJs in the EOMs contained desmin and nestin, but still differed from those in skeletal muscle in general by lacking keratin 19.
## Lack of desmin at mtjs
The intermediate filament proteins desmin and nestin are highly expressed at MTJs in skeletal muscles. [bib_ref] Nestin is expressed during development and in myotendinous and neuromuscular junctions in..., Carlsson [/bib_ref] [bib_ref] Specific and innervation-regulated expression of the intermediate filament protein nestin at neuromuscular..., Vaittinen [/bib_ref] [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] [bib_ref] Desmin at myotendinous junctions, Tidball [/bib_ref] The fundamental functional role of desmin at MTJs has been proposed to be related to mechanical behavior at the junctional site, where myofibers are exposed to greatest mechanical stress and may easily be damaged. [bib_ref] Specific and innervation-regulated expression of the intermediate filament protein nestin at neuromuscular..., Vaittinen [/bib_ref] [bib_ref] Desmin at myotendinous junctions, Tidball [/bib_ref] Approximately 15% of MTJs in the human EOMs showed absence of desmin whereas the majority (85%) of the examined MTJs had desmin. The present results suggest that desmin is not of crucial importance for the stability of MTJs in the human EOMs, given that these myofibers and their MTJs had normal appearance. Furthermore, MTJs show a similar morphology in control and in desmin knockout mice, indicating that desmin might not play the functional role that it has been assigned previously. [bib_ref] Nestin is expressed during development and in myotendinous and neuromuscular junctions in..., Carlsson [/bib_ref] Given a number of particular properties of the EOMs, our results may be interpreted as that lateral force transmission may be of particular importance in myofibers lacking desmin at MTJs. Eye movements are very different from movements across a bone joint, as the eyeballs are not rigidly attached to anything and have a minimal weight. Thus, there is no weight lifting involved in muscle movement and lower force is generated during EOM muscle contraction as compared to limb muscle. [bib_ref] Specific force of the rat extraocular muscles, levator and superior rectus, measured..., Frueh [/bib_ref] [bib_ref] Measurement of contractile force of skeletal and extraocular muscles: effects of blood..., Croes [/bib_ref] On the other hand, lateral force transmission across the muscle cell membrane may put different requirements on the myofibers of the EOMs, given that they are surrounded by a very rich connective tissue bed and, in addition, a large number of myofibers have tapered ends, terminating within the muscle itself. [bib_ref] Complex threedimensional patterns of myosin isoform expression: differences between and within specific..., Mcloon [/bib_ref] [bib_ref] Organization of the orbital surface layer in rabbit superior rectus, Davidowitz [/bib_ref] [bib_ref] Internal structure of cat extraocular muscle, Mayr [/bib_ref] [bib_ref] Myofascial force transmission is increasingly important at lower forces: firing frequency-related length-force..., Meijer [/bib_ref] [bib_ref] Pedrosa Domellöf F. Composition, architecture, and functional implications of the connective tissue..., Mcloon [/bib_ref] We recently demonstrated enriched desmin immunoreactivity in the subsarcolemma of EOM myofibers in both human [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] [bib_ref] Complex correlations between desmin content, myofiber types, and innervation patterns in the..., Liu [/bib_ref] and zebrafish, [bib_ref] Pedrosa Domellöf F. Absence of desmin in myofibers of the zebrafish extraocular..., Dennhag [/bib_ref] with weak or lack of desmin immunoreactivity inside the myofibers. Such strong immunoreactivity for desmin is typically seen at NMJs or MTJs but not seen subsarcolemmally in limb muscles. [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] [bib_ref] Complex correlations between desmin content, myofiber types, and innervation patterns in the..., Liu [/bib_ref] The high concentration of desmin in the subsarcolemmal region may partially provide the required sarcolemmal strength and the connections to the underlying contractile
## Lack of desmin at mtjs is most likely not related to palisade endings in eoms
The palisade endings are found in the close vicinity of MTJs, exclusively in myofibers containing MyHCsto/I in the global layer of EOMs. [bib_ref] The fine structure of innervated myotendinous cylinders in extraocular muscles of rhesus..., Ruskell [/bib_ref] [bib_ref] The palisade endings of cat extraocular muscles: a light and electron microscope, Alvarado-Mallart [/bib_ref] [bib_ref] Palisade endings are a constant feature in the extraocular muscles of frontal-eyed,..., Blumer [/bib_ref] In the present study, we could not study the palisade endings as the tendon end of the muscle had been cut very close to the end of the myofibers and the typical morphology of palisade endings could not be ascertained. However, we have previously established that desmin is always present in myofibers containing MyHCsto/I found in the close vicinity of MTJs.
In other words, our previous studies indicate that most likely myofibers containing MyHCsto/I and which may have received palisade endings, contain desmin. [bib_ref] Complex correlations between desmin content, myofiber types, and innervation patterns in the..., Liu [/bib_ref] In the current study, approximately 96% of the myofibers labeled with the antibody against desmin along their length were also labeled at their MTJs. Altogether, the evidence available suggests that the absence of desmin at MTJs in the EOMs is not likely associated with the presence of palisade endings.
## Lack of desmin, nestin, keratin 19, and vimentin at mtjs
Desmin, nestin, keratin 19, and vimentin belong to different classes of intermediate filaments present in myofibers of skeletal muscles and have different temporal expression patterns during the course of development. [bib_ref] Overview of the muscle cytoskeleton, Henderson [/bib_ref] These intermediate filament proteins interact closely with each other during development, as illustrated by colocalization of desmin, nestin, and/or vimentin. [bib_ref] Sejersen T. Colocalization of nestin and vimentin/desmin in skeletal muscle cells demonstrated..., Sjöberg [/bib_ref] [bib_ref] Early appearance of desmin, the muscletype intermediate filament protein, in the rat..., Bignami [/bib_ref] However, in mature limb muscle, desmin is continually expressed throughout the full length of the myofibers, whereas nestin is only expressed at the NMJs and MTJs of the myofibers, [bib_ref] Nestin is expressed during development and in myotendinous and neuromuscular junctions in..., Carlsson [/bib_ref] [bib_ref] Specific and innervation-regulated expression of the intermediate filament protein nestin at neuromuscular..., Vaittinen [/bib_ref] and vimentin is not present in the myofibers. [bib_ref] Different intermediate-sized filaments distinguished by immunofluorescence microscopy, Franke [/bib_ref] Similarly, keratin 19 is transiently expressed in developing skeletal muscle [bib_ref] Transient cytokeratin expression in skeletal muscle during murine embryogenesis, Kosmehl [/bib_ref] and although it has been shown to be localized at both M-band and Z-disk in the close vicinity of the sarcolemma in mature skeletal muscle, keratin 19 is only present in trace amounts in myofibers, and it is hard to detect beyond the MTJ. [bib_ref] Cloning and characterization of cytokeratins 8 and 19 in adult rat striated..., Ursitti [/bib_ref] In the human EOMs, nestin was detected in the vast majority of MTJs and it is tempting to speculate whether the presence of nestin, to some extent, may compensate for the lack of desmin at MTJs. The present study did not address this question, but because the other two patterns (lack of both desmin and nestin or presence of desmin and absence of nestin) were observed in approximately the same proportion of MTJs, it is fair to suggest that the presence of nestin and desmin are independent of each other at MTJs in the human EOMs. Furthermore, 6% of MTJs had neither desmin nor nestin, and no MTJs examined had keratin 19 or vimentin, indicating that these MTJs lack the fundamental cytoskeletal organization proposed to be required for force transmission. The lack of all major intermediate filament proteins at MTJs in the EOMs represents a paradigm shift questioning the proposed roles of these proteins as necessary to maintain integrity and provide mechanical strength to the myofibers. [bib_ref] Overview of the muscle cytoskeleton, Henderson [/bib_ref] In summary, our results provide strong evidence that approximately one-third of the MTJs in the human EOMs lacked desmin and/or nestin and all MTJs lacked keratin [bib_ref] Three myosin heavy chain isoforms in type 2 skeletal muscle fibres, Schiaffino [/bib_ref]. Approximately 6% of MTJs lacked all the cytoskeletal proteins examined. The present findings are in line with our previous reports that lack of desmin in myofibers is not compensated by nestin in human EOMs, [bib_ref] Pedrosa Domellöf F. Intermediate filaments in the human extraocular muscles, Janbaz [/bib_ref] nor in EOMs of control or desmin knock out mice. [bib_ref] Pedrosa Domellöf F. The cytoskeleton in the extraocular muscles of desmin knockout..., Rodriguez [/bib_ref] We propose that the particular cytoskeletal organization of the MTJs may suggest adaptation to meet the functional requirements of the human EOMs, which include significant lateral force transmission, very rich connective tissue bed surrounding each individual myofiber, and myofibers that do not run the full length of the muscle.
## Changes of myhc isoforms along eom myofibers
In contrast to limb muscles in which myofibers typically contain one or two MyHC isoforms uniformly along their length, 42 the myofibers in the human EOMs contain a mixture of up to five different MyHC isoforms in a single myofiber. 14,21 Three major myofiber groups are found in the human EOMs taking into account their MyHC composition: myofibers containing MyHCI, myofibers containing MyHCIIa, and myofibers containing MyHCeom, but lacking MyHCI and MyHCIIa. [bib_ref] Myosin heavy chain isoforms in human extraocular muscles, Kjellgren [/bib_ref] [bib_ref] A novel type of multiterminal motor endplate in human extraocular muscles, Liu [/bib_ref] MyHCeom is present in EOMs of most mammals, [bib_ref] Fibre types in extraocular muscles: a new myosin isoform in the fast..., Sartore [/bib_ref] [bib_ref] Expression and identification of 10 sarcomeric MyHC isoforms in human skeletal muscles..., Mascarello [/bib_ref] and the human MyHCeom has been detected at both the gene and the protein levels. [bib_ref] Myosin heavy chain isoforms in human extraocular muscles, Kjellgren [/bib_ref] [bib_ref] Myosin heavy chain composition of muscle spindles in human biceps brachii, Liu [/bib_ref] [bib_ref] Expression and identification of 10 sarcomeric MyHC isoforms in human skeletal muscles..., Mascarello [/bib_ref] A particularly interesting finding of the current study was that almost all MTJs examined in the EOMs were present at the distal end of myofibers containing either MyHCI or MyHCIIa. MTJs were hardly ancountered in myofibers containing MyHCeom but lacking MyHCI and MyHCII. This finding provided indirect evidence of heterogeneity in MyHC composition along myofiber length in the human EOMs, and fits previous findings that MyHC content of individual myofibers varies along the myofiber length, both in human EOMs 14 and in other species. [bib_ref] The distribution of myosin heavy chain isoforms among rat extraocular muscle fiber..., Rubinstein [/bib_ref] [bib_ref] The development of longitudinal variation of Myosin isoforms in the orbital fibers..., Rubinstein [/bib_ref] The expression of MyHCeom has been shown to be limited to the endplate zone of singly innervated myofibers, which contain MyHCIIa at their proximal and distal portions in both the global and the orbital layers, in the EOMs of rabbits [bib_ref] The superfast extraocular myosin (MYH13) is localized to the innervation zone in..., Briggs [/bib_ref] [bib_ref] Changes in myosin heavy chain isoforms along the length of orbital fibers..., Lucas [/bib_ref] and rats. [bib_ref] The distribution of myosin heavy chain isoforms among rat extraocular muscle fiber..., Rubinstein [/bib_ref]
[fig] FIGURE 1: MTJs of lumbrical muscle doubled-labeled with antibodies (green) against desmin (A), nestin (D), keratin 19 (G), vimentin (J), and antibody against laminin (red; B, E, H, and K). The right column shows the merged images for each intermediate filament protein and laminin. Note the strong labeling at MTJs with the antibody against desmin (thick arrows in A and C), nestin (thick arrows in D and F), and keratin 19 (thick arrows in G and I). Notice the weak labeling with nestin at MTJs (thin arrows in D and F). Arrowheads (J-L) denote unlabeled MTJs with the antibody against vimentin on the muscle side of the MTJs. Bars = 50 μm. [/fig]
[fig] FIGURE 2: MTJs in the human EOMs labeled with the antibody against laminin. The arrowheads (A-C) denote MTJs facing the tendon. Arrow (C) denotes MTJ facing away from the tendon. Bars = 20 μm. [/fig]
[fig] FIGURE 3: MTJs in EOMs double-immunolabeled with antibodies against desmin (green in A, D, and G) and laminin (red in B, E, and H). Merged images for desmin and laminin are shown in C, F, and I. Immunolabeling with desmin is increased at MTJs in myofibers containing desmin (thick arrows in A-C) or present in the proximity of the MTJs in myofibers lacking desmin in the remaining of their length (thin arrows in D-F). Note the absence of desmin at MTJs in myofibers lacking desmin in the remaining of their length (arrowheads in G-I). Bars = 20 μm. [/fig]
[fig] FIGURE 4: Immunoreactivity with the antibody against nestin (green in A and D; red in G) at MTJs labeled with antibodies against laminin (red in B and E) or tenascin (green in H) in EOMs. Merged images for desmin and either laminin or tenascin are shown in C, F, and I. Strong immunolabeling with the antibody against nestin was present at MTJs in myofibers containing nestin (thick arrows in A-C) and in myofibers lacking nestin (thin arrows in D-F) in the remaining of their length. Note the absence of nestin in MTJs of myofibers completely lacking nestin in the remaining of their length (arrowheads in G-I). Bars = 20 μm. [/fig]
[fig] FIGURE 5: Five (I-V) different patterns of immunostaining with antibodies against desmin (green in A, K, O, and S; and gray in F) or nestin (gray in B, L, P, and T; and green in G) at MTJs in myofibers containing MyHCIIa (A-R) and MyHCI (S-V) in the EOMs. Merged images of desmin and laminin are shown in C, M, Q, and U. Merged image of nestin and laminin is shown in H. Bars = 20 μm. for nestin in approximately 6% of MTJs and the remaining portion of the myofiber (Figs. 5K-N; Fig. 6III); (IV) absence of both desmin and nestin in 6% of MTJs and along the myofibers (Figs. 5O-R; Fig. 6IV); and (V) absence/weak immunolabeling with desmin but presence of nestin at the MTJs only (approximately 5%) and absence of desmin and nestin along the myofiber (Figs. 5S-V; Fig. 6V). [/fig]
[fig] FIGURE 6: Schematic illustration of the major patterns of immunoreactivity regarding desmin and nestin at MTJ and in the remaining portion of the myofiber present in the muscle sections of the human EOMs. I: Desmin and nestin were present in both MTJs and the remaining portion of the myofibers. II: Desmin and nestin were present at MTJs but nestin was lacking in the remaining portion of the myofibers. This pattern is similar to that of myofibers and their MTJs in skeletal muscle in general. III: Desmin was present in both the MTJs and along the myofibers whereas nestin was totally absent. IV: Desmin and nestin were absent at MTJs and along the remaining of the myofibers. V: Only nestin was present at the MTJs whereas neither desmin nor nestin were present in the remaining portion of the myofiber present in the muscle section. [/fig]
[fig] FIGURE 7: Immunolabeling with antibodies against desmin (green in A, E, and I), laminin (red in B, F, and J) and keratin 19 (gray in D, H, and L) in EOMs. Merged images of desmin and laminin are shown in C, G, and K. Keratin 19 was absent from MTJs and the remaining parts of the myofibers containing desmin along their length (thick arrows in A-D) or only in the vicinity of the MTJs (thin arrows in E-H), or in myofibers completely lacking desmin (arrowheads in I-L). Bars = 20 μm.apparatus, to ensure lateral force transmission in the EOMs.34 [/fig]
[fig] FIGURE 8: Immunolabeling with antibodies against vimentin (green in A, B, D, and E) and desmin (gray in C and F) at MTJs identified with the antibody against laminin (red in B, C, E, and F). Note that vimentin was absent from the MTJs and along the myofibers either containing (C) or lacking (F) desmin. Bar = 20 μm. [/fig]
|
A High-Speed and Low-Offset Dynamic Latch Comparator
Circuit intricacy, speed, low-offset voltage, and resolution are essential factors for high-speed applications like analog-to-digital converters (ADCs). The comparator circuit with preamplifier increases the power dissipation, as it requires higher amount of currents than the latch circuitry. In this research, a novel topology of dynamic latch comparator is illustrated, which is able to provide high speed, low offset, and high resolution. Moreover, the circuit is able to reduce the power dissipation as the topology is based on latch circuitry. The cross-coupled circuit mechanism with the regenerative latch is employed for enhancing the dynamic latch comparator performance. In addition, input-tracking phase is used to reduce the offset voltage. The Monte-Carlo simulation results for the designed comparator in 0.18 m CMOS process show that the equivalent input-referred offset voltage is 720 V with 3.44 mV standard deviation. The simulated result shows that the designed comparator has 8-bit resolution and dissipates 158.5 W of power under 1.8 V supply while operating with a clock frequency of 50 MHz. In addition, the proposed dynamic latch comparator has a layout size of 148.80 m × 59.70 m.
# Introduction
Analog-to-digital converters (ADC) have become a significant element driving the semiconductor industry over the past few years. Increased integration of different functional blocks within a single chip makes ADCs more conventional and they are able to provide high speed with low power dissipation. In addition, some features of ADCs like small size processes, low power indulgences, and a reduced propagation delay make them more acceptable to the semiconductor industry. However, it is not straightforward to scale down transistor dimensions, as it requires high channel doping, gate-induced drain leakage, and band to band tunneling across the junction. The difficulty of short channel effects also needs to be controlled . Moreover, analog circuit design happens to be more complex to carry out the necessity of reliability, where supply voltages need to be decreased according to the small dimensions of the transistors [bib_ref] Low power/low voltage high speed CMOS differential track and latch comparator with..., Fayomi [/bib_ref]. All these concerns apply to the most usable representative of the ADCs: the comparator.
The comparator is the key building block in the design process for ADCs. The comparators measure the smallest voltage differences in ADC's inputs, resolving the performance and the precision of any ADCs. An application that requires digital information recovery from analog signals, such as I/O receivers and radio frequency identification (RFID) memory circuits, widely uses high performance comparators to intensify a little input voltage to a big voltage level [bib_ref] Beyond the WIFI: introducing RFID system using IPV6, Rahman [/bib_ref] [bib_ref] Designing a ring-VCO for RFID transponders in 0.18 m CMOS process, Jalil [/bib_ref]. Moreover, digital logic circuits can detect these signals within a short period. Therefore, a faster and precisionmaking comparator requires high gain and high bandwidth [bib_ref] High speed current dq PI controller for vector controlled PMSM drive, Marufuzzaman [/bib_ref] [bib_ref] Design techniques for highspeed, high-resolution comparators, Razavi [/bib_ref].
Several structures of high-speed comparators exist, such as the multistage open loop comparator, the preamplifier latch comparator, and the regenerative latch comparator. Among the different structures, high resolution and high speed can be obtained easily by using the multistage open loop comparator. On the other hand, the latch-type comparator is the most usable one in the abovementioned applications due to its high-speed and low power consumption features. Latch-type comparators are able to accomplish decisions more rapidly with no static power indulgence and strong positive feedback [bib_ref] The behavior of flip-flops used as synchronizers and prediction of their failure..., Veendrick [/bib_ref]. Moreover, latch-type comparators are able to generate high gain in regeneration mode due to their positive feedback features. However, to design circuits for low voltage operations capable of decreasing the dynamic range of the inputs and the corresponding differential process [bib_ref] Low power/low voltage high speed CMOS differential track and latch comparator with..., Fayomi [/bib_ref] [bib_ref] A 65-fJ/conversion-step 0.9-V 200-kS/s rail-to-rail 8-bit successive approximation ADC, Hong [/bib_ref] , the power dissipations in rail-to-rail operations are often increased. Consequently, the most vital limitations of the dynamic latch comparator are the kickback noises generated by high transmission currents [bib_ref] A high-speed highresolution latch comparator for pipeline Analog-to-Digital Converters, Wang [/bib_ref]. In addition, employing a transmission gate can also induce spikes at the differential input voltage signals, which affects the performance of the dynamic latch comparator due to random noise, input offset voltages, and component mismatch.
In 2013, Zhu et al. designed an ultra-high-speed latched comparator with a controlled amount of positive feedback cell [bib_ref] A high-speed latched comparator with low offset voltage and low dissipation, Zhu [/bib_ref]. However, in this design, transmission gate switches are employed to reduce the power dissipation and the effect of charge injection. In other research work, Kapadia and Gandhi implemented a dynamic latch comparator using the CMOS charge-sharing concept, which also employed an extra buffer stage [bib_ref] Implementation of CMOS charge sharing dynamic latch comparator in 130 nm and..., Kapadia [/bib_ref]. To obtain better performances, the uniqueness of the comparator, such as offset, input common mode range, propagation delay, and power dissipation, has been analyzed in both 130 nm and 90 nm technologies. Nevertheless, employing a buffer stage increases the power dissipation and the chip size. Singh and Gupta proposed a wideband flipped voltage follower (FVF) circuit using an inductive-peaking based bandwidth, which is able to engender low output impedance at high frequency [bib_ref] High frequency flipped voltage follower with improved performance and its application, Singh [/bib_ref]. On the other hand, this FVF design consumes more power. In 2013, Bhumireddy et al. introduced a novel latch-based comparator for successive approximation ADC with sub-32 nm double gate MOSFETs (DG-MOSFET) [bib_ref] Design of low power and high speed comparator with sub-32-nm double gatemosfet, Bhumireddy [/bib_ref]. In this design, the regeneration time of the latch is enhanced by employing an extra positive feedback, which in turn increases the offset voltage and the propagation delay. An amplifier is employed before the latched comparator, which decreases the offset voltages caused by the device mismatch. To achieve high gain to the output signal of the amplifier, a transmission gate can be utilized between the preamplifier and the latch, which in turn controls the signal path by using the insertion trend. Conventionally, a latch proceeded by preamplifier stages is utilized to employ a faster and accurate comparator [bib_ref] Design techniques for highspeed, high-resolution comparators, Razavi [/bib_ref] [bib_ref] A new architecture for area and power efficient, high conversion rate successive..., Dabbagh-Sadeghipour [/bib_ref]. As a result, more area and power are dissipated by employing the preamplifier stages that also border the frequency bandwidth of the input signal. Miyahara et al. proposed a dynamic comparator with a selfcalibration feature based on output averaging [bib_ref] A low-noise self-calibrating dynamic comparator for high-speed ADCs, Miyahara [/bib_ref]. Moreover, in this design, a charge pump is required to regulate the corresponding input-referred offset voltage, making the approach inefficient. However, the involvement of this charge pump circuit limits its accuracy. In 2007, Verma and Chandrakasan proposed a novel structure for an offset compensated latch comparator. Nevertheless, convoluted timing necessities and a high number of offset annulment capacitors limit using this comparator in high-speed applications [bib_ref] An ultra low energy 12-bit rate-resolution scalable SAR ADC for wireless sensor..., Verma [/bib_ref].
In this paper, a dynamic latch comparator is proposed based on differential pair input stages and one cross-coupled stage. Moreover, the proposed comparator is able to provide more high resolution and high speed with low power dissipation than conventional dynamic comparators at low supply voltage. The design is implemented in a Cadence Virtuoso 0.18 m CMOS process. The prelayout and postlayout simulation results prove that the circuit topology makes it applicable to work at very low supply voltage applications.
## Design of proposed dynamic latch comparator
All the transistors should be properly matched in layout and biased in the saturation region to make a dynamic latch comparator more vigorous against mismatch and process variations. A fully differential dynamic latch comparator based on cross-coupled differential pairs is shown in , which is based on the design of "Lewis-Gray" dynamic comparator [bib_ref] 10-bit, 20-MS/s, 35-mW pipeline A/D converter, Cho [/bib_ref].
In and M5 is utilized in the proposed topology to produce positive feedback that allows the output to switch faster. The switching mainly depends on the inputs VINP and VINN. In addition, the output signals SWM and SWP are unchanged (combine with the above). When the voltage VINP is bigger than VINM, the drain voltage of M0 will fall at a faster rate than the drain voltage of M7. Once the positive feedback from the cross-coupled NMOS transistors M5 and M3 kicks in, the node n1 will drop even faster and pull node VON low creating a logic low at the RS latch and the output Q = low. The overall circuit performance is depending on the devices size and dimensions which are shown in [fig_ref] Table 1: Transistor dimensions used in this proposed topology [/fig_ref].
In this research, no offset cancelling techniques are introduced. However, there is a tradeoff between high speed and high accuracy (e.g., 8 bits) because of MOS device mismatches [bib_ref] A high-speed CMOS comparator with 8-b resolution, Yin [/bib_ref]. Effects of offset voltage can be reduced but cannot be circumvent completely. The total offset voltage of the comparator has the well-known dependency on the mismatch of the threshold voltage Δ , load resistance Δ ,
[formula] = Δ + − 2 ( Δ + Δ ) .(1) [/formula]
The Scientific World Journal In (1), the offset voltage is dominated by the Δ , which is the mismatch of the transistor dimension, the overdrive voltage − . Moreover, the threshold voltage also has an effect in . If the common mode voltage becomes lower ( , low), the offset is found to be smaller. The effect of the mismatches of the transistors from simulations M1, M8, M9, M10, M11, M13, M14, and M15 in this topology is not very critical. Moreover, the transistors in the input differential pair and the cross-coupled NMOS are significant, as these M0, M3, M5, M7, and M12 transistors determine the overdrive of the input differential pair, which is related proportionally to the − in (1).
# Results and comparison
The proposed dynamic latch comparator circuit has been verified using the SPECTRE simulator (CADENCE). The CADENCE Virtuoso in a 0.18 m CMOS process parameter is utilized in this design. The simulated behavior of the comparator is illustrated in [fig_ref] Figure 2: Transient simulation of the comparator input signals, VLATCH signal, and output signals [/fig_ref]. It is observed from [fig_ref] Figure 2: Transient simulation of the comparator input signals, VLATCH signal, and output signals [/fig_ref] that with a 2 mV positive step size for the input VINP and keeping VINN fixed at 0.7 V, the proposed dynamic latch comparator can switch successfully. In this topology, the stepping of the input signal VINP (going up and coming back down) is used to check whether there is any hysteresis or not. From the simulated results of [fig_ref] Figure 2: Transient simulation of the comparator input signals, VLATCH signal, and output signals [/fig_ref] , it is found that when VINP > VINN (VINP = 695 mV and VINN = 700 mV) on the rising edge of the VLATCH signal, the proposed dynamic latch comparator can switch successfully. Similarly, switching also happens whenever VINP < VINN during VLATCH = high. In this case, the input signals VINP and VINN values are chosen in such a way to find out the output variation for small input changes. [fig_ref] Figure 3: Propagation delay waveform between the VLATCH and SWP signal [/fig_ref] shows the propagation delay between the latch and the output signals. When input signal VINN > VINP and the VLATCH is high, output signal SWM is in the rising state and output signal SWP is in the falling state. The simulated result shows that during the rising edge of the SWP maximum propagation delay between the VLATCH and SWP signal is about 4.2 nS.
To verify the robustness of the proposed dynamic latch comparator, different increments of environmental conditions like temperature, voltage clock frequency, and so forth need to be tested. To test the process variation, all 45 corners, 3 Vcc (1.7, 1.8, and 1.9), 3 temperatures (27, 0, and 90), and 5 corners (typical, snsp, snwp, wnsp, and wnwp) are analyzed [bib_ref] Design techniques for highspeed, high-resolution comparators, Razavi [/bib_ref] The Scientific World Journal for the proposed design. In addition, different stepping sizes of VINP and a fixed VINN at 0.7 V are analyzed. [fig_ref] Figure 4: Corner analysis of the comparator input signals, VLATCH signal, and output signals [/fig_ref] shows that at different variations of Vcc and temperature the proposed dynamic latch comparator has been switching properly. Whenever VNP > VINN and VLATCH is high, SWM and SWP switch properly. Similarly, switching also happens whenever VINP < VINN during VLATCH = high. The chip layout is shown in , where the chip occupies an area of 148.80 m × 59.70 m. In this layout, all the transistors are placed symmetrically to reduce mismatch in the parasitic capacitance.
The postlayout Monte-Carlo simulation results for 100 runs are shown in , which found that a higher offset value was obtained at a sampling frequency of 50 MHz using VDD 1.8 V with the overdrive voltage of 3.44 mV, which corresponds to 0.5 LSB at 8-bit precision. [fig_ref] Table 2: Comparison study of the proposed latch comparator performance [/fig_ref] summarizes the proposed latch comparator performance with recently published research works. Compared to the research works of [bib_ref] A low-voltage low-power comparator with currentcontrolled dynamically-biased preamplifiers for DCM buck regulators, Lee [/bib_ref] , the proposed dynamic latch comparator has less offset voltage. The comparator of [bib_ref] A low-voltage low-power comparator with currentcontrolled dynamically-biased preamplifiers for DCM buck regulators, Lee [/bib_ref] [bib_ref] Low-voltage, high-speed CMOS analog latched voltage comparator using the flipped voltage follower..., Achigui [/bib_ref] works only in 20 MHz sampling rate, whereas this design is able to run in 50 MHz sampling rate. Moreover, propagation delay is significantly lower than the research works published in [bib_ref] Low-voltage, high-speed CMOS analog latched voltage comparator using the flipped voltage follower..., Achigui [/bib_ref]. Apart from the research work of [bib_ref] A 7-bit 50MS/s single-ended asynchronous SAR ADC in 65nm CMOS, Xu [/bib_ref] , this design has more resolution (8 bits instead of 7 bits).
To compare the performance of different comparators, a well-known figure of merit (FOM) is used. Therefore, in this research, to measure the performance of the design, the FOM is calculated using the following equation:
[formula] FOM = 2 * ,(2) [/formula]
The Scientific World Journal 7 where is the power dissipation, is the number of bits (resolution), and is the sampling frequency of the comparator. From the comparison study of different recently published works as shown in [fig_ref] Table 2: Comparison study of the proposed latch comparator performance [/fig_ref] , the proposed dynamic latch comparator has the lowest FOM energy dissipated per conversion among all the recently published research works.
In this research, the proposed dynamic latch comparator is able to work for 8-bit resolution, whereas the resolution of [bib_ref] Low-voltage, high-speed CMOS analog latched voltage comparator using the flipped voltage follower..., Achigui [/bib_ref] is found to be 12 bits. To improve the overall resolution of the proposed dynamic latch comparator reduced offset voltage, different transistor sizing for reducing mismatch and layout methods in the realized chip can be implemented. In this research, V T mismatches is reduced by using big transistors like M0, M3, M5, and M7. Mismatch among the devices can be found from the following equation:
[formula] Δ = √ .(3) [/formula]
As a result, overdrive voltage has been increased and − is decreased. Therefore, to improve the resolution of the proposed design, tail current of M12 can be reduced. However, the proposed design has removed the preamplifier stage and employed a dynamic latch, resulting in significant power saving, especially in flash and pipelined A/D architectures and RFID transponders.
# Conclusion
A novel high-speed, low power, and low-offset dynamic latchtype comparator method is presented in this research work. The proposed design does not use any preamplifier stages before the latch stage, which reduces the power dissipation and the area dramatically. The corner analysis and the Monte-Carlo simulation results clearly reveal that the dynamic latch comparator is able to switch properly with different input stepping sizes. Though the proposed design has 8-bit resolution with 50 MHz sampling rate it consumes much lower power. Moreover, the comparison study shows that the novel design is able to operate at a higher clock frequency of 50 MHz with offset voltage 3.44 mV and propagation delay 4.2 nS in 1.8 V supply voltage, which is better than recently published research works.
[fig] Figure 1: (a) Schematic diagram of the proposed differential pair dynamic latch comparator and (b) schematic diagram of the R-S flip-flop with digital signal PD. [/fig]
[fig] Figure 2: Transient simulation of the comparator input signals, VLATCH signal, and output signals (SWP and SWM) using Virtuoso Spectre. [/fig]
[fig] Figure 3: Propagation delay waveform between the VLATCH and SWP signal. M1, M8, M10, and M15 reset the outputs VON and VOP and nodes n1 and n2 to VDD. On the other hand, when VLATCH = high, the outputs are disconnected from the positive supply and switching current source M12 begins to conduct. M12 mainly determines the bias current of the input transistors M0, M7, M3, and M5. The cross-coupled NMOS pair M3 [/fig]
[fig] Figure 4: Corner analysis of the comparator input signals, VLATCH signal, and output signals (SWP and SWM). [/fig]
[fig] Figure 5, Figure 6: A layout design of the proposed dynamic latch comparator. Postlayout Monte-Carlo simulation result with process and mismatch variation. [/fig]
[table] Table 1: Transistor dimensions used in this proposed topology. [/table]
[table] Table 2: Comparison study of the proposed latch comparator performance. [/table]
|
Efficacy of artemether-lumefantrine and artesunate-amodiaquine as first line therapy of uncomplicated malaria in Burkina Faso, 11 years after policy change
Introduction: artemether-lumefantrine (AL) and artesunate-amodiaquine (ASAQ) are the first line therapy of uncomplicated malaria in Burkina Faso. We assessed the treatment efficacy, tolerability of these drugs 11 years following its adoption as first line treatment. Methods: in this opened randomized controlled trial carried out in 2016, participants with age over 6 months who consented to participate were randomly assigned treatment with artemether-lumefantrine or artesunate-amodiaquine and followed up for 28 days. Primary endpoint was the treatment efficacy over 28 days of follow up unadjusted by Polymerase chain reaction (PCR). Results: two hundred and eighty-one (281) participants were enrolled and the completion rate was 92.9%. No early treatment failure was found. Adequate clinical and parasitological responses were significantly higher in artesunateamodiaquine group (97% versus 85.2%, p = 0.0008). On day 28, the risk of failure was 4 times higher in AL group 20.14%, 95% CI (13-30.47) against 5.16%, 95% CI (1.91-13.54) in ASAQ group. All treatments had a similar and good tolerability profile. Conclusion: eleven years following artemether-lumefantrine and artesunate-amodiaquine adoption as first line therapy for uncomplicated malaria in Burkina Faso, artemetherlumefantrine retained fairly good efficacy even though its efficacy fell below WHO threshold of 90% considering uncorrected o utcome.
# Introduction
The cornestone of malaria treatment has been since the last fithteen years the artemisinin based combination therapy (ACTs) mainly artemether-lumefantrine and artesunate-amodiaquine. Procurement of ACTs faced severe shorted during the first five to ten years of its adoption. Currently, the procurement market has improved and ACTs This situation inevitably sustains the use of the ACTs and thus the increase in drug pressure to the used drugs (artemether-lumefantrine and artesunate-amodiaquine). As drug pressure is known to contribute to the alteration of drug efficacy , this continuous ACT use over 11 years warrants the investigation of this first line therapy efficacy; the present report is intented to provide updated efficacy data for artemether-lumefantrine and artesunate-amodiaquine in Burkina Faso to guide the national malaria control program.
# Methods
Study site: the study was carried out in two peripheral health facilities, Colsama and Sakaby during the transmission season between June and December 2016. Each health facility was staffed by two to three nurses and one midwife. They were equipped with one outpatient department and one inpatient wall. All sites were located in Bobo-Dioulasso the second capital city of Burkina Faso at 365 km from Ouagadougou. Transmission of malaria is long but seasonal occurring over 6 months and containing at least 60-80% of clinical cases.
Study participants: subjects presenting to the study health facilities with fever (axillary temperature ≥ 37.5°C) or history of fever in the previous 24 hours were screened against the following inclusion criterias in accordance with WHO guideline 2015: age at least 6 months, weight at least 5 kg, no evidence of concomitant illness, the provision of informed consent by the parents and the ability to participate in 28 days follow up, no history of antimalarial treatment within the last two weeks, no danger sign or evidence of severe malaria, P.falciparum mono infection, parasite density of 2000-200 000 parasites per µl of blood and hemoglobin ≥ 5 g/dL. During the screening process, participants satisfying the above criterias were enrolled while those who were excluded were referred to the health facility staff for standard care.
Evaluation at entry and randomization: potential participants were assigned a number (ID), interviewed on past medical history, examined and then referred to the study nurses dedicated only to treatment allocation. Participants were randomly allocated to oral three days treatments based on a computer generated code provided by a staff not involving in the patient's evaluation. The treatment was either artemether-lumefantrine or artesunate-amodiaquine. The dosage was given on weight basis as per national malaria control program recommendation as this study is intented to evaluate the efficacy of first line treatments of malaria as implemented by the national malaria control program: (a) Artemether-lumefantrine (Coartem® Novartis administrated in tablet containing 20 mg of artemether plus 120 mg of lumefantrine); (b) artesunate-amodiaquine as given as tablet containing 67.5 mg plus 25mg respectively). The study as per its objectives is to inform straight away, it was not blinded to the study staff and patient. The treatment was administrated as DOT (directly observed treatment) and participants were retained for 30 minutes and in case of vomiting within this time period, a replacement dose was given. Any repeated vomiting led to the participant exclusion from further follow up and referral to health facility staff for standard care. Children presenting with hemoglobin less than 10 g/dL were treated according to the Integrated management of childhood Illness guidelines with ferrous sulfate for 14 days and anthelmintic treatment, if appropriate.
Follow up and classification of treatment outcome: patients were asked to return to the study clinics on day 1,2,3,7,14,21 & 28 and at any time they were feel illness. Field workers visited those who failed to return to the clinics and each visit consisted on standard case report completion, physical examination, a finger pick for thin and thick blood smear. Blood smears were read to assess the parasite density and gametocytes. Patients were followed up for 28 days and Gametocytes were recorded as present or absent. Hemoglobin level was measured from finger prick blood sample using a portable photo spectrophotometer (HemoCue).
Statistical analysis: the study was designed to check the efficacy of first line treatment of malaria in Burkina Faso, artemetgherlumefantrin and artesunate-amodiaquine. We anticipated that 136 participants will be needed in each group to assess the efficacy of artemether-lumefantrin versus artesunate-amodiaquine assuming 90% uncorrected cure rate in artesunate-amodiaquine group (detecting a difference of 15%) with 80% and significance level of 0.05 accounting for 20% of lost to follow up. Baseline characteristics of enrolled participants will be summarized; participants in each category (anemic or not) will be presented as proportion and compared between drug regimen. Categorical variables will be compared dosing chi square test while continuous are compared using either independent or dependant t-test. Kaplan Meyer estimate will be used to calculate the risk of treatment failure using strick formula and patient were censored the day they left the study.
## Ethics approval and consent to participate
# Results
Baseline characteristics of enrolled participants: a total of 281 children were enrolled over the study period and treated with either AL or ASAQ for three days. About 43% of the participants treated with AL and 56.7% of those treated with ASAQ were male. Children 5-9 years were more represented. Mean hemoglobin was 10.82±1,2 g/dL and 10.92±1,85 g/DL respectively in AL and ASAQ groups. Other characteristics are described inand.
Treatment efficacy: we did not record any early treatment failure over the 28 days follow up period in all treatment groups. Late clinical failures were recorded in all groups: 5 cases 3.9% in AL and 1.5% in ASAQ group meaning that patients treated with AL were 2.6 more likely to have late clinical failure compared to patients treated with ASAQ (p = 0.2). All cases were above 59 months except the failure in the AL group which occurred in 16 months old boy. On the other hand, late parasitological failures were more common, 10.9% and 1.5% respectively in AL and ASAQ groups; patients treated with AL were almost 8 times likely to experience late parasitological failure than those treated with ASAQ (p = 0.0016). Finally, treatment was significantly successful in ASAQ group .
Clinical symptoms: clinical symptoms were common at entry for all two drugs regimen. Weakness was observed in approximately 58% of participants in each arm but vomiting was more frequently observed with ASAQ (78.83% versus 65.94%, p = 0.1).
Treatment tolerability: over the study period, we found no serious adverse event. The registered adverse events were mild to moderate and none required active treatment or intervention's discontinuation.
Common adverse events were weakness in either group. No pruritus or nausea were found.
# Discussion
This clinical trial was a study that is part of routine monitoring of first Thus, we can expect these drugs to retain some efficacy for a while.
In this study, all treatments doses were directly observed and this may have contributed to reach such high efficacy level.
## What this study adds
Overall these drugs remain effective even though the uncorrected efficacy is below the threshold of more than 90% recommended by the WHO; This study rises the discussion around the Polymerase Chain Reaction correction to consider the drug efficacy in real life.
## Competing interests
The authors declare no competing interests.
# Authors' contributions
# Acknowledgments
This research was conducted as part of the routine National Malaria Control Program Monitoring and Evaluation of first line therapy efficacy. We thank the nurses and the laboratory technicians who collected the data. We thank also the administration of the health district of Do and health facility of Colsama and Sakaby for their strong collaboration. We are grateful for all parents and children who participated in this study. almost all patients were apyretic on day 2 and 3 for artemether-lumefantrine treated participants and very few remained on day 3
# Tables and figures
|
Mining Biological Pathways Using WikiPathways Web Services
WikiPathways is a platform for creating, updating, and sharing biological pathways[1]. Pathways can be edited and downloaded using the wiki-style website. Here we present a SOAP web service that provides programmatic access to WikiPathways that is complementary to the website. We describe the functionality that this web service offers and discuss several use cases in detail. Exposing WikiPathways through a web service opens up new ways of utilizing pathway information and assisting the community curation process.
# Introduction
WikiPathways is an online resource for biological pathway information and a platform for community-based curation [bib_ref] WikiPathways: pathway editing for the people, Pico [/bib_ref]. Currently, WikiPathways contains more than 600 pathways, representing various species from bacteria and fungi to plants and animals. New species are being added on demand, and pathways are being created and edited almost daily. These pathways combine different types of biological knowledge, including protein interactions, metabolic reactions, annotations from gene, protein and metabolite databases and references to scientific literature. Using a Java-based pathway editor, researchers can create, edit and annotate pathways directly on the website [bib_ref] Presenting and exploring biological pathways with PathVisio, Van Iersel [/bib_ref].
WikiPathways is accessed by the biology community mainly via a wiki-style website. In addition to the website, we recently implemented a web service that provides programmatic access to WikiPathways (figure 1). This makes it easier to integrate biological pathways in existing applications and provides a framework for pathway-centered analysis of experimental data. In contrast to other pathway resources with managed curation teams (e.g., KEGG [bib_ref] KEGG: kyoto encyclopedia of genes and genomes, Kanehisa [/bib_ref] and Reactome [bib_ref] Reactome knowledgebase of human biological pathways and processes, Matthews [/bib_ref] or focused on distributing and querying pathway information (e.g., Pathway Commons [5]), WikiPathways is a primary source for richly annotated pathway content open to community curation. With the addition of web services, users have expanded access and advanced methods that uniquely apply to WikiPathways content. In this article, we discuss the web service functionality in more detail and highlight several use cases.
Supplementary data, including full documentation, example client implementations, and source code, are available at http://
# Results and discussion
## Web service functionality
The web service provides an interface to WikiPathways that can be accessed through the Simple Object Access protocol (SOAP), and the data structure and available functions are described in the Web Service Description Language (WSDL). Both SOAP and WSDL are widely supported standards. The web service provides access to all pathway information on WikiPathways in several different forms. Complete pathways can be downloaded in XML format (GPML) or a plain text file listing the biological entities and their identifiers. Image versions of a pathway can be retrieved in several graphics formats, including rasterized formats (e.g., Portable Network Graphics (PNG)) and vector graphics formats (e.g., Scalable Vector Graphics (SVG) and Portable Document Format (PDF)). Additionally, color information can be specified to highlight specific elements of the pathway (e.g., to color protein entities according to their measured expression). Individual interactions between biological entities that are defined in the pathways can be retrieved separately. Furthermore, information related to community-based curation, such as revision history and recently edited pathways, can be queried. A full list of available functions can be found in the supplementary data.
An index of all pathways is maintained using the Apache Lucene library [6]. This feature makes it possible to perform advanced search queries through the web service. All textual information on the pathway is included in the index to allow for keyword searches. In addition, the index uses synonym databases [bib_ref] Presenting and exploring biological pathways with PathVisio, Van Iersel [/bib_ref] to cross-reference between various biological databases and the entities on the pathways. This allows for queries to find all pathways for any given biological identifier, regardless of the identifier system that is used to annotate the pathway. For example, when the query is an Affymetrix probeset identifier, all pathways containing genes that map to that probeset will be returned.
The web service also allows client software to publish information to WikiPathways. New pathways can be uploaded and pathways can be modified or labeled according to quality standards. This enables scripts to perform quality monitoring and notification to assist the manual community-based curation process. This concept has already been successfully applied to other wiki's, such as Wikipedia [bib_ref] A gene wiki for community annotation of gene function, Huss [/bib_ref]. To prevent scripts from systematically overwriting pathways with invalid data, write access to the web service is restricted to a subset of user accounts. Users who require write access for their script can request it from the WikiPathways site administrators.
To assist programmers in building applications that use the WikiPathways web service, several toolkits and programming libraries exist. Libraries to handle SOAP requests and responses are available for practically any programming language. Additionally, several bioinformatics tools, such as Taverna and GenePattern, support plugging in SOAP web services by writing only little or no extra code. This makes it easy to integrate WikiPathways in existing pipelines. To facilitate working with the GPML pathway format, we maintain an open-source Java library that provides a high-level API to process GPML. This library contains methods to read and write pathways in several file formats and to modify information in the pathway. Furthermore, it provides an object-oriented interface to the WikiPathways web service, including support for caching downloaded pathway information locally to improve performance. For each described use-case, example code is provided that demonstrates the use of available toolkits and libraries. The supplementary data include a list of useful libraries in several programming languages.
## Web applications
The web service can be used to build web applications that provide end-users access to specific WikiPathways functionality. Research groups can build a website that queries, processes and presents information from WikiPathways in a fully customized way. As an example, we implemented two web applications, each highlighting a unique functionality of the web service. The first application is an improved search application with more advanced functionality, available at http://search.wikipathways.org. The default WikiPathways search is based on a Google Custom Search Engine and allows searching through all the text on pathway pages, but doesn't use the biological context that is defined in the GPML format. The improved search application uses the web service functions that query the Lucene index and allows users to perform searches with biological context, such as using biological identifiers and filtering by species. Furthermore, the results are presented as thumbnail images, making it easier to choose the relevant result.
The second example is a web application that demonstrates integration of pathway information with other types of data. This application visualizes gene expression information from ArrayExpress Atlas [8] on a WikiPathways pathway. ArrayExpress Atlas is a curated set of gene expression datasets that are publicly available. In this example, the user can specify a pathway from WikiPathways and a set of experimental conditions defined in ArrayExpress Atlas. First, all gene identifiers on the pathway will be mapped to Ensembl using the synonym database. The resulting Ensembl identifiers are passed to the ArrayExpress Atlas web service, which returns the corresponding experiments and p-values for the differentially expressed genes. Second, the WikiPathways web service will be used to download a colored version of the pathway image that will be displayed to the user (figure 2). This application can be used to get a quick overview of how known pathway interactions relate to observed gene expression in a given experimental condition and is available at http://atlas.wikipathways.org. Currently, 716 of the 1000 experiments on ArrayExpress Atlas can be mapped onto at least one of the WikiPathways pathways, representing 9357 unique genes affected in one or more experiments across 7 different species.
Research groups are encouraged to build their own client-side web applications based on the WikiPathways web service and our open source libraries. This could include applications that present WikiPathways content in a customized way or integrate pathways with other data. For example, research groups focused on metabolic pathways could create an application that presents the pathways in combination with detailed enzymatic information, while the genetics community could create a web application that combines polymorphism information with the genes from a pathway.
## Assisting community-based curation
The biology community manually moderates the content on WikiPathways. Being able to quickly respond to mistakes as well as acts of vandalism is an important aspect of community-based curation. Many types of erroneous data can be identified automatically. For example, wrongly annotated gene, protein, or metabolite entities are identified by a simple computer script. The web service enables us to create ''bots'' that continually watch the pathway content for errors and notify curators when incorrect information is introduced. Currently, various types of bots run on . WikiPathways can be accessed by end-users from the wiki-style website. In addition, the WikiPathways web service provides a programmatic interface that can be used in many programming languages, including R, python, Java and perl and in workflow tools such as Taverna. Using this interface, new pathway analysis tools can be built and existing bioinformatics tools can be extended with pathway-based functionality. doi:10.1371/journal.pone.0006447.g001 a daily basis and check for incorrect database annotations, missing literature references, or incorrectly defined interactions. Each bot generates an HTML report that provides statistics and an overview of pathways that need to be corrected. By monitoring these reports over time, we can assess the effectiveness of community-based curation.
Community involvement in pathway curation on WikiPathways could be further stimulated by using the web service to build applets that can be included in any webpage or desktop. These applets could display user-specific information, such as recent changes on pathways that have been edited by the user, recent discussion items, or listing other users that are interested in similar pathways. Users could install this applet on their own webpage or on their desktop, for example using Google Gadget interface [10].
## Querying interactions in cytoscape
WikiPathways includes interactions between biological entities that can be represented as interaction networks. For example, metabolic reactions or protein activation events that are defined in a pathway can be viewed as binary interactions that form a graph. This enables network-based analysis [bib_ref] Protein networks in disease, Ideker [/bib_ref] and integration with various datasets, such as protein-protein interactions [bib_ref] A new, structurally nonredundant, diverse data set of protein-protein interfaces and its..., Keskin [/bib_ref] [bib_ref] An in Vivo Map of the Yeast Protein Interactome, Tarassov [/bib_ref]. Cytoscape [bib_ref] Cytoscape: a software environment for integrated models of biomolecular interaction networks, Shannon [/bib_ref] is a widely used tool for visualization and analysis of biological networks. The core Cytoscape functionality can be extended by external developers via a plug-in mechanism. We used this mechanism to build a plug-in that allows Cytoscape to work with GPML pathways [bib_ref] Presenting and exploring biological pathways with PathVisio, Van Iersel [/bib_ref]. We recently extended the plug-in with functionality from the WikiPathways web service. This new functionality makes it possible to search for pathways on WikiPathways directly from within Cytoscape, and these pathways can be loaded as an interaction network and then analyzed and integrated with experimental data [bib_ref] Integration of biological networks and gene expression data using Cytoscape, Cline [/bib_ref]. Furthermore, existing networks in Cytoscape can be expanded with interactions defined in WikiPathways. Network analysis of biological pathways can also be a useful tool to augment pathways with knowledge from various resources. For example, cross-talk between biological pathways can be identified by projecting pathways on large protein-protein interaction networks [bib_ref] A Global Pathway Crosstalk Network, Li [/bib_ref]. The Cytoscape plug-in is a typical example of how the web service can be used to enrich external analysis tools with pathway content.
Future versions of WikiPathways will allow the user to define the semantic meaning for each interaction in a pathway. This can be used to improve the web service with functions that query and filter interactions based on this semantic information. For example, queries, such as ''show me all proteins that inhibit phosphorylation of protein X'' could be performed. Including semantics also opens new possibilities for analysis in Cytoscape, for example the signaling pathway impact factor analysis method [bib_ref] A Novel Signaling Pathway Impact Analysis (SPIA), Tarca [/bib_ref] , which can use the distinction between activation or inhibition interactions.
## Integration in bioinformatics workflows
A primary difficulty in bioinformatics is integrating erratically formatted data from different resources [bib_ref] A Novel Signaling Pathway Impact Analysis (SPIA), Tarca [/bib_ref] [bib_ref] Towards a cyberinfrastructure for the biological sciences: progress, visions and challenges, Stein [/bib_ref]. The development of web services for biological databases makes this easier and enables usage of workflow tools, such as Taverna [bib_ref] Taverna: a tool for the composition and enactment of bioinformatics workflows, Oinn [/bib_ref]. Taverna provides a framework for creating workflows that integrate and process different types of data. For instance, a workflow could be used to integrate microarray experiment data with quantitative trait loci to filter for relevant genes [bib_ref] A systematic strategy for large-scale analysis of genotype phenotype correlations: identification of..., Fisher [/bib_ref]. Knowledge represented as a pathway is especially suited for integration with other biological data, since it typically covers multiple levels of the biological information hierarchy. Transcriptomics, proteomics, or metabolomics data can be integrated with pathway information to aid in the understanding of the underlying biological mechanisms. We created several Taverna workflows based on the WikiPathways web service, such as a workflow that finds relevant pathways based on a list of over-expressed genes, proteins, or metabolites. These workflows can be downloaded from http://www.myexperiment. org/packs/30. Other possible workflows that could be implemented in Taverna include cross-species comparisons of pathway information, or batch visualization of expression data.
The WikiPathways web service could be used as framework for building data analysis tools that make use of pathway information. Pathways can be used for visualizing experimental data in a biological context [bib_ref] GenMAPP 2: new features and resources for pathway analysis, Salomonis [/bib_ref] , finding relevant biological mechanisms and improving statistical power of an experimental analysis [bib_ref] MAPPFinder: using Gene Ontology and GenMAPP to create a global geneexpression profile..., Doniger [/bib_ref]. The WikiPathways web service provides all of the functionality necessary to perform these methods. Pathways can be downloaded as images where the pathway components can be given user defined colors, allowing for experimental data analysis. Genes, proteins and metabolites are linked to several identifier systems, providing information to perform enrichment analyses.
## Integration with online databases
Pathways typically consist of different types of biological entities, such as genes, proteins and metabolites. For each entity type, different biological databases are available, and each presents unique information about the entities in a different way. With the WikiPathways web service, we aim to encourage database developers to integrate pathway information into their online data presentation to provide more biological context. WikiPathways links pathway components to over 50 supported biological databases, including Entrez Gene, Ensembl, Affymetrix, ZFIN, TAIR, and ChEBI. Tools that use information from any of these databases can use the web service to retrieve relevant pathway information per biological entity. For example, the list of pathways that contain a given gene identified by an Ensembl ID could be retrieved and displayed on the Ensembl web page for that gene or any website indexed by Ensembl identifiers. The website could display the pathway name, URL and/or thumbnail image of the pathway. A similar approach could be taken by metabolic databases (e.g., PubChem, ChEBI, or ChemSpider), protein databases (e.g., UniProt), model organism databases (e.g., MGI, ZFIN, WormBase, FlyBase, or TAIR), measurement platforms (e.g., Affymetrix, Illumina, or Agilent), or even literature databases, such as PubMed.
Future versions of WikiPathways will support the export of pathways in BioPAX format. This will make it easier to integrate WikiPathways with other pathway databases and resources, such as PathwayCommons. This functionality will be available in the web service, so that integrated pathway resources can easily keep the pathway information from WikiPathways up-to-date.
# Conclusions
The WikiPathways web service provides an interface for programmatic access to community-curated pathway information. It provides a flexible framework for building or extending tools that use pathway information from WikiPathways. The web service can be used by software developers to build or extend tools for analysis and integration of pathways, interaction networks and experimental data. The web services are also useful for assisting and monitoring the community-based curation process. By providing this web service, we hope to help researchers and developers build tools for pathway-based research and data analysis.
[fig] Figure 2: Web application that integrates pathways with gene expression information. Pathway and gene expression information are retrieved from the WikiPathways and ArrayExpress Atlas web services respectively. A Java servlet integrates this information and publishes it to an interactive web application. In this web application, users can view the information on an interactive pathway diagram. doi:10.1371/journal.pone.0006447.g002 [/fig]
|
First report of testis‐sparing surgery for sertoliform cystadenoma: case presentation and review of literature
Abbreviations & Acronyms AE1 = Cytokeratin AE1 CAM5.2 = Cytokeratin CAM5.2 CK7 = Cytokeratin 7 CT = computed tomography EMA = epithelial membrane antigen PAX8 = Paired-box gene 8Introduction: Sertoliform cystadenoma is a very rare, benign lesion of the rete-testis difficult to distinguish from other malignancies of the testicle. Case presentation: We present the case of a 42-year-old male who presented with a right testicular mass, asymptomatic for 1 year. Clinical examination revealed a palpable, painless, and well-delimited right testicular superior pole nodule. Testicular ultrasound confirmed the nodule, whereas serum tumoral markers were normal. The patient underwent inguinal partial orchiectomy. Intraoperative excisional biopsy and frozen section pathology were performed, reporting undetermined tumoral origin with negative surgical margins. Ischemia time was 12 minutes. The final pathology report showed a Sertoliform cystadenoma of rete testis, with immunomorphology positive for AE1, CK7, and negative surgical margins. Conclusion: To our knowledge, this is the first report of testicular sparing surgery for Sertoliform cystadenoma, a very rare benign tumor of rete testis. All previously reported cases were managed by radical inguinal orchidectomy.Keynote messageTo our knowledge, this is the first report of testis-sparing surgery for Sertoliform cystadenoma, a very rare benign tumor of the rete testis. All previously reported cases were managed by radical inguinal orchidectomy. Partial inguinal orchiectomy with frozen section pathology could be an alternative to radical surgery in selected cases.
# Introduction
Testicular tumors account for 1% of male cancers and approximately 5% of urogenital malignancies, with an increasing incidence due to the widespread use of ultrasound examination. Paratesticular tumors, especially masses of the rete testis, are a rare finding and can be classified in nonneoplastic developmental pathologies (adenomatous hyperplasia, acquired or secondary cystic dysplasia), benign tumors (cystadenomas or Sertoliform cystadenoma), and malignant lesions such as adenocarcinoma of the rete testis. The first report in the literature of Sertoliform cystadenoma was made by Jones MA in 1997. All previously reported cases of sertoliform cystadenoma have been managed by radical surgery-inguinal orchiectomy.We present the first case of a Sertoliform cystadenoma in a 42-year-old male managed by conservative surgery.
## Case report
A 42-year-old male with no prior medical history presented with a palpable mass in the right testicle, asymptomatic, and without any increase in size for 1 year. The patient is married, without children. At clinical examination, a firm, painless, well-delimited superior pole right testicular lesion was objectified. Scrotal ultrasound revealed a 1.7/ 0.5 cm mass in the upper pole of the right testicle, with normal contralateral testicle and normal testicular volumes . Serum tumor markers were collected and in normal range, as following: alpha-fetoprotein (0.5 ng/ml, N = 0.89-8.78), beta-human chorionic gonadotropin (0, N < 1.4 UI/L), and lactate dehydrogenase (10 UI/L, N = 120-280 UI/L). A staging chest-abdominal-pelvic contrast-enhanced CT scan was performed which did not show any evidence of visceral or lymph node metastasis.
Given the fact that the lesion was perceived as stable by the patient for 1 year, as described during the first consultation, we suspected that a benign etiology could be possible and offered the patient the option of testis-sparing surgery. The patient was informed of the risk of radical orchiectomy in case of intraoperative aspect or frozen section pathology report highly suggestive for malignancy and salvage radical orchiectomy if the final pathology report would be conclusive for malignancy. After informed consent was given, the patient underwent right partial orchiectomy by inguinal approach. We identified and dissected the elements of the spermatic cord, after which we separated the vas deferens from the testicular vessels. After clamping the testicular vessels, we incised the tunica albuginea and proceeded with the partial orchiectomy within safety surgical margins. Frozen section assessment was demanded and revealed testicular tumor of indeterminate origin, with negative surgical margins. Thus, we closed the albuginea and reintegrated the ipsilateral testicle in the scrotum, at the same time performing an orchidopexy . The total ischemia time was 12 min.
The final pathology report showed a Sertoliform cystadenoma of the rete testis, with the tubular architecture of variable sizes, without atypical cells and rare mitosis visible. The testicular parenchyma present on the specimen was without pathological characteristics and the surgical margins were negative. Immunohistochemistry was intense positive for AE1 and CK7, negative for Inhibin and EMA, without any signs of malignancy .
The patient was discharged the same day, no immediate or long-term postoperative complications were reported. The 1-month postoperative scrotal ultrasound examination was normal, with no sign of local recurrence. Due to the fact that it is a rare histology, we performed also 2 and 6-months postoperative serum tumoral markers, which were all negative.
# Discussion
Sertoliform cystadenoma is a rare benign tumor of the rete testis, being for the first time mentioned in the literature in 1997 by Jones et al. The largest case series in the literature was presented by Paluru et al.and is comprised of 15 cases; mean age reported was 46 years, mean tumor size 1.5 cm, mean follow-up of 8.1 years with no complications and no evidence of disease recurrence. All of the patients underwent radical inguinal orchidectomy.
Several single case reports have been also identified in the literature. Sahnan et al. and Bremmer et al. describe the cases of a 19 and 66-year-old man, respectively, with a 6-month history of smooth, painless right testicular mass, stable over time, with normal serum tumoral markers in which right radical inguinal orchidectomies were performed.The first pediatric case was reported by Kacar et al.in a 6year-old boy who presented bilateral gynecomastia and left scrotal tumefaction. Serum tumoral markers were in the normal range but estradiol and testosterone levels were slightly increased (48.4 ng/mL and 20 ng/dL, respectively). No other adult patient reported in the literature displayed increased levels of sexual hormones. The patient underwent left inguinal orchidectomy and the final pathology report revealed Sertoli cell proliferations, compatible with Sertoliform cystadenoma.
The most recent case of Sertoliform cystadenoma was presented by Lahouti et al. in a 39-year-old male with a right superior pole testicular mass, 5 in which serum tumor markers were negative. The final pathology report revealed sertoliform cystadenoma with immunohistochemical examination intense positive for Inhibin, cytokeratin CAM5.2, PAX8.
Histologically, the most important differential diagnosis is made with rete testis adenocarcinoma. Sertoliform cystadenoma has been reported in patients between 19 and 84 years, Scrotal ultrasound showing a 1.7/0.5 cm mass in the upper pole of the right testicle having variable size characteristics (a few mm to 4 cm) with a predominant tubular architecture, columnar to cuboidal tumor cells with eosinophilic cytoplasm, and basally located nucleus. Moreover, adenocarcinoma of rete testis is also a rare but aggressive finding, with an incidence peak around 70 years of age. The most important morphologic criteria required for diagnosis are the involvement of testicular mediastinum with the preservation of the parietal tunica vaginalis, in the absence of other lesions.Although current recommendations are in favor of considering every scrotal mass as malignant until proven otherwise, we propose in selected cases where clinical history is in favor of benign lesion and patients desire maximum fertility preservation the alternative of organ sparing surgery. Testispreserving surgery should always be accompanied by frozen section examination, due to the fact that it has a high concordance with the final pathology result.However, all patients must be informed of the risk of immediate or salvage radical orchidectomy if the frozen section or final pathology report are consistent with malignancy.
# Conclusions
In conclusion, we report the first case of Sertoliform cystadenoma managed by conservative surgery-partial orchidectomy, without any early or late postoperative complications. Sertoliform cystadenoma remains a rare finding among intratesticular mass findings, with only 16 cases reported in the literature, all treated by radical inguinal orchidectomy.
# Funding source
No funding was received for this study.
## Registry and the registration no. of the study/trial
Not applicable.
## Editorial comment editorial comment
Editorial Comment to First report of testis-sparing surgery for sertoliform cystadenoma: case presentation and review of literature Orchiectomy is the gold standard for managing testicular tumors. According to the American Urological Association guideline, patients with a suspected malignant testicular lesion and a normal contralateral testis should undergo radical inguinal orchiectomy, instead of testis-sparing surgery (TSS). 1 TSS may be offered as an alternative for treating testicular tumors in highly selected patients having tumors of diameter 2 cm or less who wish to preserve testicular function or whose testicular function is likely to be substantially impaired by orchiectomy. However, TSS is controversial for cases that are strongly suspected to be benign. Sertoliform cystadenoma is a very rare benign tumor originating from the rete testis. Ultrasound findings show a cystic mass in many cases, but sometimes a solid mass is observed with an epididymal cyst. If malignancy cannot be completely ruled out after examination with various modalities, orchiectomy is still considered preferable. In particular, given the rarity of sertoliform cystadenoma, it may be difficult to definitely rule out malignancy preoperatively. However, when the mass is obviously benign and no malignant findings on rapid intraoperative pathology after ischemia of the testicular vessels exist, TSS with tumor enucleation with a wide enough margin may be considered. Preoperative evaluation of testicular function, including analysis of semen, serum testosterone levels, and gonadotropin levels, is advised when considering the possibility of conversion to orchiectomy.
In this case, the tumor was judged to be benign because there was no change over a long period, and the authors successfully performed TSS. If possible, preoperative testicular function should be evaluated to determine if TSS is truly necessary. Simultaneous testicular sperm extraction should be considered for patients who wish to have children and have azoospermia in the preoperative semen analysis or who are expected to have decreased testicular function. © 2021 The Authors. IJU Case Reports published by John Wiley & Sons Australia, Ltd on behalf of Japanese Urological Association
Conflict of interestThe authors disclose no conflicts of interest.Approval of the research protocol by an Institutional Reviewer BoardThis study was approved by the local ethics committee of the Croix du Sud hospital (register number 57/2021) and was conducted in line with the principles of the Declaration of Helsinki.Informed ConsentNot applicable.Conflict of interestThe author declares no conflict of interest.This 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.
|
Pilot study assessing the Rotterdam Healthy Aging Score in a cohort of HIV-positive adults in Toronto, Canada
Objective: The Rotterdam Healthy Aging Score (HAS) is a validated multidimensional index constructed from five health domains. We describe the HAS distribution in a cohort of HIV-positive adults and correlate it with health outcomes.Design: A cross-sectional pilot study of 101 adults aged at least 40 years, on suppressive antiretroviral therapy attending a tertiary HIV clinic in Toronto, Canada.Methods: Participants completed questionnaires to calculate their HAS (range 0-14). Demographics, HAS and sub-scores were compared by age and sex. The HAS was compared with results of the Fried Frailty Score, Short Performance Physical Battery score (SPPB) and measures of health utilization. Kruskal-Wallis Rank-Sum and Fisher's exact tests were used for all comparisons.Results: Median (IQR) age was 56 (50-62), 81 (80%) men and 50 (50%) born in Canada. Median (IQR) CD4 þ cell count was 574 (417-794) cells/ml. Median (IQR) HAS was 12 (10-13) with 39 (39%) achieving a score more than 12 (considered healthy aging). Younger participants experienced more depression, whereas women had greater pain. The HAS score correlated with the Fried Frailty Score (P ¼ 0.008) and trended with the SPPB Score (P ¼ 0.077). Those with the poorest HAS scores were more likely to have been hospitalized in the preceding 6 months (P ¼ 0.034).Conclusion: The HAS ranged from 5 to 14 in this cohort of older HIV adults with 39% attaining scores in the 'healthy' range. The HAS correlated with measures of physical performance and health utilization. Further validation of an objective outcome in HIVpositive patients will facilitate evaluation of interventional studies to improve healthy aging.
# Background
The aging of the HIV population is a remarkable success but comes with new challenges and questions for persons living with HIV (PLWH) and their caregivers [bib_ref] Aging in HIV-infected subjects: a new scenario and a new view, Negredo [/bib_ref] [bib_ref] The transition from co-morbidities to geriatric syndromes in HIV, Guaraldi [/bib_ref] [bib_ref] Clinical implications of aging with HIV infection: perspectives and the future medical..., Guaraldi [/bib_ref] [bib_ref] Predictions of geriatric HIV in 2030, Vance [/bib_ref]. Although long-term survival of effectively treated PLWH approaches that of the general population [bib_ref] Narrowing the gap in life expectancy between HIVinfected and HIV-uninfected individuals with..., Marcus [/bib_ref] [bib_ref] Survival of HIV-positive patients starting antiretroviral therapy between 1996 and 2013: a..., Therapy Cohort [/bib_ref] , agerelated comorbidities are more prevalent and the overall quality of life may differ [bib_ref] Clinical implications of aging with HIV infection: perspectives and the future medical..., Guaraldi [/bib_ref] [bib_ref] HIV infection, antiretroviral treatment, ageing, and non-AIDS related morbidity, Deeks [/bib_ref] [bib_ref] Premature age-related comorbidities among HIV-infected persons compared with the general population, Guaraldi [/bib_ref] [bib_ref] Aging with HIV and disability: the role of uncertainty, Solomon [/bib_ref]. Critical knowledge gaps exist on how HIV and its therapies might impact and interact with normal aging processes.
As this cohort continues to grow, it is critical to develop strategies to support healthy aging in PLWH [bib_ref] Trajectories of episodic disability in people aging with HIV: a longitudinal qualitative..., Solomon [/bib_ref]. In order to do so, we need to define health and identify the modifiable factors that impact it. Researchers and advocates have called for a revision to the UN AIDS 90 : 90 : 90 goals for HIV. Not only should 90% of those with HIV be aware of their disease, 90% be engaged in care, 90% be receiving ART and 90% have viral suppression but that 90% should have good health-related quality of life [bib_ref] Beyond viral suppression of HIV -the new quality of life frontier, Lazarus [/bib_ref].
There are little data on how to measure healthy aging in the context of HIVand no HIV-specific indexes of health. Although older adults may present with more advanced HIV disease and have less immune recovery than younger adults [bib_ref] Short communication: effects of age on virologic suppression and CD4 cell response..., Szadkowski [/bib_ref] [bib_ref] The absence of CD4R T cell count recovery despite receipt of virologically..., Gazzola [/bib_ref] , for PLWH on effective combination antiretroviral therapy (ART) neither absolute CD4 þ cell counts, CD4 þ /CD8 þ ratios, nor HIV viral load are useful surrogate markers of healthy aging [bib_ref] A frailty index predicts survival and incident multimorbidity independent of markers of..., Guaraldi [/bib_ref] [bib_ref] VACS Project Team. Does an index composed of clinical data reflect effects..., Justice [/bib_ref] [bib_ref] GEPPO (GEriatric Patients living with HIV/AIDS: a Prospective Multidimensional cOhort) Study Group...., Calcagno [/bib_ref]. The Veterans Aging Cohort Study Index [bib_ref] VACS Project Team. Does an index composed of clinical data reflect effects..., Justice [/bib_ref] creates a score by summing preassigned points for age, CD4 þ cell count and HIV-1 RNA, and general indicators of organ system injury including hemoglobin, platelets, aspartate and alanine transaminase, creatinine, and viral hepatitis C infection. The VACS Index predicts all-cause mortality, cause specific mortality, and other outcomes in those living with HIV infection. Despite these predictions, it is focused on physical health, does not capture other health domains or quality of life and has not been used to assess interventions to maintain or improve health.
In geriatric medicine, frailty is often used to assess health status [bib_ref] The interplay between frailty and intrinsic capacity in aging and HIV infection, Guaraldi [/bib_ref]. A commonly used measure is the Fried frailty phenotype, which assesses five specific features: selfreported weight loss, self-reported exhaustion, low levels of physical activity, measured 15 feet walk time and measured grip strength. The Short Physical Performance Battery (SPPB) has also been used for over 20 years to measure physical performance or functional status [bib_ref] A short physical performance battery assessing lower extremity function: association with self-reported..., Guralnik [/bib_ref]. It is a brief battery with a timed walk, repeated chair stands, and balance tests. Although useful, these two measures are largely focused on physical status.
There are many components to healthy aging beyond physical function or the simple presence or absence of comorbid disease [bib_ref] Definitions and predictors of successful aging: a comprehensive review of larger quantitative..., Depp [/bib_ref]. The new framework of the WHO's Global Strategy and Action Plan on Aging and Health [bib_ref] The World report on ageing and health: a policy framework for healthy..., Beard [/bib_ref] defines healthy aging as 'the process of developing and maintaining the functional ability that enables well being in older age'. The framework acknowledges that older people experience significant losses in their physical and cognitive capacities and that although some are inevitable, others may be avoided. The framework stresses that the study of healthy aging must consider a composite of an individual's physical and mental capacities (intrinsic capacity) in addition to factors in the extrinsic world that form the context of an individual's life (environments). In the Action Plan of the strategy, the WHO challenges researchers to improve measuring, monitoring, and understanding healthy aging in different populations [bib_ref] The World report on ageing and health: a policy framework for healthy..., Beard [/bib_ref].
The Rotterdam HAS was constructed using factor analysis from a prospective population-based study of 3527 Dutch participants at least 55 years old [bib_ref] Development of a healthy aging score in the population-based Rotterdam study: evaluating..., Jaspers [/bib_ref]. It considers five biopsychosocial domains of health: mental health, cognitive function, physical function (three subcomponents of pain, comorbidity, and activities of daily living), social support, and quality of life. For each of the seven measures a score of 0 (low, corresponding to a worse status within the domain), 1 (moderate), or 2 (high, corresponding to an optimal status within the domain) is assigned and the total summed to determine the HAS score.
To respond to the WHO challenge for standardization, we selected The Rotterdam Healthy Aging Score-HAS [bib_ref] Development of a healthy aging score in the population-based Rotterdam study: evaluating..., Jaspers [/bib_ref] to examine the concept of healthy aging in an HIV cohort. Determination of a valid, standardized measure of healthy aging would be useful in the clinic to follow the health trajectory of patients and as an objective outcome of clinical trials of interventions designed to improve it.
# Methods
We evaluated the HAS in a cross-sectional cohort of HIVpositive older adults attending a tertiary care HIV clinic in Toronto, Canada between May and August 2018. To be eligible, patients were required to be on ART with HIV RNA less than 50 copies/ml for at least 6 months. A priori, we elected to include 30, 35, and 35 participants in the age categories 40-50, 51-60 and greater than 60 years, respectively, in order to assess the impact of age on the score. Exclusion criteria were: pregnancy, previous liver cirrhosis, and current treatment with systemic steroids, chemotherapy, immunosuppressive agents or radiation therapy. A convenience sample of 101 PLWH was chosen from consecutive patients attending the clinic who consented to participate and met the inclusion criteria. Participants completed a series of questionnaires to calculate the HAS (score range 0-14).
To assess construct validity of the score in our population, we also calculated the Fried Frailty Index [24] and determined the Short Performance Physical Battery Score (SPPB) [bib_ref] A short physical performance battery assessing lower extremity function: association with self-reported..., Guralnik [/bib_ref] for each participant. In addition, we estimated a measure of health utilization by administering questionnaires on the number of emergency room visits and overnight hospital stays in the previous 6 months and the need for home assistance.
The study was approved by the University Health Network Ethics review board, and each participant provided written informed consent. The study was conducted in accordance with the Declaration of Helsinki.
## Sample size considerations
This pilot study was not powered for statistical significance but rather to determine the patient acceptability of completing the tasks and determining some preliminary performance characteristics of the score. With 100 participants, we would be able to detect a difference in HAS of 1.30 between two groups of participants, one with and one without a characteristic of interest, assuming a standard deviation in the HAS score of 2.3, a significance level of 0.05, power of 80% and that 50% of the study population has the characteristic of interest.
# Statistical analysis
The Kruskal-Wallis Rank-Sum and Fisher's Exact tests were used to make the following comparisons: demographics, HAS, and HAS domains by age group
# Results
Of 107 persons approached, 104 agreed to participate and 3 were excluded for not meeting eligibility criteria. The demographics of the participants are shown in [fig_ref] Table 1: Demographic and clinical characteristics of HIV cohort by age category [/fig_ref]. The median [interquartile range (IQR)] age was 56 years (50-62), 81 (80%) were men, and 50 (50%) were born in Canada reflective of our clinic population. Participants aged at least 61 years were more often Caucasian (75%) HAS, healthy aging score. [fig_ref] Table 3: Comparison of the Fried Frailty Score and Components by healthy aging score... [/fig_ref] shows the outcome of the Fried Frailty Score, stages, and individual components by HAS category. Overall, 10.4% are classified as frail, 79% as prefrail and 10.4% as not frail. A lower HAS score was associated with increased frailty (P ¼ 0.008). This was most strongly correlated with the sub-component of the frailty index that related to exhaustion. There was also a positive correlation between the HAS and grip strength (P ¼ 0.03). shows the relationship between the overall SPPB score and the SPPB component sub-scores with the HAS categories. In this case, there was a trend towards poorer performance and with lower HAS scores (P ¼ 0.077).
We also explored the relationship between measures of health utilization and the HAS. Of the 101 participants, seven had been hospitalized in the preceding 6 months. This included two of 39 (5.1%) participants who scored in the healthy range of the HAS, and five of 30 (16.7%) who scored in the poor range of the HAS, P ¼ 0.034. The length of hospitals stay was a median of 2 days (IQR 2-4) and did not vary by HAS score, P ¼ 0.32.
Overall 24 (23.8%) participants had an emergency room visit in the preceding 6 months, the number did not vary by HAS score, P ¼ 0.915 (data not shown), nor did the number of emergency room visits during this period vary by HAS score. Overall only three (3%) of participants had received help for activities of daily Evaluation of a healthy aging score in HIV Walmsley et al. 863 living at home during the previous 6 months (data not shown).
# Discussion
This is the first study to evaluate the HAS in a cohort of HIV-infected individuals. In our pilot study of aging PLWH, the outcome of the score ranged from 5 to 14 with a median score of 12 [IQR (interquartile range) [bib_ref] Trajectories of episodic disability in people aging with HIV: a longitudinal qualitative..., Solomon [/bib_ref] [bib_ref] Beyond viral suppression of HIV -the new quality of life frontier, Lazarus [/bib_ref] [bib_ref] Short communication: effects of age on virologic suppression and CD4 cell response..., Szadkowski [/bib_ref]. The median score did not vary by age category or sex. According to the criteria developed by the Rotterdam group, only 39% of our participants scored in the healthy range of the score.
The WHO Global Strategy and Action Plan on Aging and Health has emphasized the need for standardizing measurements of healthy aging to use in care and research to enable comparisons across populations and disciplines. Although there is currently no consensus or gold standard, researchers are attempting to develop and validate indexes to operationalize healthy aging in different disease settings [bib_ref] Definitions and predictors of successful aging: a comprehensive review of larger quantitative..., Depp [/bib_ref] [bib_ref] Indicators for healthy ageing-a debate, Fuchs [/bib_ref] [bib_ref] Towards measurement of the Healthy Ageing Phenotype in lifestyle-based intervention studies, Lara [/bib_ref] [bib_ref] Healthy aging' concepts and measures, Michel [/bib_ref] [bib_ref] A case-controlled study of successful aging in older HIVinfected adults, Moore [/bib_ref] [bib_ref] Healthy ageing: how is it defined and measured?, Peel [/bib_ref]. The indexes are broadly similar, capturing multiple dimensions of health status within the domains of the WHO framework. It remains unclear, which health dimensions should best be captured in a global score [bib_ref] Building bridges for innovation in ageing: synergies between action groups of the..., Bousquet [/bib_ref] [bib_ref] Operational definition of Active and Healthy Ageing (AHA): a conceptual framework, Bousquet [/bib_ref] [bib_ref] A meta-analysis of the correlates of successful aging in older adults, Kim [/bib_ref]. It is likely that different domains of health have more impact in different disease cohorts [bib_ref] The interplay between frailty and intrinsic capacity in aging and HIV infection, Guaraldi [/bib_ref].
After the review of existing scores, we chose to study the HAS score in our HIV population for the following reasons. The HAS was developed in a cohort of 3527 older Dutch participants [bib_ref] Development of a healthy aging score in the population-based Rotterdam study: evaluating..., Jaspers [/bib_ref]. We felt it ideal to study as the domains included in the score enabled an evaluation of health as a multidimensional state consistent with the framework established by the WHO. When developing the Rotterdam model, a number of socioeconomic and health behavioral factors that are relevant to those living with HIV were considered as covariates. Other strengths of this score included the demonstration of a mean decrease in score with age; sex differences in domain subscores, and correlations between scores of certain domains with each other. The score can be assessed as a mean score, or scores can be compared in the separate domains. Limitations of this score included: the threepart ordinal outcome, equal weighting of each dimension in the total score when some may have greater impact on healthy aging, and the inclusion of participant comorbidities without considering their severity.
In the Rotterdam study [bib_ref] Development of a healthy aging score in the population-based Rotterdam study: evaluating..., Jaspers [/bib_ref] , the mean (SD) of the HAS was 11.2 (2.2) in men and 10.7 (2.30) in women and the score outcomes ranged across the entire score continuum.
In the Rotterdam cohort, men had poorer scores in the chronic diseases domain than women. Women had poorer mental health, worse physical function, more pain, and lower quality of life compared with men. The Rotterdam study was also able to validate the score to mortality data. The age-adjusted hazard ratio per unit increase in the HAS with mortality was 0.86 (0.83-0.89) in men and 0.89 (0.87-0.91) in women.
The population studied in the Rotterdam cohort [bib_ref] Development of a healthy aging score in the population-based Rotterdam study: evaluating..., Jaspers [/bib_ref] has significant differences to that of our HIV cohort with 39.8% men, and a mean (SD) age of 75.3 (6.0) years. The majority were of Caucasian decent (>97%). Potential limitations of applying this score to an HIV population include the chronic diseases included in the domains. Many chronic diseases in the HAS are similar to those that aging PLWH are experiencing (cardiovascular disease, stroke, diabetes, chronic obstructive lung disease, chronic renal failure, and cancer) [bib_ref] Time trends for risk of severe age-related diseases in individuals with and..., Rasmussen [/bib_ref] ; however, Parkinson's disease was included but not seen in our clinic population and liver disease commonly seen in HIV was not included [bib_ref] Aging and HIV infection: a comparison between older HIV-infected persons and the..., Onen [/bib_ref]. There is some evidence that aging is 'accelerated' or 'accentuated' in HIV because of the residual immune inflammation and activation that persist despite control of HIV replication, which may affect the development of and age of presentation of comorbidities [bib_ref] Accelerated immune senescence and HIV-1 infection, Appay [/bib_ref] [bib_ref] Premature aging and premature age-related comorbidities in HIV-infected patients: facts and hypotheses, Capeau [/bib_ref] [bib_ref] Systemic effects of inflammation on health during chronic HIV infection, Deeks [/bib_ref].
Whether or not this would be reflected in the score is uncertain. Further, other factors that can affect health in PLWH were not captured in the score, such as stigma, trauma, and discrimination [bib_ref] The impact of HIV-related stigma on older and younger adults living with..., Emlet [/bib_ref] as only a depression scale was included in the mental health domain. Additionally, the underlying demographics of PLWH vary [bib_ref] Socio-demographic profile of older adults with HIV/AIDS: gender and sexual orientation differences, Brennan [/bib_ref] and social determinants of health, such as income, food, and housing could affect quality of life and healthy aging [bib_ref] Multidimensional approaches to examining gender and racial/ethnic stratification in health, Brown [/bib_ref] [bib_ref] Women's health, men's health, and gender and health: implications of intersectionality, Hankivsky [/bib_ref] [bib_ref] Indicators of 'healthy aging' in older women (65-69 years of age). A..., Swindell [/bib_ref]. Finally, the HIV cohort consists of many individuals who assume nontraditional sex roles and may have different caretaking and advocacy roles than aging heterosexual men and women, which could impact the domain of the score-related to activities of daily living. Despite these limitations, the mean and range of scores was similar in our cohort to the original Rotterdam study suggesting that it will be useful in this population.
Multiple approaches have been used to operationalize health in the general aging population but the best tools to use in the clinical and research setting remain controversial [bib_ref] Towards measurement of the Healthy Ageing Phenotype in lifestyle-based intervention studies, Lara [/bib_ref] [bib_ref] Healthy aging' concepts and measures, Michel [/bib_ref] [bib_ref] Healthy ageing: how is it defined and measured?, Peel [/bib_ref] [bib_ref] Indicators of 'healthy aging' in older women (65-69 years of age). A..., Swindell [/bib_ref] [bib_ref] Successful aging, Rowe [/bib_ref] [bib_ref] Operational definitions of successful aging: a systematic review, Cosco [/bib_ref]. It will be difficult to come up with a simple tool that can dichotomize aging as successful or not [bib_ref] Definitions and predictors of successful aging: a comprehensive review of larger quantitative..., Depp [/bib_ref]. Despite the challenges, a global healthy aging score would be useful. An ideal score would be valid for the population under study, be relatively easy and inexpensive to use, and be responsive to detecting changes in health over time. A standardized score could support clinical decision-making around a person's strengths and quality of life, the need for investigation or intensity of follow-up, and monitoring health trajectories over time or with specific therapies. A score could help researchers identify areas for directed interventions to maintain or improve health and provide an outcome to monitor response. Population level assessments could provide policy makers with comprehensive data to inform the allocation of funds and services to keep persons healthy.
Other groups have evaluated the Fried Frailty phenotype and the SPPB in the context of HIV [bib_ref] Physical functioning among patients aging with human immunodeficiency virus (HIV) versus HIV..., Crane [/bib_ref] [bib_ref] Frailty and physical function in older HIV-infected adults, Branas [/bib_ref] [bib_ref] Lower frailty is associated with successful cognitive aging among older adults with..., Wallace [/bib_ref] [bib_ref] Frailty is strongly associated with increased risk of recurrent falls among older..., Tassiopoulos [/bib_ref] [bib_ref] Frailty and circulating markers of inflammation in HIVR and HIV-men in the..., Margolick [/bib_ref] [bib_ref] Frailty in people aging with human immunodeficiency virus (HIV) infection, Brothers [/bib_ref]. The limitations of using these as indicators of health is that they fail to take into consideration other domains of health, such as cognitive function, mental health, and social well being. We were able to demonstrate correlation of the HAS scores with these indices especially with respect to the subscales addressing physical function adding to the validity of the HAS.
In attempts to improve on the frailty phenotype in the context of HIV, Brothers et al. [bib_ref] Predictors of transitions in frailty severity and mortality among people aging with..., Brothers [/bib_ref] In order to be a useful endpoint in interventional studies, a measure of healthy aging needs to be predictive of clinically relevant endpoints, such as disease progression, hospitalization, and mortality. The original Rotterdam cohort was able to demonstrate an association between the HAS score and mortality. In our cohort, we were able to demonstrate a significant correlation between poor scores on the HAS and hospitalization in the preceding 6 months. However, as the numbers were small, this observation needs to be confirmed with a larger population.
This is the first step to assess a validated healthy aging score in an HIV-infected population. The strengths of our study are that we were able to demonstrate within a small cohort that individuals scored across the range of the score values without a significant floor or ceiling effect. We were able to correlate the score with other indicators of primarily physical health, such as the Fried Frailty phenotype and the SPPB score. We were also able to demonstrate a correlation between health and hospitalization, although this conclusion is limited by small numbers. Limitations include our small sample size and the small percentage of women. Further study is warranted in a larger HIV aging population and to validate it against clinical outcomes, such as disease progression, hospitalization, and mortality. Furthermore, as our study was cross-sectional, more work is required to determine if the score is responsive to change.
In conclusion, in this pilot study of 101 older PLWH, we demonstrated a range of healthy aging scores using an index that has been validated in a population of older adults in the Netherlands [bib_ref] Development of a healthy aging score in the population-based Rotterdam study: evaluating..., Jaspers [/bib_ref]. The score correlated with other indexes that measure health and with recent hospitalization. Only 39% of participants scored in the 'healthy' range despite viral suppression and good immunologic response to ART suggesting that other domains are important in evaluating health in this population. As the HIV population continues to age, we will need measures to assess outcomes and test interventions to maintain or optimize health [bib_ref] HIV and ageing: improving quantity and quality of life, Althoff [/bib_ref]. This will enable us to determine whether or not we can achieve the proposed fourth 90 of the UN AIDS goals for HIV that 90% of those with HIV have good health and quality of life [bib_ref] Beyond viral suppression of HIV -the new quality of life frontier, Lazarus [/bib_ref]. The data from our pilot study suggest we have work to do to reach this goal.
[fig] Figure 1: Distribution of the Healthy Aging Score in the HIV cohort. HAS, healthy aging score. [/fig]
[table] Table 1: Demographic and clinical characteristics of HIV cohort by age category. IDU, injection drug use; IQR, interquartile range. P ¼ 0.051). More women were black (55%) than men (7%, P < 0.001). Participants at least 61 years of age had longer median durations of HIV infection (25 years, P ¼ 0.001) and lower CD4 þ nadirs (153 cells/ml, P ¼ 0.047) compared with 40-50 years (16 years, 267 cells/ml) or 51-60-year categories (22 years, 180 cells/ml). Median (IQR) CD4 þ cell count at enrollment was 574 (417-794) cells/ml. [/table]
[table] Table 2: Health aging score and sub-categories by age category. [/table]
[table] Table 3: Comparison of the Fried Frailty Score and Components by healthy aging score status.Table 4. Comparison of Short Performance Physical Battery Score and subcomponents by healthy aging score status. [/table]
|
Dominant Th1 and minimal Th17 skewing in discoid lupus revealed by transcriptomic comparison with psoriasis
Discoid lupus erythematosus (DLE) is the most common skin manifestation of lupus. Despite its high frequency in systemic lupus in addition to cases without extracutaneous manifestations, targeted treatments for DLE are lacking, likely due to a dearth of knowledge of the molecular landscape of DLE skin. Here, we profiled the transcriptome of DLE skin in order to identify signaling pathways and cellular signatures that may be targeted for treatment purposes. Further comparison of the DLE transcriptome to that of psoriasis, a useful reference given our extensive knowledge of molecular pathways in this disease, provided a framework to identify potential therapeutic targets. Although a growing body of data supports a role for IL-17 and Th17 cells in systemic lupus, we show a relative enrichment of IFN-γ-associated genes without that for IL-17associated genes in DLE. Extraction of T cells from the skin of DLE patients identified a predominance of IFN-γ-producing Th1 cells and an absence of IL-17-producing Th17 cells, complementing the results from whole skin transcriptomic analyses. These data therefore support investigations into treatments for DLE that target Th1 cells or the IFN-γ signaling pathway.Users may view, print, copy, and download text and data-mine the content in such documents, for the purposes of academic research, subject always to the full Conditions of use:
# Introduction
Lupus erythematosus (LE) is a heterogeneous disease whose hallmark is the formation of circulating autoantibodies. Systemic LE (SLE) is a multiorgan form of LE with manifestations that can variably affect the skin, joints, kidneys and CNS, among other organ systems. The etiology of SLE is multifactorial, and both genetic and environmental influences are believed to contribute.
The most common specific skin manifestation of lupus is discoid LE (DLE), and its presence is one of the eleven American College of Rheumatology diagnostic criteria for SLE [bib_ref] The 1982 revised criteria for the classification of systemic lupus erythematosus, Tan [/bib_ref]. DLE represents up to 80% of all cutaneous lupus cases [bib_ref] Cutaneous lupus erythematosus and the association with systemic lupus erythematosus: a population-based..., Gronhagen [/bib_ref] , and this skin manifestation is notable for its tendency to cause disfigurement, alopecia, and scarring [bib_ref] Chronic Discoid Lupus Erythematosus, Rothfield [/bib_ref]. Approximately 1 in 4 patients with SLE have DLE [bib_ref] Phenotypic associations of genetic susceptibility loci in systemic lupus erythematosus, Sanchez [/bib_ref] [bib_ref] Trends in the incidence and mortality of systemic lupus erythematosus, Uramoto [/bib_ref] , and, in some DLE cases, the skin can be the only end-organ affected without extracutaneous involvement. Interestingly, a recent observational, population-based study of over 1000 cutaneous lupus patients noted an overall rate of progression to SLE of 18% within a three year follow-up period [bib_ref] Cutaneous lupus erythematosus and the association with systemic lupus erythematosus: a population-based..., Gronhagen [/bib_ref]. The ease of sampling affected skin potentially makes DLE a convenient and accessible model to study end-organ pathology in SLE.
Currently, antimalarials are the most commonly used systemic treatment for cutaneous lupus, despite the fact that the mechanism of action of this class of drugs is not completely understood [bib_ref] New insights into mechanisms of therapeutic effects of antimalarial agents in SLE, Wallace [/bib_ref]. Further, a need exists for new treatments because many patients do not respond to antimalarials or other available immunosuppressants. Insights into the pathophysiology of damage to this end-organ have the potential to identify therapeutic targets and new treatments.
Initial studies have focused on the types of cells present in DLE lesions. Histopathologically, a dense perivascular, periadnexal lymphocytic infiltrate with interface dermatitis can be appreciated, with most analyses indicating CD4 T cells are the predominant inflammatory cell type. A growing body of evidence supports the importance of Th17 CD4 T cells in SLE pathogenesis [bib_ref] Balance between Regulatory T and Th17 Cells in Systemic Lupus Erythematosus: The..., Alunno [/bib_ref] [bib_ref] Treatment of alopecia areata: modern principles and perspectives, Brajac [/bib_ref] [bib_ref] Plasma IL-17A is increased in new-onset SLE patients and associated with disease..., Chen [/bib_ref] [bib_ref] Frequency and functional activity of Th17, Tc17 and other T-cell subsets in..., Henriques [/bib_ref] [bib_ref] The relation of interleukin 17 (IL-17) and IL-23 to Th1/Th2 cytokines and..., Mok [/bib_ref] [bib_ref] Type 17 T-helper cells might be a promising therapeutic target for systemic..., Pan [/bib_ref] [bib_ref] Dysregulated balance of Th17 and Th1 cells in systemic lupus erythematosus, Shah [/bib_ref] , although comparisons of the relative frequencies of T helper subtypes in DLE have not been rigorously undertaken. Targeting the specific subset of T cells or their effector cytokines that are critical to the pathogenesis of DLE is a strategy for the design of new therapeutic agents, as has been done for other cutaneous autoimmune diseases. Psoriasis, now known to be a Th17-mediated disease, is a clear example; ustekinumab, a monoclonal antibody that blocks Th17-and Th1related pathways [bib_ref] Efficacy and safety of ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients..., Leonardi [/bib_ref] [bib_ref] Efficacy and safety of ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients..., Papp [/bib_ref] was recently approved for the treatment of psoriasis, and two IL-17/Th17-pathway inhibitors were recently shown to be efficacious in clinical trials [bib_ref] Brodalumab, an antiinterleukin-17-receptor antibody for psoriasis, Papp [/bib_ref].
We were interested in identifying global perturbations in immune related pathways or immune cell signatures that were present in skin lesions of DLE. Here, we provide a molecular description of the transcriptome of DLE. Analysis of activated gene expression pathways was performed, identifying a preponderance of immune-related pathway activation. Next, we compared the transcriptional profile of DLE skin with psoriasis, the most well-characterized autoimmune skin disease to date and therefore a useful basis for comparison. Our interest was to characterize the type of T cells infiltrating DLE lesions. Our reference skin disease, psoriasis, is caused by interactions between skin keratinocytes and T cells that make up the Th17, Th22, and Th1 subsets. In contrast, we report relatively minimal Th17 signatures, as ascertained by Gene Set Enrichment Analysis, gene expression and T cell protein production, but rather skewing towards a predominantly Th1 signature in whole skin and infiltrating T cells in DLE. We thus provide a rationale to exploring targeting cells and molecular pathways associated with the formation and function of Th1 cells.
# Results
In order to define the transcriptional profile of discoid lupus skin lesions, we first performed punch biopsies of the affected skin of discoid lupus patients and healthy control patients. RNA extracted from skin samples was processed and hybridized to GeneChip Human Genome U133 2.0 microarrays. Gene lists of differentially expressed genes between DLE and normal skin were generated using criteria of greater than two-fold change and less than 0.05 false discovery rate and subsequently interrogated using Ingenuity Pathway Analysis. The majority of pathways were indicative of changes in immune-related pathways and immune cell signatures [fig_ref] Figure 1: Transcriptomic analysis of DLE [/fig_ref]. Upregulated pathways include that for "SLE signaling" and "Interferon signaling," consistent with reports of a prominent interferon signature in the peripheral blood of patients with SLE [bib_ref] Interferoninducible gene expression signature in peripheral blood cells of patients with severe..., Baechler [/bib_ref] [bib_ref] Elevated serum levels of interferon-regulated chemokines are biomarkers for active human systemic..., Bauer [/bib_ref] [bib_ref] Interferon and granulopoiesis signatures in systemic lupus erythematosus blood, Bennett [/bib_ref] [bib_ref] Coordinate overexpression of interferon-alpha-induced genes in systemic lupus erythematosus, Kirou [/bib_ref] as well as the upregulated interferon genes in DLE lesions reported by others [bib_ref] Enhanced type I interferon signalling promotes Th1-biased inflammation in cutaneous lupus erythematosus, Wenzel [/bib_ref]. Quantitative PCR confirmations supported increased cellular signatures of T cells (CD3D), CD8 T cells (CD8A, GZMB, GNLY), B cells (IGJ), macrophages (CD163), natural killer cells (KLRK1, KLRD1), and dendritic cells (CD86, SIGLEC1, MR1) as well as interferon (MX1, OASL, STAT1), IL-15, and IL-1β signatures. DLE skin, therefore, is marked by potent induction of immune pathways and immune cell signatures.
Because of the high number of pathways and immune cell signatures that were present in our analysis of the DLE transcriptional profile, a well described inflammatory skin disease, psoriasis, was used as a reference for comparison. Although histological and immunohistochemical features of the two diseases are well-known, a comparative tissuelevel appraisal of these disease entities has not previously been done and provides context to our molecular analyses. Histological and immunohistochemical comparisons highlighted differences in these skin diseases [fig_ref] Figure 2: Figure 2. [/fig_ref]. In psoriasis, a superficial perivascular infiltrate is appreciated on hematoxylin and eosin staining with overlying hyperplasia of the epidermis, whereas prominent dermal edema and inflammation of the dermoepidermojunction is appreciated in DLE. DLE was noted to have more inflammatory cell infiltrates of CD3 + and CD11c + cells. Also, differences were noted in cellular localization; CD8 + infiltrates were predominantly epidermal in psoriasis compared with a dermal preponderance in DLE [fig_ref] Figure 2: Figure 2. [/fig_ref]. It is clear that, despite both entities being autoimmune skin diseases, differences exist in the relatively quantities and tissue localizations of immune cells.
A global comparison of the transcriptional profiles of DLE and psoriasis was performed to identify contrasting elements. DLE and psoriasis samples formed distinct groups on Principal Components Analysis plots, indicative of distinct molecular profiles [fig_ref] Figure 3 J: Transcriptomic comparison between DLE and psoriasis [/fig_ref]. By comparing 'DLE versus normal skin' and 'psoriasis versus normal skin,' we found that 591 genes were uniquely upregulated and 358 genes were uniquely downregulated in DLE, and, in psoriasis, 770 genes were uniquely upregulated and 989 genes were uniquely downregulated [fig_ref] Figure 3 J: Transcriptomic comparison between DLE and psoriasis [/fig_ref]. A heat map of the most differentially expressed genes between DLE and psoriasis, generated by directly comparing DLE and psoriasis, highlighted the major differences in the transcriptomes of these diseases [fig_ref] Figure 3 J: Transcriptomic comparison between DLE and psoriasis [/fig_ref]. For example, genes that were among the most highly expressed in psoriasis when compared with DLE included DEFB4A, S100A12 and IL8, genes associated with the IL-17-regulated pathway , and quantitative RT-PCR confirmed this relationship (Supplementary [fig_ref] Figure 1: Transcriptomic analysis of DLE [/fig_ref]. In total, these data indicate DLE has a distinct molecular profile when compared with psoriasis, and indicated that these diseases are the result of distinct pathological mechanisms.
In order to further define differences in T helper associated cytokine pathways, such as the IL-17 pathway, differentially expressed genes between DLE and psoriasis were then interrogated with Gene Set Enrichment Analysis (GSEA) [bib_ref] Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles, Subramanian [/bib_ref]. The goal of GSEA is to determine whether a defined set of genes (described below) are overrepresented at the extremes (either increased or decreased expression) throughout a reference gene list (in the analysis here, we use the gene list from DLE versus psoriasis). The degree to which a gene set is overrepresented at the extremes of a reference gene list is reflected in an enrichment score. Because we were interested in T helper cell associated cytokine pathways, gene sets from keratinocytes co-cultured with the T helper cytokines IL-17, IFN-γ, and IL-22 were used in this analysis. Additionally, we included gene sets from keratinocytes cocultured with TNF [bib_ref] Effective treatment of psoriasis with etanercept is linked to suppression of IL-17..., Zaba [/bib_ref] , as inhibiting the TNF signal is an efficacious form of treatment in patients with psoriasis, or IFN-α [bib_ref] The psoriatic transcriptome closely resembles that induced by interleukin-1 in cultured keratinocytes:..., Mee [/bib_ref] , a signature cytokine in the peripheral blood of SLE patients and identified in DLE lesions [bib_ref] Interferoninducible gene expression signature in peripheral blood cells of patients with severe..., Baechler [/bib_ref] [bib_ref] Elevated serum levels of interferon-regulated chemokines are biomarkers for active human systemic..., Bauer [/bib_ref] [bib_ref] Interferon and granulopoiesis signatures in systemic lupus erythematosus blood, Bennett [/bib_ref] [bib_ref] Coordinate overexpression of interferon-alpha-induced genes in systemic lupus erythematosus, Kirou [/bib_ref]. GSEA analysis was conducted for the differentially expressed genes between DLE and psoriasis; the set of genes with increased expression in DLE when compared with psoriasis showed enrichment of the following sets of genes: upregulated genes in keratinocytes cocultured with IFN-α or IFN-γ . In contrast, the set of genes with decreased expression in DLE when compared with psoriasis (therefore higher expression in psoriasis compared with DLE) showed enrichment of the upregulated gene sets from keratinocytes cocultured with IL-17, IL-22, or TNF . In sum, the GSEA analysis is consistent with prior reports implicating IL-17, IL-22, and TNF pathways in the pathogenesis of psoriasis and type I interferon for DLE, but further supports that activation of the IFN-γ pathway is more characteristic of DLE than is activation of the IL-17 pathway.
One potential caveat of our GSEA analysis is that the IL-17-associated signature, however, may be active in DLE, but not to a sufficient extent to overcome or render statistically insignificant the signal for psoriasis. Because of this possibility, we next examined the expression of the primary T helper associated cytokines. Quantitative RT-PCR for IFN-γ, IL-17A, and IL-22, corresponding to Th1, Th17, and Th22 cells, respectively, revealed substantial mRNA expression of these cytokines in psoriasis skin samples [fig_ref] Figure 4: Figure 4. [/fig_ref] , consistent with prior reports for psoriasis [bib_ref] Role of the immune system and immunological circuits in psoriasis, Jabbari [/bib_ref] [bib_ref] Psoriasis vulgaris lesions contain discrete populations of Th1 and Th17 T cells, Lowes [/bib_ref]. In contrast, DLE skin samples expressed relatively low levels of IL-17A and IL-22 and high levels of IFN-γ. IL-13 expression was not detected in psoriasis or DLE samples (data not shown). These data support a Th1/Th17/Th22 mixed signature in psoriasis, but a predominantly Th1-skewed signature, with minimal Th17 activation, for DLE.
In order to assess the relative frequencies of T cells among T helper subsets in DLE lesions, we isolated T cell emigrants from skin samples from patients with DLE and psoriasis for comparison. T cells were stained for intracellular IFN-γ and IL-17A after in vitro stimulation with phorbol-12-myristate-13-acetate and ionomycin. A detectable population of CD4 T cells from psoriasis skin samples were able to readily produce both IFN-γ and IL-17A, while only IFN-γ producing CD4 T cells from DLE skin were appreciated in DLE lesions [fig_ref] Figure 4: Figure 4. [/fig_ref]. All DLE samples had < 1% CD4 T cells expressing IL-17, while psoriasis samples analyzed in parallel and previously published data demonstrated substantial IL-17 production by lesional CD4 T cells (Supplementary [fig_ref] Figure 2: Figure 2. [/fig_ref]. These data demonstrating IFN-γ protein elaboration support that the T helper infiltrate of DLE skin samples predominantly consist of Th1 cells.
# Discussion
DLE is a cutaneous manifestation of lupus that causes scarring and disfigurement. Treatment usually requires systemic immunosuppressive agents with ill-defined mechanisms of action, the potential for harmful side effects, and a requirement for frequent laboratory surveillance. In many cases, available systemic agents are unable to adequately control the disease. There is therefore a critical need for the development of a targeted therapeutic agent with a favorable side effect profile.
The rational identification of potential therapeutic targets requires a well-developed understanding of the pathogenesis of disease. Psoriasis is an example of an autoimmune skin disease in which our growing understanding of the cytokines and T cells critical to the disease have led to the identification of new treatment targets. Biologics that inhibit IL-17and IL-22-mediated signaling, recently identified to play critical roles in this disease, are at various stages of development [bib_ref] Brodalumab, an antiinterleukin-17-receptor antibody for psoriasis, Papp [/bib_ref] , with some now entering the final phases of clinical trials (http://clinicaltrials.gov).
A molecular characterization of DLE, however, has not been carried out to the same extent as that for psoriasis. The current knowledge of DLE pertains to targeted assessments of specific signaling pathways and has been most well described for type I interferons [bib_ref] The interferon-regulated gene signature is elevated in subacute cutaneous lupus erythematosus and..., Braunstein [/bib_ref] [bib_ref] Scarring skin lesions of discoid lupus erythematosus are characterized by high numbers..., Wenzel [/bib_ref] [bib_ref] Enhanced type I interferon signalling promotes Th1-biased inflammation in cutaneous lupus erythematosus, Wenzel [/bib_ref] [bib_ref] Type I interferon-associated cytotoxic inflammation in cutaneous lupus erythematosus, Wenzel [/bib_ref]. Cytokines associated with specific T cell subsets have been explored in two separate prior studies [bib_ref] T-cell cytokine network in cutaneous lupus erythematosus, Stein [/bib_ref] [bib_ref] Detection of type 1 cytokines in discoid lupus erythematosus, Toro [/bib_ref] , although their results conflicted with each other and were published prior to the discovery of the Th17 and Th22 T cell subsets.
Recent descriptions of the presence of Th17 cells in the blood of DLE patients as well as a growing literature on the role of Th17 cells in SLE [bib_ref] Plasma IL-17A is increased in new-onset SLE patients and associated with disease..., Chen [/bib_ref] [bib_ref] Frequency and functional activity of Th17, Tc17 and other T-cell subsets in..., Henriques [/bib_ref] [bib_ref] Imbalance between Th17 and regulatory T-cells in systemic lupus erythematosus, Kleczynska [/bib_ref] [bib_ref] The relation of interleukin 17 (IL-17) and IL-23 to Th1/Th2 cytokines and..., Mok [/bib_ref] [bib_ref] Dysregulated balance of Th17 and Th1 cells in systemic lupus erythematosus, Shah [/bib_ref] seemed to support the hypothesis that Th17 cells would be present at appreciable numbers and may be playing a role in the development of discoid lesions-our data illustrating minimal Th17 involvement in DLE lesions were unexpected. Several reports have described the presence of IL-17 related activity or signals in DLE lesions by immunohistochemistry [bib_ref] Expression of interleukin-17 is correlated with interferon-alpha expression in cutaneous lesions of..., Oh [/bib_ref] [bib_ref] IL-17 in cutaneous lupus erythematosus, Tanasescu [/bib_ref] or in the serum of DLE patients . IL-17-related markers have also been reported as upregulated compared to healthy controls in the serum of SLE patients in some [bib_ref] Dysregulated balance of Th17 and Th1 cells in systemic lupus erythematosus, Shah [/bib_ref] [bib_ref] Increased serum interleukin 17 in patients with systemic lupus erythematosus, Zhao [/bib_ref] , but not all [bib_ref] Treatment of alopecia areata: modern principles and perspectives, Brajac [/bib_ref] [bib_ref] Identification of activated cytokine pathways in the blood of systemic lupus erythematosus,..., Higgs [/bib_ref] studies. In a few cases, enhanced numbers of Th17 cells were only significantly increased in those patients with higher SLEDAI scores, a research instrument that integrates clinical data and laboratory results to assess SLE activity and identify SLE flares [bib_ref] Increased expression of costimulatory markers CD134 and CD80 on interleukin-17 producing T..., Dolff [/bib_ref]. The results of our analysis, however, indicated that the expected Th17-associated gene set was minimally enriched in our comparative DLE data, and this was further corroborated by the absence of IL-17-producing infiltrating T cells in DLE.
The study presented here provides a global description of gene expression in lesional skin of discoid lupus; although others have used customized microarray technologies in the examination of specific aspects of DLE [bib_ref] Identification and molecular analysis of glycosaminoglycans in cutaneous lupus erythematosus and dermatomyositis, Chang [/bib_ref] , we found no substantial global description of lesional skin in DLE in the literature to date. By comparing our transcriptomic data to that of psoriasis, an autoimmune skin disease whose gene expression profile in skin has been described in dozens of subjects [bib_ref] Insights into psoriasis and other inflammatory diseases from large-scale gene expression studies, Bowcock [/bib_ref] [bib_ref] Global gene expression analysis reveals evidence for decreased lipid biosynthesis and increased..., Gudjonsson [/bib_ref] [bib_ref] Transcriptional profiling of psoriasis using RNAseq reveals previously unidentified differentially expressed genes, Jabbari [/bib_ref] [bib_ref] Evaluation of the psoriasis transcriptome across different studies by gene set enrichment..., Suarez-Farinas [/bib_ref] [bib_ref] Novel mechanisms of T-cell and dendritic cell activation revealed by profiling of..., Zhou [/bib_ref] , we were able to identify signatures that may be useful in the pursuit of novel therapeutics. Our gene expression and protein data does not support the use of an IL-17 modulating strategy in the treatment of this disease. Rather, these data invite exploring interventions that target IFN-γ or Th1 cell infiltration in the treatment of DLE.
Therapeutic agents that target interferon signaling are currently in development for use in autoimmune diseases. Inhibitors of proximal signaling mediators of IFN-γ signaling, including JAK1 and JAK2, have recently been approved for the treatment of myelofibrosis [bib_ref] JAK inhibition with ruxolitinib versus best available therapy for myelofibrosis, Harrison [/bib_ref] [bib_ref] Safety and efficacy of INCB018424, a JAK1 and JAK2 inhibitor, in myelofibrosis, Verstovsek [/bib_ref] [bib_ref] A double-blind, placebocontrolled trial of ruxolitinib for myelofibrosis, Verstovsek [/bib_ref]. Additionally, some studies have examined the use of these small molecular inhibitors in topical form [bib_ref] Preclinical evaluation of local JAK1 and JAK2 inhibition in cutaneous inflammation, Fridman [/bib_ref] , which would be especially useful in the context of DLE treatment in the absence of SLE.
It is unclear whether similar pathogenic mechanisms are present in other affected end-organs in lupus, and it is possible that systemic inhibition of pathways activated in DLE may be useful in the treatment of SLE. Interestingly, some immunomodulating biologics that are currently in use for the treatment of psoriasis additionally address extracutaneous manifestions of the disease. Another possibility is that DLE is a marker of a global inflammatory profile in a subset of SLE patients, and the presence of DLE identifies those SLE patients that have one particular molecular pattern of end-organ pathology that is distinct from that seen in SLE patients without DLE. In this scenario, molecular profiling of skin lesions may be useful to identify therapeutic targets in a specific subset of SLE patients. Genetic studies have defined SLE-associated single nucleotide polymorphisms that are enriched in SLE patients with DLE, supporting that the presence of DLE may be indicative of a particular "flavor" of SLE [bib_ref] Phenotypic associations of genetic susceptibility loci in systemic lupus erythematosus, Sanchez [/bib_ref]. Furthermore, whether the pathogenic mechanisms in DLE lesional skin is different in the context of systemic disease or cutaneous involvement alone, will require further studies with adequate patient samples to power this form of stratification. While our study does not stratify based on the presence of SLE, our data provide a global description of gene expression in DLE skin and provide a springboard to future work potentially culminating in a specific, efficacious treatment.
# Materials & methods
## Skin samples
Eleven patients with active DLE were enrolled in the study . Punch biopsy and shave biopsy specimens of psoriasis (n=5) and normal (n=3) skin samples were from patients with active moderate to severe disease or healthy subjects, respectively. Additional sample data from prior studies were added as indicated in the text or Supplementary Methods. All procedures were performed under Institutional Review Boardapproved protocols at New York University or at the Rockefeller University Hospital. Punch biopsy specimens were frozen in optimal cutting temperature medium for immunohistochemistry or flash frozen with liquid nitrogen for RNA extraction. Shave biopsy specimens were prepared for T cell emigrant isolation and functional cytokine expression (see below) as previously described .
## Sample preparation for rt-pcr and gene chip analysis
Total RNA was extracted from flash frozen punch biopsy samples using the RNeasy Mini Kit (Qiagen, Valencia, CA). DNA was removed by on-column DNase digestion by the RNase-free DNase Set (Qiagen).
# Dna microarray analysis
Human Genome U133A 2.0 gene chips (Affymetrix, Inc., Santa Clara, CA) were used for this study. The U133A 2.0 array includes more than 22,000 probe sets to analyze expression level of more than 18,400 transcripts, including 12,681 genes with known identity. Descriptions of methods for total RNA preparation for microarray hybridization, microarray data analysis, and Ingenuity Pathway Analysis (www.ingenuity.com) have been included in the Supplementary Materials.
# Quantitative rt-pcr analysis
Quantitative PCR was performed using RNA or amplified cDNA using Taqman gene expression assays. Predesigned primer and probe sets were purchased from Applied Biosystems (Carlsbad, CA). Data normalized to human acidic ribosomal protein housekeeping gene were quantified by software provided with Applied Biosystems. Statistical analysis compared log 2 transformed, normalized values from DLE and psoriasis samples by unpaired two-tailed students t-test.
## Functional cytokine expression, intracellular cytokine staining, and flow cytometry
Shave biopsy specimens were cultured for two hours in 0.5% dispase (Sigma Aldrich Corp., St Louis, MO) to separate the epidermis and dermis. Dermal T cells were obtained by culturing the dermis for three days in RPMI 1640 (Gibco-BRL Life Technologies, Carlsbad, CA) supplemented with 10% human pooled serum (Mediatech, Inc., Manassas, VA), and 1% 1 M HEPES buffer (Sigma Aldrich, St. Louis, MO); one DLE sample did not yield a sufficient number of live cells to perform further analyses. T cells were cultured for 4 hours with 10 μg/ml brefeldin A with or without 25 ng/ml phorbol myristate acetate (PMA) and 2 μg/ml ionomycin (all from Sigma Aldrich Corp.). Cells were frozen in 10% DMSO (ATCC, Manassas, VT) in RPMI-1640 (Gibco-BRL Life Technologies) with 1 mM HEPES buffer (Sigma Aldrich), 0.1% gentamicin (Gibco-BRL Life Technologies) and 5% normal human serum (C-Six Diagnostics, Germantown, WI). Frozen cells were thawed, washed, and subjected to an intracellular staining protocol. Briefly, cells were stained for 30 minutes at room temperature with antibodies specific for human CD3 and CD4. After washing with FACS Buffer, cells were permeablized and stained with antibodies specific for human IFN-γ and IL-17A. Samples were acquired using an LSRII (BD Biosciences, Rockville, MO) and analyzed using FlowJo software (Treestar, Ashland, OR).
## Immunohistochemistry
Frozen sections were stained with hematoxylin (Fisher Scientific, Pittsburgh, PA, U.S.A.) and eosin (Shandon, Pittsburgh, PA, U.S.A.). As previously described for immunohistochemistry [bib_ref] Transcriptional profiling of psoriasis using RNAseq reveals previously unidentified differentially expressed genes, Jabbari [/bib_ref] , frozen sections were blocked with 10% normal horse serum, and endogenous peroxidases were quenched by incubation with diluted hydrogen peroxide (1:10 dilution of 3% hydrogen peroxide). Sections were incubated overnight at 4°C with primary monoclonal antibodies. Biotin-labeled horse anti-primary antibodies were used for secondary binding, and thereafter the signals were amplified with avidin-biotin complex (Vector Laboratories, Burlingame, CA, U.S.A.). Subsequently, the sections were developed using chromogen 3-amino-9-ethylcarbazole (Sigma-Aldrich, St Louis, MO, U.S.A.).
# Human subjects declaration
All studies have been approved by the appropriate Institutional Review Boards and were conducted under the Declaration of Helsinki principles. Informed written consent was received from participants prior to inclusion in the study.
# Supplementary material
Refer to Web version on PubMed Central for supplementary material. Quantitative RT-PCR of selected upregulated genes in the DLE versus normal transcriptome corresponding to interferon-associated genes (MX1, OASL, STAT1), immune cell associated genes, and selected cytokines in DLE (n=7) and normal (n=6) samples. Genes were normalized to human ribosomal acidic protein. Box and whisker plots (middle line: median; box: lower to upper quartile; whiskers: minimum and maximum), * p < 0.05, ** p < 0.005, *** p < 5 × 10 −4 , **** p < 5 × 10 −5 , † p < 5 × 10 −6 , † † p < 5 × 10 −7 , † † † † p < 5 × Microscopic and immunohistologic comparison of DLE and psoriasis. Skin sections from patients with DLE, psoriasis, or healthy control patients were stained with hematoxylin and eosin or were subjected to immunohistochemistry with antibodies specific for CD3, CD11c, or CD8. Scale bar = 100 μm. Comparison of T helper subsets in DLE and psoriasis. (A) Quantitative RT-PCR comparison of T helper subset-associated cytokines in DLE and psoriasis (PsO). * p < 0.05, ** p < 5 × 10 −5 , † p < 5 × 10 −6 . (B) Intracellular cytokine staining of CD4 T cells extracted from lesions from patients with DLE (n=4) or psoriasis (n=4) for IFN-γ or IL-17A, stimulated with (+ PMA/I, right column) or without (− PMA/I, left column) PMA and ionomycin. Numbers are percentages of CD3 + CD4 + events positive for either IFN-γ or IL-17A.
[fig] Figure 1: Transcriptomic analysis of DLE. (A) Selected top canonical pathways from interrogation of the DLE (n=7) versus normal transcriptome (n=13) with Ingenuity Pathway Analysis. (B) [/fig]
[fig] Figure 2: Figure 2. [/fig]
[fig] Figure 3 J: Transcriptomic comparison between DLE and psoriasis. (A) Principal components analysis plot comparing DLE (n=7), psoriasis (PsO, n=18), and normal control (n=13) samples. (B) Venn diagram comparing upregulated and downregulated genes from DLE versus normal skin (DLE) and psoriasis versus normal skin (PsO). (C) Heat map of the most upregulated and downregulated genes in DLE versus psoriasis comparison. Invest Dermatol. Author manuscript; available in PMC 2014 July 01. [/fig]
[fig] Figure 4: Figure 4. [/fig]
|
Cardiopulmonary Sleep Spectrograms Open a Novel Window Into Sleep Biology—Implications for Health and Disease
# Introduction
The prevalence of sleep disorders has been increasing over the last two decades. Disorders like insomnia and sleep apnea have a prevalence of as much as 20% in the general population. There is a need for nimble sleep state estimation, diagnostics, and tracking. One approach seeing increasing utilization both in formal medical and consumer wearable devices is through analysis of heart rate and respiration. There is a strong correlation between changes in heart rate variability and sleep during health and disease. High frequency (HF) components mainly present parasympathetic activity, while low frequency (LF) components is partly a quantitative marker of sympathetic modulation. LF and the LF/HF ratio are high in Wake and decrease in NREM sleep, peaking once more during REM sleep, while HF follows the opposite trend. Deep NREM sleep (N3) typically has the greatest HF power. Sleep disruptive influences such as sleep apnea, insomnia, and depressionare associated with an increase in the LF components. Different techniques have been used to assess for such changes one of which is analysis of cardiopulmonary coupling and coherence (CPC) patterns. In this technique a single lead electrocardiogram (ECG) or photo plethysmogram (PPG) is used to extract heart rate variability and ECG or PPG signal derived respiration (EDR/PDR). Contrary to the stage/grade approach to conventional sleep characterization, CPC analysis based on coupling and coherence provides a novel and complementary view of sleep, that of bistability, which are particularly well defined during NREM sleep. Thus, distinct patterns of CPC are observed: high frequency coupling (HFC) which is associated with stable NREM sleep and low frequency coupling (LFC) which is associated with unstable and often fragmented NREM sleep. A third CPC pattern named very low frequency coupling occur in both REM sleep and wake, which may be differentiated by analysis of signal quality and motion artifact (Al. High and low frequency coupling are mutually exclusive and shift logically with disease states and treatments. For example, there is an increase LFC in patients with insomnia and this can be tracked in the ambulatory setting . There is increased LFC during sleep in unmedicated patients with major depression and improvement with therapy of major depression. Integrating CPC with oximetry allows generating a true FDA approved apnea-hypopnea index (AHI), which shows good correlation with conventional polysomnogram-derived AHI.
The overall goal of this article is to review physiological basis, techniques, and applications of CPC spectrograms in sleep in health and disease. There are three aims for this review, to show that-(a) CPC shows a fundamental characteristic of NREM sleep-bimodality, across a number of physiologies; (b) CPC has several uses in sleep apnea care-diagnosis, phenotype, tracking outcomes; (c) CPC can diagnose and track non-apneic sleep fragmentation and medication effects, and should be used in the appropriate clinical context.
# Physiology background
Entrainment between heart rate and respiration in humans has been described since the early twentieth century. It has been suggested that such synchrony between heart rate and respiration improves pulmonary gas exchange and computational models have shown that healthy cardiopulmonary coupling minimizes the heart workload while maintaining adequate ventilation. Such strong cardiopulmonary coupling is seen at its best during deep sleep, sedation, and anesthesia. There is a critical influence of the autonomic nervous system on cardiopulmonary coupling. Non-rapid eye movement (NREM) sleep is associated with decreased sympathetic activity, a decrease in heart rate, and a decrease of average blood pressure and blood pressure variability in comparison to the wake state. Respiratory sinus arrythmia is a phenomenon in which the heart rate variability is synchronized beat-to-beat with respiration and is most pronounced during deep NREM sleep. In contrast, during rapid eye movement (REM) sleep there is dominance of sympathetic control and a burst frequency of sympathetic activity that is actually higher than during wakefulness leading to increases in blood pressure variability and heart rates similar to what is seen during wakefulness. Using spectral analysis, a frequency of 0.1 Hz and above has been associated with parasympathetic activity dominance and frequencies below 0.1 Hz have been associated with dominance of sympathetic activity.
# Cardiopulmonary coupling methodology
Cardiopulmonary sleep spectrograms were first obtained from a single lead ECG. ECG-derived respiration (EDR) is obtained either by using R-S wave amplitudes or variations in QRS complexes area. Several studies have looked at improving the accuracy of deducing EDR from single lead ECG and reducing noise but is beyond the scope of this paper. In parallel to extracting the EDR, ectopic beats are identified and removed and normal sinus-normal sinus (NN) intervals are extracted and outliers are filtered. After extracting the N-N interval series on ECG and its associated EDR, the signals are then resampled using cubic splines at 2 Hz. The Fast Fourier Transform is applied to 3 overlapping 512 sample sub-windows within the 1,024 coherence window. The 1,024 coherence window is then advanced by 256 samples (2.1 min) and the calculation repeated until the entire N-N interval/EDR series is analyzed. Thus, the cross-spectral power and coherence of these two signals are calculated over a 1,024 sample (8.5 min) window.
For each 1,024 window the product of the coherence and cross-spectral power is used to calculate the ratio of coherent cross power in the low frequency (0.01-0.1 Hz.) band to that in the high frequency (0.1-0.4 Hz.) band. The logarithm of the high to low frequency cardiopulmonary coupling ratio [log (HFC/LFC)] is then computed to yield a continuously varying measure of cardiopulmonary coupling (Al. While originally the ECG signal was used as input, any signal or signal set which encodes respiration and heart rate variability may be used to compute the CPC sleep spectrogram.shows the steps in computing CPC.
## Distinct cardiopulmonary coupling patterns in sleep
## Sleep stages and cyclic alternating pattern
The conventional characterization of sleep stages dictates a "graded" approach to NREM sleep, from lightest (N1) to deepest (N3). The difference between N2 and N3 are relatively arbitrary, dependent on the proportion of high amplitude slow waves (20% threshold) for a given epoch. However, there is a great variability of depth of NREM sleep, as can be readily objectively demonstrated by techniques such as the Odds Ratio Product. A unique and key feature of CPC sleep states is poor correlation of HFC and LFC with conventional NREM sleep stages. Thus, in health, the majority if N2 is also HFC, N3 is usually HFC but at times LFC, while N1 is always LFC. There is a moderate correlation with a well-described stability dimension of NREM sleep, Cyclic Alternating Pattern (CAP). CAP is a distinct pattern that can be seen on electroencephalography (EEG) during unstable NREM sleep. High frequency coupling dominates when CAP is sparse or absent, while LFC is reliably associated with CAP. Conventional NREM stage N3 is usually HFC, but so is the majority of healthy N2, where non-CAP periods also dominate. Thus, CAP and CPC capture significantly overlapping domains of NREM sleep stability while both measures correlate only partially with conventional measures of sleep depth.
## Slow wave (delta) power
Slow-wave power in the sleep EEG has highly characteristic spatial and temporal evolution patterns across a night. Power in the 1-4 Hz frequencies dominates the first half of the night, but the ebb and flow of slow-wave power continues throughout the night. High frequency coupling strongly covaries with slowwave power across the whole night, while low frequency power in heart rate variability is inversely related to EEG delta power. One interesting finding when aligning HFC with delta power is that there is a consistent lag of delta power after HFC where HFC usually precedes an increase in delta power by an average of 6 min, suggesting that subcortical/brain-stem mechanisms may lead large-scale cortical synchrony during sleep.
## Sleep blood pressure
A key dimension of health is a reduction of blood pressure during sleep (blood pressure "dipping"). Sleep blood pressure is known to dip during stage N3, and rise during REM sleep, yet the majority of sleep is N2, and dipping profiles occupy the entire sleep period, not just N3-enriched zones. This discrepancy has been solved by correlating blood pressure during sleep with stable sleep as determined by CPC (high frequency coupling). Blood pressure dipping occurs only during HFC periods. In a randomized trial targeting sleep apnea treatment in patients with cardiovascular risk factors, it was shown that those who were treated with CPAP had more HFC during sleep, which was in turn associated with improvement in blood FIGURE 2 | Photoplethysmogram/oximetry-based CPC-heart rate analysis. The oximetry-based analysis provides a full CPC-sleep spectrogram, an apnea-hypopnea index by integrating CPC LFC and oxygen desaturation events, and a profile of heart rate across the night. In the upper segment of the figure, "dipping" of heart rate is noted along with abundant high frequency coupling/stable sleep. In the lower sample, there is less stable state, but the heart rate profile is distinctly abnormal, with an elevation even during stable state. Such relative tachycardia during stable NREM sleep may suggest obstructive hypoventilation. In both examples, oxygen desaturation itself is mild. Stable and unstable sleep (HFC and LFC, respectively) occur intermittently through the night. HFC, high frequency coupling; LFC, low frequency coupling; e-LFC, elevated low frequency coupling; HR, heart rate; CHVR, cyclic heart rate variation.
Frontiers in Neuroscience | www.frontiersin.org pressure dipping and mean arterial pressures during sleep.
## Autonomic regulation during sleep
There is normally a reduction of the heart rate (HR) during sleep, a HR-dip, which roughly follows the blood pressure dipping pattern. Heart rate during sleep is, however, far simpler to measure than blood pressure and provides a window into cardiovascular health. By aligning heart rate profiles with the CPC spectrogram , unique insights and cardiovascular risk profiles are potentially extractable, and can be tracked over time. Normally, HR dipping occurs during HFC periods.
## Vertically integrated multi-component sleep states
Cardiopulmonary coupling analysis established that sleep is bimodal than graded. That is, while conventional NREM sleep stages moves across the N1 to N3 grades, the CPC-spectrogram FIGURE 3 | The oximeter-extracted CPC spectrogram. The basic graphical representation of the CPC-spectrogram has high, low, and very low frequency coupling (HFC, LFC, and VLFC, respectively) components. Actigraphy is integrated, and VLFC without movement is considered REM sleep, whereas VLFC with movement is Wake. Cyclic variation of heart rate is also displayed, as well as e-LFC as a measure of sleep fragmentation. The oximeter signal itself provides standard oximetry metrics, such as an oxygen desaturation index. As shows, periods of HFC and LFC alternate throughout the night. LFC, low frequency coupling; VLFC, very low frequency coupling; HFC, high frequency coupling; CHVR, cyclic heart rate variation. shows that NREM sleep has only two distinct and completely non-overlapping forms-stable and unstable (HFC and LFC, respectively), which intermittently switch across the entire night. While N3 dominates in the first half of the night, HFC occurs throughout . This bimodality or stability domain is especially clear when incorporating autonomic and respiratory variables with electrocortical activity, specifically, delta power and the < 1 Hz slow oscillation. Stable NREM is characterized by high probability of occurrence of the < 1 Hz slow oscillation, high delta power, non-CAP EEG, stable breathing, blood pressure dipping, strong sinus arrhythmia and vagal dominance, and high frequency CPC. Conversely, unstable NREM exhibits opposite features: a fragmented and discontinuous < 1 Hz slow oscillation, CAP patterns on the EEG, non-dipping of blood pressure, unstable respiration, cyclic variation in heart rate, and low frequency CPC.
## Cardiopulmonary coupling in sleep apnea
## Diagnosis of sleep apnea
Sleep apnea reliably induces strongly coupled low-frequency oscillations in heart rate and respiration. This results in strong ECG or PPG amplitude fluctuations, besides cyclic variation in heart rate, enabling computing an AHI. This computation requires knowing the number of oxygen desaturation events, the amount of time in coupled low-frequency oscillations, the mean frequency of these computed oscillations, and the total sleep period. The first step in using CPC for sleep apnea detection involves a second-levels analysis of the LFC zone, where a spectral band designated as elevated-LFC (e-LFC) was found which correlated highly with scored apneas and hypopneas. Within e-LFC, two further patterns were discernable, one with a wide dispersion of coupling spectra and another with a narrow band of coupling spectra (broad and narrow-band e-LFC, or e-LFC BB and e-LFC NB ). However, other causes of sleep fragmentation may also cause similar patterns, especially e-LFC BB , a limitation which may be minimized by integrating oxygen saturation fluctuations into the computation. In two recent large studies combining CPC AHI with oximetry desaturation index events in one index have improved the accuracy of derived AHI in comparison to PSG AHI. This derived AHI was approved to be equivalent to PSG AHI in adults and children in 2019 by the FDA (K182618).
These LFC NB and LFC BB indices have been used in several studies in the adult and pediatric populations for automated detection of sleep apnea, section A). There are several advantages for using CPC through wearable devices, especially the current embodiment of a ring-form oximeter, in the sleep apnea population. These include: (1) cost-effective screening of high risk adult and pediatric populations; (2) minimizing patient (wearing) and system (scoring) burdens; (3) detection of expressed high loop gain (central apnea and periodic breathing), which can be a risk stratification approach, as such patients are at risk for treatment-emergent central sleep apnea, reduced adherence to therapy, and persistent respiratory instability during apnea therapy.
## Sleep apnea treatment effects
Successful sleep apnea treatment is expected to increase HFC relative to LFC, including following oral appliance therapy and upper airway surgery. A similar pattern is noted in pediatric patients with OSA after adenotonsillectomy. The same results are seen with CPAP treatment of OSA. Successful treatment of OSA with CPAP is associated with improvement in HFC/LFC ratio.randomized patients with heart failure (ejection fraction less than 45%) who had moderate to severe OSA to CPAP vs. usual care. After 1 month the CPAP treated group showed an increase in HFC compared to the control group.looked at CPAP titration studies and defined successful CPAP titration and optimum CPAP pressures as AHI less ≤ 5 / h of sleep for 30 min during supine REM; higher HFC was found in successful CPAP PSGs and higher LFC in unsuccessful titrations.
## Endotyping and phenotyping sleep apnea
Endotypes are the mechanisms which drive pathology, while phenotypes are the expression of these endotypic effects. Multiple driver endotypes are now recognized as important in the pathogenesis of obstructive sleep apnea, including high loop gain, low arousal threshold, airway collapsibility, impaired negative pressure response, and sleep fragmentation resulting in amplified wake-sleep transitional instability. Thus, what is considered "obstructive sleep apnea" can be caused by one or more of the above driving mechanisms, which can be classified into anatomical and non-anatomical. High loop gain, reflecting respiratory control instability and an imbalance between input (oxygen and carbon dioxide levels) and output (neural drive to respiratory muscles and upper airway) of the respiratory system, is perhaps the most important non-anatomical endotype. When loop gain is more than 1, self-sustained oscillations are inevitable. The Frontiers in Neuroscience | www.frontiersin.org importance of high loop gain is that treatment failure risk is high and options such as oxygenand acetazolamidecan be beneficial. Though mathematical methods can accurately quantify endotypes, analysis of the expressed phenotypes can also accurately identify sleep apnea with high loop gain.
When high loop gain is manifested, the polysomnographic patterns include classic central sleep apnea, periodic breathing, complex apnea with codominant loop gain and airway pathology, treatment-emergent central sleep apnea, and NREMdominant obstructive sleep apnea. A common theme across all these conditions is self-similar (metronomic timing, identical morphology of consecutive events) of respiratory abnormality, which induce e-LFC NB , which is a marker of this expressed high loop gain;. In a study of 671 subjects with sleep apnea which compared CPC indices to conventional PSG scoring, e-LFC NB was associated with respiratory instability during CPAP titration. Since e-LFC NB is a marker of expressed high loop gain and "central" sleep apnea,evaluated if could be used as a marker of adaptive servoventilation titration success in 106 patients with complex sleep apnea. Overall ASV titration success as defined as AHI < 10/h on ASV was found in 81% of patients and no correlation was found between percentage of LFC NB and ASV titration success. One limitation of this study was the use of opiates, which causes ataxic breathing, and is unlikely to cause the exact self-similarity needed to induce e-LFC NB .
## Cardiopulmonary coupling in other sleep disorders
CPC has been used to study other sleep disorders beyond OSA. Patients with insomnia have been shown to exhibit increased LFC even in the absence of sleep disordered breathing. It appears that LFC BB is the main LFC pattern seen in pure insomnia so the coexistence of LFC NB should raise the suspicion for coexisting sleep apnea.studied CPC in a group of primary insomnia and compared to a group of good sleepers. They found increased LFC and a lower HFC/LFC ratio among the insomnia group .studied insomnia patients with cognitive impairment and found decreased HFC indicating predominance of unstable sleep compared to insomnia patients with normal cognition. However,found that improvements in some sleep parameters in insomnia patients subjected to 6 weeks of cognitive behavior therapy was associated with decreased HFC. One of the limitations of this study was absence of control group. A systematic review of cardiovascular autonomic activity in insomnia patients showed that increased LFC/HFC ratio is a consistent finding in those patients. Similar findings of increased LFC/HFC ratio were also seen in CPC studies of populations with conditions that would predispose them to secondary insomnia/short sleep durations including: sleep deprivation, fibromyalgia, and periodic limb movement disorder.
Since insomnia is common in patients with uncontrolled psychiatric disorders; CPC could be used in studying and tracking treatment response in such patients. In comparison to controls, patients with untreated major depression have reduced HFC and increased LFC.studied 41 patients with depression and showed that the increase in HFC following 2 weeks of antidepressant medications treatment was associated with improvement in psychiatric questionnaire scores and suggested that this can be used to predict early response to treatment in such patients.
# Conclusion
The CPC sleep spectrogram provides a novel window into sleep physiology and key information about sleep during health and disease. Because such data can be obtained from simple/reduced and even contactless signal acquisition methods, it allows studying sleep in greater numbers, and with greater ease, in a wider range of conditions, with nearly limitless repeatability, than typically possible with traditional polysomnograms or current home sleep apnea testing devices.
# Author contributions
HA and RT wrote the manuscript. All authors contributed to study design and the literature search, and revised the manuscript.
|
The clinical efficacy of relocation pharyngoplasty to improve retropalatal circumferential narrowing in obstructive sleep apnea patients
Lateral pharyngeal wall appears to be a critical culprit of obstructive sleep apnea (OSA) subjects and relocation pharyngoplasty has been expected to be a promising surgical option to correct retropalatal circumferential narrowing in OSA patients. The purpose of our study is to evaluate the therapeutic outcomes of relocation pharyngoplasty and its clinical effectiveness in OSA patients with retropalatal circumferential narrowing. We performed relocation pharyngoplasty combined with nasal surgery in 133 OSA patients with the following characteristics: apnea-hypopnea index (AHI) over 10, retropalatal circumferential narrowing greater than grade I when awake, and redundant soft tissue around the lateral pharyngeal wall. The analysis of surgical success rate was performed with the data of 68 subjects who underwent pre and postoperative polysomnography. The objective success rate of relocation pharyngoplasty was 52.9%, and significant reduction of mean AHI with improvement of lowest SpO2 was seen in 69% of patients 3 months after the surgery. The median AHI was decreased from preoperative 37.3 to postoperative 21.4. Median lowest SpO2 changed from 78.4 to 84.1%. Total sleep time, daytime sleepiness, and visual analogue scale for snoring showed improvement as well. Postoperative complications including pain or bleeding were minimal in 133 subjects and a few patients complained of subtle taste loss. Our data demonstrate that relocation pharyngoplasty can be a favorable surgical option fighting against retropalatal circumferential narrowing.Obstructive sleep apnea (OSA) is a common sleep disorder characterized by upper-airway collapse that causes reduction or cessation of airflow during sleep. Clinically, abnormal anatomy in the upper airway is obvious in OSA patients and reduced airflow from the nasal cavity and narrowing of the upper airway increase negative pressure in the pharyngeal airway and predispose the pharynx to collapse 1,2 . Both upper airway narrowing and increased airway resistance reportedly contribute to the underlying pathogenesis of OSA, leading to sleep-related symptoms such as loud snoring, apnea, fatigue, daytime sleepiness and systemic complications if not properly treated 3-8 . OSA occurs due to fixed or dynamic upper-airway obstruction caused by anatomical factors or abnormal upper-airway motor tone, and upper-airway obstruction can be caused by collapse at multiple levels, such as the soft palate, uvula, palatine tonsils, lateral pharyngeal walls, and base of the tongue 5,9 . A palatal pattern of collapse is most frequent and numerous surgical techniques have been designed to modify the palate anatomy in OSA patients 10 . Palatal surgeries for OSA aim to correct the pharyngeal tissues narrowing the upper airway, enhance the tension of the pharyngeal muscle, and widen the pharyngeal lumen. Multiple studies have demonstrated the clinical benefits of palatal surgeries, including relief from both subjective symptoms and life-threatening conditions in OSA patients, and the addition of new surgical options can potentially improve the success rate of management of OSA patients with palatal narrowing or collapse in their upper airway[11][12][13][14][15].The lateral pharyngeal wall is a complex structure composed of numerous pharyngeal muscle groups, such as the palatopharyngeus, superior pharyngeal constrictor, and palatoglossus muscles in addition to lymphoid tissue,
Lateral pharyngeal wall appears to be a critical culprit of obstructive sleep apnea (OSA) subjects and relocation pharyngoplasty has been expected to be a promising surgical option to correct retropalatal circumferential narrowing in OSA patients. The purpose of our study is to evaluate the therapeutic outcomes of relocation pharyngoplasty and its clinical effectiveness in OSA patients with retropalatal circumferential narrowing. We performed relocation pharyngoplasty combined with nasal surgery in 133 OSA patients with the following characteristics: apnea-hypopnea index (AHI) over 10, retropalatal circumferential narrowing greater than grade I when awake, and redundant soft tissue around the lateral pharyngeal wall. The analysis of surgical success rate was performed with the data of 68 subjects who underwent pre and postoperative polysomnography. The objective success rate of relocation pharyngoplasty was 52.9%, and significant reduction of mean AHI with improvement of lowest SpO2 was seen in 69% of patients 3 months after the surgery. The median AHI was decreased from preoperative 37.3 to postoperative 21.4. Median lowest SpO2 changed from 78.4 to 84.1%. Total sleep time, daytime sleepiness, and visual analogue scale for snoring showed improvement as well. Postoperative complications including pain or bleeding were minimal in 133 subjects and a few patients complained of subtle taste loss. Our data demonstrate that relocation pharyngoplasty can be a favorable surgical option fighting against retropalatal circumferential narrowing.
Obstructive sleep apnea (OSA) is a common sleep disorder characterized by upper-airway collapse that causes reduction or cessation of airflow during sleep. Clinically, abnormal anatomy in the upper airway is obvious in OSA patients and reduced airflow from the nasal cavity and narrowing of the upper airway increase negative pressure in the pharyngeal airway and predispose the pharynx to collapse. Both upper airway narrowing and increased airway resistance reportedly contribute to the underlying pathogenesis of OSA, leading to sleep-related symptoms such as loud snoring, apnea, fatigue, daytime sleepiness and systemic complications if not properly treated. OSA occurs due to fixed or dynamic upper-airway obstruction caused by anatomical factors or abnormal upper-airway motor tone, and upper-airway obstruction can be caused by collapse at multiple levels, such as the soft palate, uvula, palatine tonsils, lateral pharyngeal walls, and base of the tongue. A palatal pattern of collapse is most frequent and numerous surgical techniques have been designed to modify the palate anatomy in OSA patients. Palatal surgeries for OSA aim to correct the pharyngeal tissues narrowing the upper airway, enhance the tension of the pharyngeal muscle, and widen the pharyngeal lumen. Multiple studies have demonstrated the clinical benefits of palatal surgeries, including relief from both subjective symptoms and life-threatening conditions in OSA patients, and the addition of new surgical options can potentially improve the success rate of management of OSA patients with palatal narrowing or collapse in their upper airway.
The lateral pharyngeal wall is a complex structure composed of numerous pharyngeal muscle groups, such as the palatopharyngeus, superior pharyngeal constrictor, and palatoglossus muscles in addition to lymphoid tissue, including the palatine tonsils. Retropalatal circumferential narrowing due to lateral pharyngeal wall collapse has been documented to be a structural cause in the pathogenesis of OSA but its clinical significance has been underestimated in OSA patients undergoing palatal surgery. The lateral pharyngeal wall is more collapsible or thicker in severe OSA patients than in normal subjects or patients with mild OSA, and retropalatal circumferential narrowing appears to be the sole independent risk factor in OSA patients. Complete retropalatal circumferential narrowing may be closely related to higher apnea-hypopnea index (AHI) scores, and higher lateral pharyngeal collapsibility is seen in OSA patients who show a relapse of snoring or apneic events after surgery. Surgical approaches that alter the retropalatal circumferential narrowing and maintenance of pharyngeal intensity to lateral dimensions appear to be critical for palatal surgeries in OSA patients with retropalatal circumferential narrowing. Therefore, diverse surgical techniques to reduce retropalatal circumferential narrowing and increase the tension and stability of the lateral pharyngeal wall have been introduced.
Lateral pharyngoplasty, expansion sphincter pharyngoplasty (ESP) and relocation pharyngoplasty have been introduced to correct retropalatal circumferential narrowing in OSA patients. Relocation pharyngoplasty is most recently introduced surgical method to correct retropalatal circumferential narrowing and is assumed to be an effective technique for creating tension in the lateral pharyngeal walls with reduction of the number of apneic events or snoring intensity. In the present study, we aimed to determine the therapeutic outcomes of relocation pharyngoplasty in patients with mild, moderate, or severe OSA and investigate the clinical efficacy of relocation pharyngoplasty in OSA patients with retropalatal circumferential narrowing.
# Methods
Ethics statement. One hundred thirty-three subjects diagnosed with OSA at the department of Otorhinolaryngology at Seoul National University Hospital from March 2015 to December 2018 were recruited, and retropalatal circumferential narrowing in these subjects was confirmed through drug-induced sleep endoscopy (DISE). Written informed consent was obtained from each participant and the study complied with the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board of Seoul National University Hospital (IRB number1801-084-915).
Patients and study design. All subjects underwent relocation pharyngoplasty, including tonsillectomy, a uvulopalatal (UP) flap procedure, and nasal surgeries to improve sleep-related symptoms and abnormal sleep parameters. Indications of relocation pharyngoplasty were (1) an AHI score greater than 10 events/hr on polysomnography (PSG), (2) retropalatal circumferential narrowing above DISE grade I (more than 50% narrowing), and (3) a narrowed oropharynx due to lateral pharyngeal collapse of the redundant soft tissue around the posterior pillar.
Patients' medical records, including the results of pre-or postoperative PSG, were reviewed retrospectively. Total sleep time, AHI (events/hr), and lowest O 2 saturation were observed before and at 6 months after the operation. A successful procedure was defined as 50% reduction in AHI and an AHI < 20, as described by Sher.
We also analyzed the therapeutic outcome of relocation pharyngoplasty with overall success and response rates which are defined commonly in the literature. "Success" was defined as being AHI < 20 and ≥ 50% reduction in AHI and "improved" was defined as > 50% reduction in postoperative AHI.
Duration of hospital stay, early complications, primary outcome parameters such as the Epworth Sleeping Scale (ESS) and the visual analogue scale (VAS) for subjective snoring or apnea, and late complications were recorded at discharge from the hospital and during follow-up visits.
Surgical technique for relocation pharyngoplasty. The surgical technique of relocation pharyngoplasty in the current study was based on the method which was originally developed by Li et al.. OSA subjects were selected for relocation pharyngoplasty depending on the presence of oropharyngeal collapse, as determined by a complete preoperative upper airway examination and DISE using the VOTE classification, which is determined by anatomical analysis of the velum, oropharynx, tongue base and epiglottis. The patients did not present with obstruction at the tongue base or epiglottis. All subjects complained of nasal obstruction, and were scheduled to receive septoturbinoplasty to obtain improved nasal airflow. The surgery was performed under general anesthesia with the patient in the Rose position and nasotracheal intubation. Bilateral tonsillectomy and the UP flap procedure were performed first. After indicating the design with a marking pen on the soft palate above the uvula, mucosa was removed with sharp pointed scissors or Bovie cautery and only uvular mucosa was carefully removed to prevent damage to the uvular muscle. The uvular tip was then turned over onto the soft palate, and the abundant uvula tip was cut properly with Bovie cautery. The palatoglossus (PG), palatopharyngeus (PP), and superior pharyngeal constrictor (SPC) muscles were identified following the UP flap procedure. The PP muscle www.nature.com/scientificreports www.nature.com/scientificreports/ was divided from the SPC muscle and posterior pillar mucosa to facilitate tension-free approximation. The PP muscle near the uvula was then rotated cephalad and laterally to approach the counterpart of the soft palate. The SPC muscle was grasped and sutured laterally to the ipsilateral PG muscle in a cephalic-to-caudal direction with Vicryl #2.0 (Ethicon, Somerville, NJ) through a mattress-style suture. The suture usually required two or three separated stitches to splint the SPC muscle in one tonsillar fossa. Mucosal closure between the anterior and posterior pillars used Vicryl #3.0. The tension of the suture on the flap was adjusted to prevent postoperative wound dehiscence. The same procedure was repeated on the opposite side. All subjects were discharged 2 days after relocation pharyngoplasty, and the last follow-up visit was performed 3 months after the surgery.
Statistical analysis. Statistical calculations were performed using SPSS 19.0 (SPSS, IBM). Postsurgical changes in AHI, lowest oxygen saturation, ESS value, and VAS for snoring or apnea were analyzed using a paired t-test and the Wilcoxon signed-rank test. Descriptive data are presented as mean ± standard deviation. Differences are considered significant if p < 0.05.
# Results
Demographic data of subjects. We recruited 133 subjects who were diagnosed with OSA and underwent relocation pharyngoplasty to resolve airway narrowing and reduce retropalatal circumferential narrowing. Of them, 117 patients were men, and 16 were women. The mean age was 43.4 years (range, 21-56 years), and the mean body mass index was 26.6 kg/m 2 (range, 16-34.2 kg/m 2 ). The severity of OSA was based on AHI: 23 patients demonstrated mild OSA, 52 showed moderate OSA, and 58 showed severe OSA. We did not recommend relocation pharyngoplasty to OSA patients with an AHI below 10 events/hr even if they showed lateral pharyngeal www.nature.com/scientificreports www.nature.com/scientificreports/ collapse on DISE. We performed relocation pharyngoplasty in OSA patients with grade I or larger tonsils and 78.9% of subjects exhibited grade II or higher tonsil grade. In addition, 81.9% of subjects showed a palatal grade above II. Based on preoperative DISE findings, all subjects showed circumferential narrowing at the retropalatal level and the mean grade of retropalatal narrowing was 2.3 ± 0.7 . Sixty-eight subjects had PSG prior to and after relocation pharyngoplasty and preoperative PSG findings revealed that the mean AHI was 35.0 ± 20.7 events/hr, mean lowest O 2 saturation was 77.9 ± 12.1%, mean total sleep time was 393.7 min, and mean REM sleep percentage was 21.6%.
Therapeutic outcome of relocation pharyngoplasty. The therapeutic outcome was determined using AHI scores from PSG, which was performed at 6 months after operation (N = 68). The mean AHI for entire subjects was 35.0 ± 20.7 events/hr prior to relocation pharyngoplasty, which decreased significantly to 21.4 ± 17.6 postoperatively. Selecting an arbitrary threshold of a 50% reduction in AHI and an AHI < 20, the number of responders was 36, for a success rate of sleep surgery (including relocation pharyngoplasty) of 52.9% in OSA patients with retropalatal circumferential narrowing. Overall success and response rates are analyzed with "success" being a ≥50% reduction in AHI and a postoperative AHI < 20, and "improved" being a ≥ 50% The overall success and response rates, with "success" being AHI < 20 and ≥ 50% reduction in AHI, and "improved" being postoperative AHI ≥ 20 and reduction > 50%, "failed" being AHI reduction under <20%. (2020) 10:2101 | https://doi.org/10.1038/s41598-020-58920-9
www.nature.com/scientificreports www.nature.com/scientificreports/ reduction in AHI. This resulted in 69.1% of patients being assigned to the success and improved group. In particular, moderate OSA patients showed higher successful outcomes following relocation pharyngoplasty (68.2%) compared with mild or severe OSA patients with retropalatal circumferential narrowing.
As a next step, we classified these 68 subjects depending on tonsil grade and compared therapeutic outcome of relocation pharyngoplasty according to tonsil size of subjects with an arbitrary threshold of a 50% reduction in AHI and an AHI < 20. The results showed that success rate of relocation pharyngoplasty in subjects with grade I tonsils (N = 28) was 35.7%, in subjects with grade II (N = 25) was 64.0%, in subjects with grade III (N = 11) was 54.5%, and in subjects with grade IV (N = 4) was 100%. We found that success rate of sleep surgeries including relocation pharyngoplasty was relatively higher in OSA patients with grade II, III, and IV compared to the OSA subjects with grade I tonsils.
Postoperative PSG results also showed that the lowest O 2 saturation improved significantly from 78.46 ± 9.5% to 84.08 ± 6.4%. In addition, postoperative total sleep time was prolonged from 390.2 ± 69.4 to 411.9 ± 54.7 minutes. In the subjective symptoms, the mean ESS value improved from 8.77 ± 3.7 to 5.76 ± 3.3, and the sleep efficiency and REM sleep percentage improved significantly in 68 subjects after surgery .
Mucosal wound dehiscence was observed in 12.7% (N = 133) of subjects after relocation pharyngoplasty combined with the UP flap procedure. Partial dehiscence was observed in 11 subjects, and total dehiscence was observed in 6. We did not suture mucosal dehiscence, and the occurrence of dehiscence was not correlated with the failure rate of relocation pharyngoplasty. The mean admission period was 4.01 ± 0.4 days, and tracheostomy was not performed in any subjects following relocation pharyngoplasty.
We investigated complications or side effects following relocation pharyngoplasty: postoperative pain, uncomfortable sensation, taste loss, velopharyngeal insufficiency, voice change, and postoperative bleeding. We did not observe serious complications related to relocation pharyngoplasty at 1 week, 1 month, or 3 months after surgery. In total, all 133 subjects complained of pain at 1 week after relocation pharyngoplasty, but only 4 patients still experienced oropharyngeal pain at 1 month after relocation pharyngoplasty. No patient experienced postoperative pain at 3 months after relocation pharyngoplasty. In addition, 24 patients felt an abnormal sensation around the soft palate area 1 week after surgery, but only 4 patients complained of it at 3 months postoperatively. Some patients complained of velopharyngeal insufficiency, voice change and taste loss after relocation pharyngoplasty, but those complaints had subsided by at 1 month. We observed postoperative bleeding in 5 patients at the oropharynx after relocation pharyngoplasty and postoperative bleeding was well controlled in all patients through conservative management without the need for electrocautery or bleeder ligation.
# Discussion
Through this study, we found that relocation pharyngoplasty can be an adequate surgical option to improve upper airway narrowing at the retropalatal level and provide positive therapeutic outcomes for OSA patients with retropalatal circumferential narrowing in the upper airway. Our current clinical results also suggest possible surgical indications for relocation pharyngoplasty in OSA patients: PSG results, DISE, and endoscopic findings such as a narrowed oropharynx due to tonsil enlargement (>grade I), lateral bulk of soft tissue around the posterior pillar, and circumferential narrowing (DISE grade > I) at the retropalatal level.
Lateral pharyngeal wall collapse contributes to the pathogenesis of OSA by increasing airway resistance and causing partial or complete obstruction of the upper airway. Retropalatal circumferential narrowing due to lateral pharyngeal wall collapse is more closely related to airway resistance and causes aggravated hypoxic tissue damage in OSA patients. In OSA patients, the lateral pharyngeal wall is more collapsible when pressured by airflow than in healthy subjects and the lateral pharyngeal wall of OSA patients can be thicker than in normal subjects, making it a predominant anatomic factor inducing upper-airway collapse in OSA patients. In addition, OSA patients with retropalatal circumferential narrowing exhibited higher AHI and respiratory disturbance index scores than OSA patients with anteroposterior upper airway narrowing. This suggests a need for an adequate surgical option to reduce retropalatal circumferential narrowing in OSA patients and adequate surgical reduction of retropalatal circumferential narrowing provide a satisfactory therapeutic outcome in OSA patients. Therefore, accurate evaluation of retropalatal circumferential narrowing and the maintenance of tension or stability of the lateral pharyngeal wall may be critical in sleep surgeries for OSA patients who show circumferential narrowing of upper airway. www.nature.com/scientificreports www.nature.com/scientificreports/ Uvulopalatopharyngoplasty (UPPP) is a popular technique in the field of sleep surgery. However, because conventional UPPP widens the retropalatal space from anterior to posterior, it appears to be ineffective in OSA patients with multiple-level obstructions, and therapeutic outcomes of UPPP are unsatisfactory in OSA patients with retropalatal circumferential narrowing. Diverse surgical techniques to replace UPPP have been introduced to correct lateral pharyngeal wall collapse in OSA subjects. Lateral pharyngoplasty, relocation pharyngoplasty, and ESP have been suggested as effective surgical procedures to correct retropalatal circumferential narrowing. These surgical techniques are associated with better improvement of AHI scores than UPPP and more effectively widen the pharyngeal lumen by reducing collapse of the lateral pharyngeal wall during sleep. We usually recommend relocation pharyngoplasty and ESP to OSA patients with retropalatal circumferential narrowing to improve lateral pharyngeal wall tension. Although lateral pharyngoplasty produced better clinical and PSG outcomes in OSA patients than did UPPP, cross-sectional measurements of the pharyngeal airway did not differ between treatments. In addition, many patients experienced significant dysphagia following lateral pharyngoplasty because it is an invasive process that undermines the superior pharyngeal constrictor muscle, and lateral pharyngoplasty exhibited more complications, such as oronasal reflux of liquid and wound dehiscence, than did relocation pharyngoplasty and ESP.
## Pre-op [mean (sd)] post-op [mean (sd)] difference [mean (sd)] p-value
Relocation pharyngoplasty appears to be similar to lateral pharyngoplasty in splinting the lateral pharyngeal wall but has the distinctive characteristic of plicating the superior pharyngeal constrictor muscle to prevent scarring contraction of the tonsillar fossa. Advancement of the soft palate is achieved by removing the supratonsillar mucosa and adipose tissue to create a space for reconstruction of the velopharynx. In addition, this procedure preserves the palatal muscle by removing only part of the soft palate mucosal layer. Our study included patients with mild, moderate, or severe OSA who showed more than 50% narrowing in their lateral pharyngeal wall with tonsil enlargement, and we expected them to need more tension to improve lateral pharyngeal wall collapse. Based on our results, relocation pharyngoplasty resulted in approximately 53% successful surgical results in OSA patients and provided a higher success rate in moderate OSA patients, with 69% of OSA patients showing an improved AHI score and superior numerical changes in sleep parameters following relocation pharyngoplasty. We conclude that relocation pharyngoplasty is a favorable surgical option that maintains pharyngeal function in moderate OSA patients with retropalatal circumferential narrowing.
It has been reported that relocation pharyngoplasty has obvious advantages in splinting the lateral pharyngeal wall. Relocation pharyngoplasty does not undermine the SPC muscle and avoids the possibility of damaging manor neurovascular structures. Our data showed that OSA patients who underwent relocation pharyngoplasty complained of minimal complications and most complications were resolved within 1 month of relocation pharyngoplasty.
Maintenance of muscle tension after pharyngeal suture in relocation pharyngoplasty may be critical for successful surgical outcomes in OSA patients with lateral pharyngeal wall collapse. We presume that the tension that relocation pharyngoplasty provides in the lateral pharyngeal wall is not as powerful as that provided by ESP, which repositions the underlying muscular structures of the pharynx and palate to widen the lateral pharyngeal airway. After revealing the arching fibers of the palate muscles, isolated PP muscle is attached to the arching fiber of the palate muscle. ESP thus creates more powerful tension in the lateral pharyngeal wall and reduces the www.nature.com/scientificreports www.nature.com/scientificreports/ bulk of lateral pharyngeal soft tissue by isolating and rotating the PP muscle super-anterolaterally. In a previous study, we found ESP combined with uvuloplasty to be an effective surgical option for lateral pharyngeal collapse in patients with severe OSA with minimal complications. Based on our clinical data, relocation pharyngoplasty also provided a high cure rate for OSA patients with lateral pharyngeal collapse and effectively reduced lateral pharyngeal bulk, resulting in enhanced soft-palate tension or widening of the pharynx without severe complications after surgery. Furthermore, subjective symptoms related to OSA, such as snoring, apnea, and daytime sleepiness, also improved significantly after relocation pharyngoplasty. The lowest oxygen saturation and valid sleep times also improved, and the extent of the retropalatal area in the upper airway widened significantly after relocation pharyngoplasty. We therefore estimated that relocation pharyngoplasty may be more useful for moderate OSA patients with greater than grade I circumferential narrowing at the retropalatal level and greater than grade I tonsil enlargement. However, we still suggest ESP for OSA patients who exhibit more severe lateral pharyngeal collapse of grade II or higher and a narrowed oropharynx due to bulky soft tissue in the lateral pharyngeal wall.
In summary, our clinical findings confirm that relocation pharyngoplasty is a useful surgical option in OSA subjects with retropalatal circumferential narrowing due to lateral pharyngeal wall collapse. They also supply evidence supporting the favorable indications of relocation pharyngoplasty and provide improved therapeutic outcomes for patients with moderate OSA and relatively less severe retropalatal circumferential narrowing.
|
Evidence of antagonistic predictive effects of miRNAs in breast cancer cohorts through data-driven networks
FiguresS1Figure 1. Comparison between Lasso and SWAG on the validation dataset [1]. We compare accuracy, sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV) of the lasso estimates with the ranges (i.e. smallest-to-largest value intervals) of the same metrics for the 112 SWAG models. We do not present the related 95% percentile bootstrap confidence intervals for the lasso estimates since they are not informative (i.e. mostly giving [0, 1] intervals). All evaluations have been made out-of-sample (i.e. on a balanced validation dataset) at the standard 0.5 cut-off of logistic regression. . ROC curve comparison between Lasso and SWAG on the validation dataset [bib_ref] Integrated analysis reveals microRNA networks coordinately expressed with key proteins in breast..., Aure [/bib_ref]. We present the ROC curve of lasso (in red) with the ROC region (in gray) produced by the 112 SWAG models. We obtain the ROC region for the set of SWAG models considering first all the 112 individual model ROC curves and then coloring in gray the area which enclose all the 112 ROC curves at the same time.. List of all the 45 miRNAs selected by the SWAG. hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-155 hsa-miR-21
hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-320d
hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-486-5p
hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-193a-5p hsa-miR-21
hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-193b* hsa-miR-21
hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-378 We present in bold the miRNAs whose single and associative β signs are confirmed in both the AHUS (primary) and validation dataset. S1 . We present the single (i.e. the estimated value of a β coefficient when considering a single miRNA in the logistic model) and associative (i.e. the different values that a miRNA specific β coefficient takes in each of the SWAG models in which it is present) coefficients (median values and range) obtained on the validation dataset [bib_ref] Integrated analysis reveals microRNA networks coordinately expressed with key proteins in breast..., Aure [/bib_ref] for the antagonistic miRNAs present in at least 10% of the models of the AHUS dataset [bib_ref] Subtype-specific micro-RNA expression signatures in breast cancer progression, Haakensen [/bib_ref] (primary study). We present in bold the miRNAs whose antagonistic role is confirmed in both the AHUS (primary) and validation dataset.
[formula] hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-328 hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-429 hsa-let-7c hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-622 hsa-miR-103 hsa-miR-1274a hsa-miR-139-3p hsa-miR-21 hsa-miR-30b hsa-miR-103 hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-497 hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-320c hsa-miR-320d hsa-miR-125b-2* hsa-miR-1274a hsa-miR-21 hsa-miR-328 hsa-miR-92a hsa-miR-103 hsa-miR-125b-2* hsa-miR-1274a hsa-miR-328 hsa-miR-342-3p hsa-miR-1274a hsa-miR-130b hsa-miR-139-3p hsa-miR-181b hsa-miR-21 hsa-miR-1274a hsa-miR-130b hsa-miR-181b hsa-miR-21 hsa-miR-21* hsa-miR-1181 hsa-miR-125b-2* hsa-miR-139-3p hsa-miR-328 hsa-miR-342-3p hsa-miR-125b-2* hsa-miR- [/formula]
[table] Table 4: S1Table 6. We present the single (i.e. the estimated value of a β coefficient when considering a single miRNA in the logistic model) and associative (i.e. the different values that a miRNA specific β coefficient takes in each of the SWAG models in which it is present) coefficients (median values and range) obtained on the validation dataset[1] for the eight most frequently selected miRNAs of the AHUS dataset[2] (primary study). [/table]
[table] miRNA Single β: Median associative β Associative β Range [/table]
|
Cyber War Game in Temporal Networks
In a cyber war game where a network is fully distributed and characterized by resource constraints and high dynamics, attackers or defenders often face a situation that may require optimal strategies to win the game with minimum effort. Given the system goal states of attackers and defenders, we study what strategies attackers or defenders can take to reach their respective system goal state (i.e., winning system state) with minimum resource consumption. However, due to the dynamics of a network caused by a node's mobility, failure or its resource depletion over time or action(s), this optimization problem becomes NP-complete. We propose two heuristic strategies in a greedy manner based on a node's two characteristics: resource level and influence based on k-hop reachability. We analyze complexity and optimality of each algorithm compared to optimal solutions for a small-scale static network. Further, we conduct a comprehensive experimental study for a large-scale temporal network to investigate best strategies, given a different environmental setting of network temporality and density. We demonstrate the performance of each strategy under various scenarios of attacker/defender strategies in terms of win probability, resource consumption, and system vulnerability.
# Introduction
Many natural and man-made systems can be modeled as complex networks consisting of nodes and links representing the interactions between nodes. One of the most important property of a network is robustness against random failures and target attacks [bib_ref] Error and attack tolerance of complex networks, Albert [/bib_ref] [bib_ref] Resilience of the Internet to random breakdowns, Cohen [/bib_ref] [bib_ref] The structure and function of complex networks, Newman [/bib_ref] [bib_ref] Complex networks: Structure and dynamics, Boccaletti [/bib_ref] [bib_ref] Increasing network resiliency by optimally assigning diverse variants to routing nodes, Newell [/bib_ref] , measured by the giant connected component size after perturbations. The percolation threshold is the fraction of non-removed nodes (or links) leading to the collapse of the network [bib_ref] Resilience of the Internet to random breakdowns, Cohen [/bib_ref] , which is often predicted by using percolation theory, a method from statistical physics. Increasing evidence shows that networks interact to each other, resulting in a new research area on interdependent networks [bib_ref] Catastrophic cascade of failures in interdependent networks, Buldyrev [/bib_ref] [bib_ref] Interdependent Networks: Reducing the Coupling Strength Leads to a Change from a..., Parshani [/bib_ref] , interconnected networks, multiplex [bib_ref] Diffusion dynamics on multiplex networks, Gomez [/bib_ref] , multilayer networks, and a network of networks [bib_ref] Robustness of a network of networks, Gao [/bib_ref] [bib_ref] Networks formed from interdependent networks, Gao [/bib_ref]. Indeed, these systems can not only model interactions between different networks, but also consider a temporal network [bib_ref] Temporal networks, Holme [/bib_ref] in which a network topology changes over time. Understanding vulnerability of these systems helps design interdependent robust infrastructures.
Unlike engineering systems, vulnerability of temporal, mobile networks can be modeled as the cyber games where the attackers intend to compromise users and and the defenders will recover the compromised users to healthy state under the nodes' resource restriction such as battery life, computational power, and/or the network's limited bandwidth [bib_ref] Dynamic Defense Strategy against Advanced Persistent Threat with Insiders, Hu [/bib_ref]. Furthermore, an entity often requires decision making based on local information in a fully distributed way and aims to take optimal strategies to maximize resource efficiency (e.g., complete a task with minimum effort) when achieving respective goals. For instance, an attacker compromises more healthy nodes to disrupt a system while a defender recovers compromised nodes to secure the system. Although many existing approaches consider cyber war games by proposing optimal strategies of attackers and defenders [bib_ref] Dynamic Defense Strategy against Advanced Persistent Threat with Insiders, Hu [/bib_ref] [bib_ref] Stackelberg-game analysis of correlated attacks in cyber-physical systems, Zhu [/bib_ref] [bib_ref] Boosting Markov Reward models for probabilistic security evaluation by characterizing behaviors of..., Zhang [/bib_ref] , they do not consider optimal strategies with minimum resource consumption in temporal networks.
An attacker-defender cyber game has been explored with various approaches such as game theory [bib_ref] Stackelberg-game analysis of correlated attacks in cyber-physical systems, Zhu [/bib_ref] or cognitive theory. Zhu and Martinez [bib_ref] Stackelberg-game analysis of correlated attacks in cyber-physical systems, Zhu [/bib_ref] model a cyber game using a twolevel Stackelberg game (leader-follower) to consider a node's inherent resource constraints in discrete-time, linear time-invariant networks. Recently, Ben-Asher and Gonzalezpropose a decision making framework using an instance-based learning technique considering dynamics of a cyber war where multiple attackers and defenders play to maximize their utility. In distributed cyberspaces, however, a network suffers from resource constraints and faces high dynamics under varying network temporality and density. In this work, we question a fundamental problem: how does the network temporality affect the performance of attackers or defenders in a cyber war game under resource-constrained, distributed network environments?
To answer this question, we aim to identify optimal strategies of attackers or defenders that allow a winning in a cyber game with minimum resource consumption in a time-varying, distributed network. In this environment, each node has a limited resource and its resource level is updated over time or upon taking actions. Due to the distributed nature of a network, a node may use local information to make decisions and often can take actions towards its adjacent nodes. That is, attackers or defenders may select a node to compromise or recover among their adjacent nodes, respectively. Considering these challenges derived from the unique characteristics of a given network environment, this optimization problem is not solvable in a polynomial time and known as a NP-complete problem. This work has the following unique contributions:
1. We consider resource efficiency of cyber strategies taken by attackers or defenders in a resource-constrained, distributed network environment where each attacker or defender can make a decision based on local information without the knowledge of global network (e.g., network topology) and node information (e.g., remaining energy).
2. We consider a time-varying network such as structural and state dynamics and study how they affect the optimal strategies. Structural dynamic refers to network topology changes that may be caused by node mobility or failure or terrains while state dynamic means resource depletion over time or upon action(s). To consider structural dynamics, we introduce a new influence metric called k-hop influence based on the concept of k-hop reachability. For state dynamics, we consider dynamic adjustment of each node's status. Both dynamics affect decisions by attackers or defenders to reach their respective system goal state.
3. We conduct comprehensive performance analysis of the proposed strategies by attackers or defenders which studies the impact of network temporality and density on our performance metrics such as a win probability, minimum resource consumption, and system vulnerability.
## System model
In this section, we explain our network model, node model, and system failure condition considered in this work.
## Network model
We consider a temporal network whose topology changes over time. In our model, at an initial time we generate a random arbitrary network using a given degree distribution. Every time step we randomly select p fraction of link and rewire the nodes between these links randomly. When p = 0, the network is static; when p = 1, we generate a new random network independent from the previous step. In addition, nodes' resource level depletes with more actions and over time. Given a network with a directed graph GðtÞ ¼ ðV; EðtÞ : WðtÞÞ at time step t, VðtÞ is a set of vertices, representing nodes (or entities) and EðtÞ is a set of edges, representing connectivity between two vertices. Depending on the existence of an edge between two nodes, i and j, the weight w ij (t) can be in WðtÞ, i.e., w ij ðtÞ 2 WðtÞ. A given network environment is characterized by: (1) it is highly distributed where each node can communicate with or take an action towards its adjacent node(s); (2) it is severely resource-constrained where a node may drain its resource (i.e., battery life or reliability) over time to maintain normal operations even without interacting with other nodes or may consume resource when it takes an action towards any adjacent node; and (3) it is time-varying, dynamic in terms of network temporality (i.e., changing network topology) and remaining resource level of nodes.
We use epidemic spreading based on susceptible-infected-removed (SIR) model [bib_ref] Complex networks: Structure and dynamics, Boccaletti [/bib_ref] [bib_ref] Heterogeneity in susceptible-infected-removed (SIR) epidemics on lattices, Neri [/bib_ref] to describe attackers' compromising behaviors and defenders' recovering behaviors. If a node is recovered from being compromised, it is immune to the attack. Thus, in terms of a node's lifetime except original attackers or defenders seeded in the network deployment, a user node only experiences one time to be compromised by attackers or recovered by defenders. A network is initialized with three types of players including attackers, defenders, and users based on their state (see [fig_ref] Fig 1: Fig 1 [/fig_ref]. We assume that a node is equipped with a host-based intrusion detection mechanism [bib_ref] Network intrusion detection, Mukherjee [/bib_ref] , characterized by probabilities of false positives and false negatives, denoted as P fp and P fn , respectively. Each node i is capable of extrapolating neighboring nodes j's resource level based on their activities and signal strength. For each node to be aware of partial or complete network topology, a node broadcasts its neighbor information to the network. Obtaining Composition of nodes and their dynamic status. OA for original attackers, Q for quarantined original attackers, OU for original users which have never compromised or recovered before, UA for compromised users becoming attackers, UD for recovered users becoming defenders, OD for original defenders, and RE for resource exhausted. All nodes except the quarantined attackers are regarded as legitimate member nodes and can become resourceexhausted, resulting in non-legitimate members. Where N ¼ jCðtÞj þ jDðtÞj þ jAðtÞj þ jI AðtÞj, we can derive DðtÞ ¼ ODðtÞ [ UDðtÞ, CðtÞ ¼ OAðtÞ [ UAðtÞ, AðtÞ ¼ UðtÞ, and I AðtÞ ¼ QðtÞ [ REðtÞ at time t > 0. a global network topology requires all nodes' neighboring information which is heavily expensive in resource-constrained environments (e.g., wireless mobile networks). To mitigate this overhead, we introduce the concept of k-hop reachability, by limiting the geographic area of disseminating information of adjacent nodes.
We consider two types of dynamics associated with a node's characteristics in terms of structural dynamic and state dynamic as follows:
1. Structural dynamic is reflected based on a node's temporal location in a network, related to the work on network of networks [bib_ref] Robustness of a network formed by n interdependent networks with a one-to-one..., Gao [/bib_ref] [bib_ref] Percolation of a general network of networks, Gao [/bib_ref] [bib_ref] Vulnerability and controllability of networks of networks, Liu [/bib_ref]. We investigate a node's influence based on the concept of k-hop reachabilityin a given time-varying network and employ it as criteria for an attacker's or defender's decision to select a node to take an action (i.e., compromising or recovering a node). We explain the computation of a node's influence in Eq (3) below.
2. State dynamic is considered in terms of a node's resource level representing battery life and/ or reliability. Structural and state dynamics may evolve at the same time [bib_ref] Emergence of structural and dynamical properties of ecological mutualistic networks, Suweis [/bib_ref] [bib_ref] Novel type of phase transition in a system of self-driven particles, Vicsek [/bib_ref] [bib_ref] Collective Motion in a Network of Self-Propelled Agent Systems, Peng [/bib_ref]. The structural and state dynamics are interwoven and affect to each other, the so called dynamics of mutualistic interactions [bib_ref] Emergence of structural and dynamical properties of ecological mutualistic networks, Suweis [/bib_ref]. For example, species abundance affects network rewiring while network structure determines the species abundance. In addition, the coevolution of state and structural dynamics leads to the nestedness of real mutualistic networks [bib_ref] Novel type of phase transition in a system of self-driven particles, Vicsek [/bib_ref].
Other examples can be observed in a collective motion of self-propelled particle systems [bib_ref] Novel type of phase transition in a system of self-driven particles, Vicsek [/bib_ref] or a network of self-propelled agent systems [bib_ref] Collective Motion in a Network of Self-Propelled Agent Systems, Peng [/bib_ref]. In our work, a network topology partially relies on state dynamic while the state dynamic fully depends on structural dynamic.
## Node model
Recall that a given cyber war game is played by attackers, defenders, and users. When an attacker compromises a user, the user is compromised and becomes an attacker who is capable of attacking another healthy user (i.e., a node which has not compromised in the past). Unless the compromised user is recovered by a defender, it remains as compromised. A defender is a node with the capability to recover a compromised user or to quarantine an original attacker. If a compromised user is recovered by the defender, it is immune to the attack such as SIR [bib_ref] Complex networks: Structure and dynamics, Boccaletti [/bib_ref] [bib_ref] Epidemic spreading in scale-free networks, Pastor-Satorras [/bib_ref] [bib_ref] Non-Markovian Infection Spread Dramatically Alters the Susceptible-Infected-Susceptible Epidemic Threshold in Networks, Mieghem [/bib_ref]. If the recovered node is an original attacker, it is quarantined and cannot perform any attack while it does not have a capability to recover another node. The defender is robust against attacks and will never be compromised by an attacker in which the defenders are trusted entities. Winning a given cyber war depends on whether or not attackers or defenders reach their system goal state, respectively. Given an initial number of nodes N in a network, the network has the following four types of nodes at time t based on their state:
## Compromised nodes (cðtþ) include original attackers or compromised users;
2. Healthy active user nodes (AðtÞ) are users who have never compromised in the past;
3. Defenders (DðtÞ) indicate original defenders or recovered users; and 4. Inactive nodes (I AðtÞ) are dead nodes due to lack of resources or original attackers being quarantined.
The total number of nodes initially given N can be derived as N ¼ jCðtÞj þ jDðtÞj þ jAðtÞj þ jI AðtÞj. We summarize the network node composition and dynamic status of nodes in [fig_ref] Fig 1: Fig 1 [/fig_ref] In a given network, a certain fraction of nodes are compromised by outside attacker(s). We study how quickly attackers or defenders reach their respective goal state from the initial state as shown in [fig_ref] Fig 1: Fig 1 [/fig_ref] As time elapses, the compromised nodes start compromising other legitimate member nodes based on their attack strategy to reach the system failure state based on System Failure Condition (SFC) (see a next subsection, System Failure Condition). We discuss attacker and defender strategies in Section Attacker and Defender Strategies later. As some nodes become compromised, defenders detect them and start performing the recovery process of compromised nodes to prevent or mitigate system failure by eliminating all compromised nodes from the system.
Next we represent the characteristics of a node as a vector by:
[formula] v i ðtÞ ¼ ½r i ðtÞ; d ðkÞ i;in ðtÞ; d ðkÞ i;out ðtÞ Tð1Þ [/formula]
where r i (t) is node i's resource level at time t and d ðkÞ i;in ðtÞ and d ðkÞ i;out ðtÞ indicate the in-degree and out-degree of node i within k-hop distance from itself at time t, respectively. The out-degree of node i with a given k indicates the concept of reachability, i.e., how many nodes are reachable from node i in a network. The in-degree of node i with k-hop distance means how many nodes can reach node i within k-hop distance. We use k-hop reachabilityto mitigate the computation or communication overhead to exchange neighbors information. We use the k-hop distance in-degree and out-degree of node i to calculate its influence in the network.
Each node disseminates adjacent nodes information in order to provide a global view of the network. To mitigate high communication overhead, it disseminates neighbors information within a k-hop distance. Given an adjacency matrix, WðtÞ, for a directed graph GðtÞ ¼ ðVðtÞ; EðtÞÞ, matrix L (k) (t) consists of elements l k ij ðtÞ with a binary value 0 or 1 representing that node j is reachable from node i within k-hop distance. L (k) (t) is computed based on the shortest path algorithm for a directed graph considered [bib_ref] Efficient algorithms for shortest paths in sparse networks, Johnson [/bib_ref].
Based on L (k) (t) calculated above, let D (k) (t) be a 2 × n matrix for the in-degree and outdegree of n number of nodes based on k-hop distance. D (k) (t) is denoted as each element with d ðkÞ i;in ðtÞ and d ðkÞ i;out ðtÞ, which are calculated based on L (k) (t) with elements l k ij ðtÞ for all i and j by:
[formula] d ðkÞ i;in ðtÞ ¼ X n j¼1;l k ij 6 ¼1 l k ij ðtÞ; d ðkÞ i;out ðtÞ ¼ X n j¼1;l k ji 6 ¼1 l k ji ðtÞ ð 2Þ [/formula]
The degree of a node's influence is used as one of criteria attackers or defenders take actions to minimize the accumulated resource consumption until they reach the respective system goal state. A node's influence is calculated by:
[formula] I k i ðtÞ ¼ d ðkÞ i;out ðtÞ d ð1Þ i;in ðtÞð3Þ [/formula]
I k i ðtÞ implies a node's influence over other nodes in a network with a given k-hop distance compared to other nodes' influence over the node itself. A node with high influence, I k i ðtÞ, means that the node has high influence over others while it is not much influenced by other nodes. For simplicity, we did not include time unit t in the equations above but the influence may be affected by the dynamics of a network topology which was examined in the simulation experiments by varying network temporality.
We consider the state r i (t) of a node i changes over time according to its incoming neighboring nodes, strategy chosen, and whether an action is taken or not. r i (t) is i's remaining resource level at time t. Based on e i (t) above, node i's remaining resource level, r i (t), is updated as: e i (t) counts the cost only when node i chooses node j to take an action; 0 otherwise. The above implies that when node j has a high resource level, an attacker or a defender needs to consume more resource to take an action towards node j. λ is a constant parameter to adjust the speed of the resource consumption per action. e i (t) implies node i consumes more resource to take an action towards node j with higher resource level. Note that node i takes an action only when
[formula] r i ðtÞ ¼ r i ðt À 1Þ À s À e iðtÞr i (t − 1) − σ − e i (t)>0. [/formula]
When node i selects node j to take an action, node i's action is effective towards node j with a probability by:
[formula] s ij ðtÞ ¼ min r j ðtÞ r i ðtÞ ; 1ð6Þ [/formula]
s ij (t) implies that when node j has high resource, node i's action is less likely to be effective, vice-versa. When r i (t) = 0, it means node i dies due to the lack of resource. This node is not part of legitimate members in the network, and accordingly the total number of active nodes at time t, N ðtÞ, decreases. IAðtÞ increments as more inactive nodes exist in the network. In this work, each attacker or defender can compromise or recover one adjacent node at a time, not allowing actions towards multiple nodes simultaneously.
## System failure condition
The goal of attackers is to reach the system state to failure. To model the attackers' target state based on the system failure state, we define the system failure condition (SFC) in terms of the loss of two system security goals: (1) loss of integrity based on the concept of Byzantine Failurewhere the system fails with too many compromised entities (e.g., the system with more than one-third of participating entities being compromised), leading to increased attack severity due to collusive attack; and (2) loss of availability based on the fact that the system does not have a sufficient number of healthy, active nodes for mission execution. Some nodes may die due to lack of resources while other nodes may be compromised due to node capture attack by attackers. Therefore, the SFC is defined by:
[formula] SFC ¼ 1 if jCðtÞj N ðtÞ ! r 1 _ jCðtÞj þ jI AðtÞj N ! r 2 0 otherwise; 8 < :ð7Þ [/formula]
where jCðtÞj is the total number of compromised nodes at time t, N ðtÞ refers to the number of active nodes at time t regardless of their status, either compromised or healthy. jCðtÞj þ jI AðtÞj indicates the total number of inactive nodes including original attackers quarantined plus dead user nodes due to lack of resource. Where N is the total number of nodes that are initially given, ρ 1 bounds the maximum number of compromised nodes that can exist without failure while ρ 2 is the fraction of the maximum number of inactive nodes that can exist without failure in the network.
## Cyber war game
This section discusses how the cyber war game is formulated as an optimization problem. In addition, we describe attacker and defender strategies proposed in this work and analyze their solution complexity.
## Problem formulation
We formalize this problem as an optimization problem that minimizes accumulated resource consumption J until the system goal state reaches by solving the following objective function as:
[formula] Minimize J ¼ Z T t¼0 X i2MðtÞ e i ðtÞdt [/formula]
Subject to w ij ðtÞ > 0; r i ðtÞ À e i ðtÞ > 0
Here MðtÞ is a set of nodes belonging to a party (i.e., either attackers or defenders) where MðtÞ includes a set of nodes taking actions to reach a respective system target state. MðtÞ is same as CðtÞ for attackers while it is DðtÞ for defenders in [fig_ref] Fig 1: Fig 1 [/fig_ref] For attackers, e(t) is a vector of the resource consumed by attackers successfully where e(t) = [e 1 (t), . . ., e i (t), . . ., e m (t)] T and m ¼ jCðtÞj and e i (t) represents resource consumed by node i to compromise another node in UðtÞ at time t (See Eq (5) for e i (t)). Similarly, for defenders, e (t) is a vector of the resource consumed by the defenders to successfully recover compromised nodes in CðtÞ, where e(t) = [e 1 (t), . . ., e i (t), . . ., e m (t)] T and m ¼ jDðtÞj and e i (t) indicates node i's resource consumption to recover a compromised node at time t. This problem is to identify a set of nodes by which attackers or defenders take actions to reach their respective goal state while minimizing resource consumption. Recall that attackers or defenders can only take actions towards their adjacent neighbors (i.e., 1-hop neighbor).
In Eq (8) above, a small amount of resource decay over time (i.e., σ) without any additional activity (e.g., compromising or recovering actions) is omitted. The imposed constraints are: (1) node i can take an action towards node j only when w ij (t)>0 which means there is a directed edge from node i to node j; and (2) node i should have sufficient resource to take an action towards node j (i.e., r i (t) − e i (t)>0).
## Attacker and defender strategies
In this section, we discuss what strategies attackers or defenders can take to win a cyber war game with minimum resource consumption, respectively.
Node i, either attacker or defender, selects an adjacent node j (it should be originally a user) with minimum resource consumption to compromise or recover node j while reaching the target state as quickly as possible. We propose two heuristic strategies based on a node's two characteristics as follows: (1) a node's resource level; and (2) a node's influence based on k-hop reachability as shown in Eq (3), called k-hop influence in this work. Therefore, each node i can have two strategies to select adjacent node j to take an action as follows:
1. Resource-First (RF): node i selects node j with the minimum resource among all adjacent nodes.
2. Influence-First (IF): node i selects node j with the maximum influence among all adjacent nodes.
Both strategies above have the goal to win a game with the minimum resource consumption by either minimizing resource consumption in each step or maximizing a chance to reach the goal state with minimum time where both strategies aim to minimize the accumulated resource consumption until the end state. We denote attackers' two strategies as Resource-First-Attack (RF-A) and Influence-First-Attack (IF-A). Similarly, defenders' strategies are notated as Resource-First-Defense (RF-D) and Influence-First-Defense (IF-D). In all cases, if node i's expected resource consumption by taking an action towards node j exceeds its current remaining resource in Eq (4), node i does not take any action towards node j to save its resource for its own survival.
# Solution complexity analysis
In this section, we analyze the solution complexity of three algorithms: optimal solution using brute-force algorithm (BFA) based on depth-first-search, resource-first (RF) and influencefirst (IF). In particular, we analyze the strategies taken by attackers and how their strategies affect the resource consumption. In order to find feasible solution space, we relax some conditions defined in this work. The time-varying network condition is relaxed by using a static network to identify optimal solution using BFA. We assume that a static network has nodes connected to other nodes with a weight e ij where e ij is computed based on e i in Eq (5) where an edge exists between nodes i and j. Note that e ij is updated whenever node i takes an action with any adjacent node. For example, i's resource will be updated as it takes actions towards adjacent nodes j's. Accordingly, e ij for all adjacent nodes j's is affected by node i's resource adjustment. In order to minimize the effect of randomness, we consider s ij = 1. Note that we remove all the relaxed conditions used above and show simulation results for a large time-varying, dynamic network later in a next section.
In this section, we analyze the complexity of solution search in three algorithms including BFA, RF and IF. We analyze the complexity in terms of attackers' perspective where the attackers require compromising more nodes to reach their goal state as defenders recover the compromised nodes over time. We approximate the complexity of solution search algorithms where a graph has n vertices and each vertex has an average of m out-degrees.
1. Brute Force Attack (BFA): we simply calculate the combination of choosing c out of n where n is the total number of nodes and c is the number of nodes required to be compromised to meet SFC. Since the loss of integrity failure is a more tight condition than the loss of availability failure unless many nodes quickly drain their resources, we treat c as the minimum number of compromised nodes to make the system failure in this case. This is computed by selecting c number of compromised nodes out of n which is the initial number of nodes given, denoted as C(n, c) = O(n2 n ) where c n/2 (i.e., ρ 1 1/2) and n! = o(n n ). Since each node computes this brute-force solution, resulting in O(n2 n ) and there is the overhead to obtain global network topology, O(n 3 ), the complexity of brute-force optimal solution is O (n2 n +n 3 ), leading to O(n2 n ).
## Resource-first-attack (rf-a)
: it is linear proportional to n as each attacker chooses one among multiple adjacent nodes based on the minimum resource. Given an initial number of attackers c 0 , and the average out-degrees m, the maximum round of compromising actions by all attackers, h, each attacker compromises another healthy node. It is estimated by:
2 0 mc 0 þ 2 1 mc 0 þ 2 2 mc 0 þ :::
[formula] þ 2 h mc 0 ¼ X h i¼0 2 i mc 0 < h2 h mc 0 where 2 h mc 0 < n < hn ¼ OðnÞ for h << n [/formula]
Therefore, RF-A has a complexity of O(n 2 ) where each node runs O(n).
## Influence-first-attack (if-a):
it is similar to RF-A for the compromising process, O(n 2 ), but it has the overhead to compute k-influence, leading to O(n 3 ).
# Experiments and analysis
We show the results and analyze their trends under two network environments: (1) a static network to identify optimal solutions; and (2) a temporal network by varying network temporality p and nodes' average degree d.
## Metrics
The following performance metrics are used:
1. Win probability (P w ) refers to attackers' average win probability. Attackers win when SFC is met; defenders win when there exist no compromised nodes in a network.
## Resource consumption(j )
is the average accumulated resource consumed by either attackers or defenders until time T which is the time that they attain their respective goal based on Eq (8).
3. System Vulnerability (VðtÞ) refers to the degree of the system vulnerability in terms of the number of compromised nodes and the number of inactive nodes, as addressed in SFC. When VðtÞ ! 1, this indicates the system failure state. This is estimated by: n
[formula] VðtÞ ¼ max jCðtÞj r 1 N ðtÞ ; jCðtÞj þ jI AðtÞj r 2 Nð9Þ [/formula]
where r 1 N ðtÞ is the maximum number of compromised nodes allowed in the system without failure at time t and jCðtÞj is the number of compromised nodes in the system at time t. r 2 N indicates the maximum number of members that are not committing for mission execution and is the maximum bound tolerated by the system. jCðtÞj þ jI AðtÞj is the number of inactive or compromised members currently in the network.
## Result analysis under a small-scale static network
Experimental Setting. Since it is not feasible to obtain optimal solution(s) of a given problem under a time-varying network consisting of a large number of nodes (i.e., NP-Complete), we first validate the optimality of the given problem by comparing the three algorithms in a static network consisting of 20 nodes. In the next section below, we will discuss results under temporal networks. We use an environment setting with λ = 0.05, σ = 0.001 (i.e., in resource calculation in Eq (4)), and k = 2 (i.e., in k-hop influence in Eq (3)). Resource levels of nodes are assigned as a real number ranged in [0.5, 1] based on the uniform distribution in order to consider vastly different resource levels of nodes in a network. We set up the network based on Erdös-Rényi (ER) model with a different probability, q, given for two nodes to be randomly connected in the network deployment. To make the given network have vastly different degrees in a directed network, we randomly select a pair of nodes to remove an edge (i.e., when edges exist between nodes i and j in both directions such as i to j or vice-versa, we keep one edge while removing the other edge).
Result. [fig_ref] Fig 2: Optimality Analysis of Attack Strategies in a Static Network [/fig_ref] we show the minimum resource consumed by three algorithms including two strategies and one optimal solution based on BFA. In order to demonstrate the optimality of the given problem, we use a simplified cyber war scenario. The scenario is that an attacker compromises user nodes while a defender may recover compromised user nodes. Thus, in [fig_ref] Fig 2: Optimality Analysis of Attack Strategies in a Static Network [/fig_ref] we show the number of compromised nodes as x-axis while plotting the resource consumption in y-axis. Each strategy shown here is a strategy taken by an attacker, such as brute-forceattack (BFA), resource-first-attack (RF-A), or influence-first-attack (IF-A).
As shown in [fig_ref] Fig 2: Optimality Analysis of Attack Strategies in a Static Network [/fig_ref] , although BFA performs the best consuming the minimum resource among three in both network conditions (i.e., sparse with q = 0.4 and dense with q = 0.7), it occurs prohibitively high overhead for solution search as shown in [fig_ref] Fig 2: Optimality Analysis of Attack Strategies in a Static Network [/fig_ref]. For the other two strategies, RF-A or IF-A, we notice IF-A becomes outperforming RF-A as attackers compromise more nodes particularly under a sparse network. Lastly we experiment the impact of varying the total number of nodes, N , in a network. [fig_ref] Fig 2: Optimality Analysis of Attack Strategies in a Static Network [/fig_ref] , we investigate the impact of N on resource consumption of the two strategies. We set the number of compromised nodes to N Â 2=3. In this case, IF-A outperforms RF-A as a network size becomes larger. In a network with higher node density but less network density (i.e., sparse with q = 0.5 in this case), attackers prefer IF-A over RF-A in compromising a large number of nodes in the network.
## Result analysis under a large-scale temporal network
Experimental Setting. For a large-scale temporal network, we use a random network based on ER network model where network temporality p is used as a rewiring probability that two nodes i and j are connected at time t. We consider the total number of nodes, N ¼ 1000 where the nodes consists of the initial number of attacker, |OA| = 1, the initial number of defenders, |OD| = 50, and the initial number of users, |OU| = 949. We set λ = 0.05, σ = 0.001, and k = 6. For network environmental conditions, network temporality p and nodes' average degree d (i.e., higher d indicates higher network density) are varied to observe their impact on performance. All data points shown in the results are collected based on 100 of realizations. We summarize all key design parameters, their meanings and corresponding default values in [fig_ref] Table 1: Key design parameters, their meanings and default values [/fig_ref]. For dependent variables, we note dependent under 'Value' in [fig_ref] Table 1: Key design parameters, their meanings and default values [/fig_ref].
Result. [fig_ref] Fig 3: Effect of network temporality [/fig_ref] how network temporality p and network density d affect attackers' win probability. We vary p to see its impact on P w under sparse or dense network with d = 0.5 or d = 2.5, respectively, as shown in [fig_ref] Fig 3: Effect of network temporality [/fig_ref]. In a sparse network of [fig_ref] Fig 3: Effect of network temporality [/fig_ref] , regardless of defenders' strategies, attackers' IF outperforms among others. In addition, higher p in a sparse network helps attackers to win. In a dense network of [fig_ref] Fig 3: Effect of network temporality [/fig_ref] , although attackers' IF outperforms the RF counterpart, when defenders choose RF, attackers have higher chances to win the game. More interestingly, in this dense network, higher network temporality p deteriorates attackers' chances to win. [fig_ref] Fig 3: Effect of network temporality [/fig_ref] shows the effect of varying d on attackers' P w . In all strategy scenarios, higher d does not help attackers to win because higher network density will increase a chance for a node to be recovered by defenders. [fig_ref] Fig 4: Effect of network temporality [/fig_ref] how network temporality and density impact attackers' resource consumption (J ). In a sparse network of [fig_ref] Fig 4: Effect of network temporality [/fig_ref] , higher J occurs as p increases and when attackers take IF strategy. Higher P w leads to higher J because attackers should take more actions. For a dense network of [fig_ref] Fig 4: Effect of network temporality [/fig_ref] , higher P w does not necessarily lead to higher J . This case is shown when attackers use RF (i.e., red and green curves). This is because choosing a node with low remaining resource does not lead to a higher chance to compromise more nodes in next decision rounds. [fig_ref] Fig 4: Effect of network temporality [/fig_ref] , a critical point of d exists in maximizing J . A network with smaller d helps attackers to win quickly due to a less chance to be recovered by defenders in a sparse network. On the other hand, a dense network with high d allows compromised nodes to be easily recovered. Thus, there exists a balance point of d maximizing J .
The trends of defenders' J observed are also very similar to the ones observed in attackers' J in [fig_ref] Fig 4: Effect of network temporality [/fig_ref] Due to the space constraint, we do not show the results here. The reason is that defenders basically follow attackers' actions as they should recover the compromised nodes. However, in our results, defenders' J is significantly lower than attackers' because attackers consume more resource than defenders by taking actions to compromise user nodes. ) for a sparse network vs. a dense network, there exists a critical point that maximizes system vulnerability but three cases out of four do not experience system failure where VðtÞ ¼ 1 implies system failure. That is, although the system has a higher chance to be endangered by system vulnerability, it can survive over time by reducing the vulnerability. However, for a dense network, the system ultimately fails due to a high chance for nodes to be compromised by attackers. On the other hand, comparing
# Conclusion
Given a cyber war game for a resource-constrained, temporal, distributed network, we studied how each party can win the game with minimum resource consumption. We devised two heuristic strategies in a greedy manner based on a node's influence and resource level to maximize a win probability while minimizing resource consumption. We investigated the effect of the proposed heuristic strategies on performance metrics including a win probability, minimum resource consumption, and security vulnerability. We investigated how network temporality and density affect performance of the strategies. We found that attackers' influence-first (IF-A) strategy outperforms resource-first (RF-A) strategy under both sparse and dense networks across a wide range of network temporality. In addition, in a dense network, IF-A consumes less resource than RF-A. Network temporality helps attackers to win a game under a sparse network while it may deteriorate the win probability under a dense network due to a higher chance for them to be recovered by defenders. Although a higher win probability generates higher resource consumption, a certain point of node degree exists to maximize resource consumption. In addition, system vulnerability is significantly affected by network temporality and density because the network characteristics are critical for a node to reach its system goal state with minimum resource consumption.
Our work may raise the following open research questions: how can network temporality be described in real networks?; how can the proposed work be validated with a real network dataset?; If each node is modeled as an agent using game theory, how can Nash Equilibrium be identified in a cyber war game?; and how do heterogeneous network or node characteristics affect optimal strategies of attackers and defenders in a cyber war game?
[fig] Fig 1: Fig 1. Composition of nodes and their dynamic status. OA for original attackers, Q for quarantined original attackers, OU for original users which have never compromised or recovered before, UA for compromised users becoming attackers, UD for recovered users becoming defenders, OD for original defenders, and RE for resource exhausted. All nodes except the quarantined attackers are regarded as legitimate member nodes and can become resourceexhausted, resulting in non-legitimate members. Where N ¼ jCðtÞj þ jDðtÞj þ jAðtÞj þ jI AðtÞj, we can derive DðtÞ ¼ ODðtÞ [ UDðtÞ, CðtÞ ¼ OAðtÞ [ UAðtÞ, AðtÞ ¼ UðtÞ, and I AðtÞ ¼ QðtÞ [ REðtÞ at time t > 0. [/fig]
[fig] Fig 2: Optimality Analysis of Attack Strategies in a Static Network. (a)-(b) For ER networks composed of N = 20 nodes with (a) q = 0.4, and (b) q = 0.7, we plot the resource consumption as a function of the number of compromised nodes for three different strategies. (c) Number of feasible solution as a function of the number of compromised nodes for two different ER networks. (d) Resource consumption as a function of the number of nodes for two different strategies. doi:10.1371/journal.pone.0148674.g002 Lastly, Fig 5 shows how system vulnerability VðtÞ evolves over time under different network temporality and density. Comparing Fig 5(a) & 5(b) and 5(c) & 5(d [/fig]
[fig] Fig 3: Effect of network temporality (p) and density (d) on win probability (P w ). (a) A win probability as a function of temporality for four pairs of strategies by defenders and attackers under a sparse network with average degree d = 0.5. (b) A win probability vs. temporality for four pairs of strategies of defenders and attackers under a dense network with average degree d = 2.5. (c) A win probability as a function of an average degree for four pairs of strategies of defenders and attackers with high network temporality p = 0.5. doi:10.1371/journal.pone.0148674.g003 [/fig]
[fig] Fig 4: Effect of network temporality (p) and density (d) on defenders' and attackers' resource consumption (J ). (a) Resource consumption of defenders vs. temporality under a sparse network with average degree d = 0.5. (b) Resource consumption of defenders vs. temporality under a dense network with average degree d = 2.5 (b). (c) Resource consumption of defenders as a function of average degree for four pairs of strategies of defenders and attackers with high network temporality p = 0.5. (d)-(f) Similar plots for attackers' resource consumption. doi:10.1371/journal.pone.0148674.g004 [/fig]
[fig] Fig 5: Effect of network temporality (p) and density (d) on system vulnerability (VðtÞ). (a) Plot of system vulnerability over time t under dense (d = 2.5) and low temporality (p = 0.05) networks. (b) Plot of system vulnerability over time t for dense (d = 2.5) and high temporality (p = 0.5) networks. (c) System vulnerability as a function of t for sparse (d = 0.5) and low temporality (p = 0.05) networks. (d) System vulnerability over time t under sparse (d = 0.5) and high temporality (p = 0.5) networks. doi:10.1371/journal.pone.0148674.g005 [/fig]
[table] Table 1: Key design parameters, their meanings and default values. Param. Meaning Value N Number of nodes deployed in a network 1000 σ Decay of resource over time to maintain its normal operations ranged in [0, 1] 0.001 λ A constant parameter value to adjust the speed of the resource consumption per action 0.05 P fp , P fn False positives and false negatives probabilities of a host-based IDS preinstalled in each node RE(t) A set of nodes with resource exhausted at time t dependent N ðtÞ A set of active nodes in a network at time t dependent CðtÞ A set of compromised nodes in a network at time t dependent DðtÞ A set of defender nodes in a network at time t dependent IAðtÞ A set of of inactive nodes in a network at time t dependent AðtÞ A set of healthy active nodes in a network at time t dependent v i (t) A vector of a node's state at time t in terms of resource consumption, in-degree and out-degree dependent r i (t) Node i's remaining resource at time t dependent Node i's in-degree using k-hop reachability dependent Node i's in-degree using k-hop reachability dependent e(t) A vector of resource consumed by attackers or defenders taking actions dependent e i (t) Resource consumed when node i takes an action towards node j dependent s ij (t) Probability that node i's action is effective against node j at time t dependent doi:10.1371/journal.pone.0148674.t001 [/table]
|
Fusion primer and nested integrated PCR (FPNI-PCR): a new high-efficiency strategy for rapid chromosome walking or flanking sequence cloning
Background: The advent of genomics-based technologies has revolutionized many fields of biological enquiry. However, chromosome walking or flanking sequence cloning is still a necessary and important procedure to determining gene structure. Such methods are used to identify T-DNA insertion sites and so are especially relevant for organisms where large T-DNA insertion libraries have been created, such as rice and Arabidopsis. The currently available methods for flanking sequence cloning, including the popular TAIL-PCR technique, are relatively laborious and slow. Results: Here, we report a simple and effective fusion primer and nested integrated PCR method (FPNI-PCR) for the identification and cloning of unknown genomic regions flanked known sequences. In brief, a set of universal primers was designed that consisted of various 15-16 base arbitrary degenerate oligonucleotides. These arbitrary degenerate primers were fused to the 3' end of an adaptor oligonucleotide which provided a known sequence without degenerate nucleotides, thereby forming the fusion primers (FPs). These fusion primers are employed in the first step of an integrated nested PCR strategy which defines the overall FPNI-PCR protocol. In order to demonstrate the efficacy of this novel strategy, we have successfully used it to isolate multiple genomic sequences namely, 21 orthologs of genes in various species of Rosaceace, 4 MYB genes of Rosa rugosa, 3 promoters of transcription factors of Petunia hybrida, and 4 flanking sequences of T-DNA insertion sites in transgenic tobacco lines and 6 specific genes from sequenced genome of rice and Arabidopsis.Conclusions:The successful amplification of target products through FPNI-PCR verified that this novel strategy is an effective, low cost and simple procedure. Furthermore, FPNI-PCR represents a more sensitive, rapid and accurate technique than the established TAIL-PCR and hiTAIL-PCR procedures.
# Background
The advent of the genomics era has had a massive impact on almost all fields of biological enquiry. However, it is still necessary to employ chromosome walking and flanking sequence cloning procedures in order to determine full gene structure. Such methods are also required to identify T-DNA insertion sites, and this is particularly relevant for species where large T-DNA insertion libraries exist, eg. rice and Arabidopsis [bib_ref] Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal..., Liu [/bib_ref] [bib_ref] Restriction Site Extension PCR: A Novel Method for High-Throughput Characterization of Tagged..., Ji [/bib_ref] [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref].
Currently, a number of PCR-based methods are available for these purposes, including random PCR, inverse PCR, panhandle PCR, and cassette PCR. However, each of these methods has drawbacks when considered for wide use in the amplification of target regions. For example, inverse PCR is rarely used for chromosome walking due to problems with the availability of restriction sites in the unknown/known regions, and/or poor circularization of the template molecule [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref] [bib_ref] Comparison and critical evaluation of PCRmediated methods to walk along the sequence..., Tonooka [/bib_ref]. Ligationmediated PCR has proved to be too inefficient for routine use in flanking sequence identification, with nonspecific amplified products accounting for the major proportion of the final PCR products [bib_ref] Restriction Site Extension PCR: A Novel Method for High-Throughput Characterization of Tagged..., Ji [/bib_ref] [bib_ref] Comparison and critical evaluation of PCRmediated methods to walk along the sequence..., Tonooka [/bib_ref] [bib_ref] T-linker-specific ligation PCR (T-linker PCR): an advanced PCR technique for chromosome walking..., Yan [/bib_ref].
In recent years, improved methodologies have been developed to address the problems outlined above [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref] [bib_ref] Comparison and critical evaluation of PCRmediated methods to walk along the sequence..., Tonooka [/bib_ref] [bib_ref] T-linker-specific ligation PCR (T-linker PCR): an advanced PCR technique for chromosome walking..., Yan [/bib_ref] [bib_ref] A highthroughput genome-walking method and its use for cloning unknown flanking sequences, Reddy [/bib_ref] [bib_ref] Site Finding-PCR: a simple and efficient PCR method for chromosome walking, Tan [/bib_ref] [bib_ref] An improved PCR method for walking in uncloned genomic DNA, Siebert [/bib_ref] [bib_ref] Improved inverse PCR scheme for metagenome walking, Uchiyama [/bib_ref] [bib_ref] Self-formed adaptor PCR: a simple and efficient method for chromosome walking, Wang [/bib_ref]. The most successful of these involve PCR procedures using random primers, notably, the thermal asymmetric interlaced (TAIL)-PCR method. Such methods are relatively low cost and have the advantage of being relatively direct, i.e. there is no requirement for intermediate steps involving restriction enzyme digestion of the DNA template or ligation reactions [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref] [bib_ref] Comparison and critical evaluation of PCRmediated methods to walk along the sequence..., Tonooka [/bib_ref] [bib_ref] T-linker-specific ligation PCR (T-linker PCR): an advanced PCR technique for chromosome walking..., Yan [/bib_ref]. Consequently, TAIL-PCR has been widely used, notably for the amplification of junction DNA at T-DNA insertion sites [bib_ref] Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal..., Liu [/bib_ref] [bib_ref] Identical promoter elements are involved in regulation of the OPR1 gene by..., He [/bib_ref] [bib_ref] Agrobacterium T-DNA integration in Arabidopsis is correlated with DNA sequence compositions that..., Schneeberger [/bib_ref] [bib_ref] Thermal Asymmetric Interlaced PCR: Automatable Amplification and Sequencing of Insert End Fragments..., Liu [/bib_ref] , the rapid isolation of promoter sequences [bib_ref] A highthroughput genome-walking method and its use for cloning unknown flanking sequences, Reddy [/bib_ref] , and the cloning of full length functional genes [bib_ref] Characterization of the gene encoding glycoprotein C of duck enteritis virus, Liu [/bib_ref]. However, TAIL-PCR is a time consuming procedure that usually involves three rounds of amplification with each round consisting of multiple PCR cycles. The method often produces nonspecific amplification products due to the short arbitrary degenerate (AD) primers and low annealing temperatures, and may also achieve non-efficient amplification of target sequences [bib_ref] Restriction Site Extension PCR: A Novel Method for High-Throughput Characterization of Tagged..., Ji [/bib_ref] [bib_ref] Comparison and critical evaluation of PCRmediated methods to walk along the sequence..., Tonooka [/bib_ref]. Furthermore, conventional TAIL-PCR tends to produce small amplification products [bib_ref] T-linker-specific ligation PCR (T-linker PCR): an advanced PCR technique for chromosome walking..., Yan [/bib_ref] [bib_ref] Evaluation of PCR-based methods for isolating flanking regions of genes, Satyanarayana [/bib_ref] [bib_ref] A Novel and Simple PCR Walking Method for Rapid Acquisition of Long..., Luo [/bib_ref]. TAIL-PCR has subsequently been adapted, but the improved high-efficiency (hi) TAIL-PCR version still remains time consuming, and is expensive in terms of primer design and synthesis. The hiTAIL-PCR technique also poses a great challenge to the activity of DNA polymerase, and so is frequently associated with inefficient amplification results.
Recently, several alternative methods have emerged to perform chromosome walking or flanking sequence cloning. However, such methods still have the drawback that they are either restriction enzyme-dependent, Phi29 DNA polymerase-dependent, or require a series of template DNA purification steps [bib_ref] Restriction Site Extension PCR: A Novel Method for High-Throughput Characterization of Tagged..., Ji [/bib_ref] [bib_ref] A highthroughput genome-walking method and its use for cloning unknown flanking sequences, Reddy [/bib_ref] [bib_ref] A Novel and Simple PCR Walking Method for Rapid Acquisition of Long..., Luo [/bib_ref]. Thus, these alternative methods are still laborious and do not display the wide application potential of conventional TAIL-PCR. Therefore, it is highly desirable that further novel methods are developed for rapid and efficient chromosome walking or flanking sequence cloning.
In order to acquire precise flanking fragments rapidly, we have developed a novel efficient method, namely fusion primer and nested integrated-PCR (FPNI-PCR). We believe that this method has the potential to have wide-spread application in genetics and genome walking studies, and is applicable throughout diverse organisms. In order to demonstrate the power of FPNI-PCR, we performed long PCR walking tests on various genomic targets namely, FT, TFL1 and SOC1 orthologs from the Spiraea cantoniensis, Pyracantha fortuneana, Photinia serrulata, Fragaria ananassa, Rosa hybrida, Prunus mume and Prunus yedoensis species of Rosaceace, 4 MYB genes of Rosa rugosa, 3 promoters from MADSbox transcription factors of Petunia hybrida, and 4 flanking sequences of T-DNA insertion sites in Nicotiana. In each case, successful amplification was achieved of the unknown genomic DNA fragments (> 100 kb sequences). In addition, comparison with genomic sequence databanks confirmed that FPNI-PCR successfully amplified various lengths of the flanking regions of 3 members of the wuschel gene family from Arabidopsis, also the flanking regions of the Osft, Osmads1 and Ostuba1 genes from rice, and an archived flanking sequence from the pCAMBIA2300 vector. Furthermore, a comparative analysis with conventional TAIL-PCR indicated that our novel FPNI-PCR method is more flexible, time saving and powerful than TAIL-PCR or hiTAIL-PCR. Thus, FPNI-PCR has high potential for application in identifying tagged sequences, or for genomic walking from known DNA regions toward unknown regions in organisms with a large genome.
# Results
## Principle of fpni-pcr
The basic principle of FPNI-PCR is outlined in and 2. The method centers on a series of primers encompassing sequence-specific primers (designed on regions of known DNA sequence), and fusion primers which contain an arbitrary degenerate (AD) section fused to a section of determined sequence (fusion primers). The fusion of a known adaptor of determined sequence to the 5'-end of an AD (or other short sitedependent primer) is the main characteristic of FPNI-PCR , and it is this trait which differentiates FPNI-PCR from TAIL-PCR . The FPNI-PCR protocol contains three key steps. In the first step, a large complex mixture of DNA reactions is prepared using 0.4 μL of a gene-specific primer (SP1) designed to the genomic region of known sequence, and 2.0 μL of a combination of nine fusion arbitrary degenerate primers (eg. FP1-9; [fig_ref] Table 1: FP primers [/fig_ref] ; other designed primers for FPNI-PCR are presented in Additional files 1 Tables S1-S5. The details of cycling parameters and PCR conditions used in this study are listed in [fig_ref] Table 2: Cycling parameters and PCR conditions for FPNI-PCR [/fig_ref]. This first step consists of 3-6 repeats of two high stringency cycles followed by a low stringency cycle. Theoretically, single stranded PCR products from the gene-specific primer are generated during the high stringency cycles, and double-stranded products utilizing the FP primers (FPs) are developed during the low stringency cycle. After 3-6 repeated cycles of this PCR regime, it is predicted that the intended target products are partially synthesized and are accompanied by other, nonspecific, products [fig_ref] Figure 2: A general [/fig_ref]. In the second and third steps, nested PCR is conducted using 1 uL of target-specific primers (SP2/ SP3, respectively) and FP-specific primers (FSP1/FSP2, respectively). These steps are high stringency PCR using high annealing temperatures so that target products are selectively amplified. Nonspecific products are not amplified in these steps, in part due to the nested approach with different primers. In addition, the large hairpin structure employed in some nonspecific products also contributes to prevent to further amplification (suppression PCR). Thus, the nonspecific products generated from the first PCR step in FPNI-PCR are not amplified in the second and third steps, and become substantially diluted in the final mix [fig_ref] Figure 2: A general [/fig_ref].
## Parameters in the design and optimization of fpni-pcr
We compared two types of AD primers. In the basic form, i.e. type I primers, the AD primer was fused to the 3' end of an adaptor of known sequence [fig_ref] Table 1: FP primers [/fig_ref]. Type II primers included a hairpin structure at the 5'end of the single-stranded adaptor (Additional files 1, [fig_ref] Table 2: Cycling parameters and PCR conditions for FPNI-PCR [/fig_ref]. We compared the results from such type I and type II primers by testing them in combination with the same gene-specific (SP) primers and under the same cycling parameters. The final results of amplification of the target products indicated no obvious differences between the two primer types, i.e. they presented similar sizes of products on the agarose gels (data not shown). Thus, the hairpin structure at the 5'end of the type II single-stranded adaptors had no apparent impact on amplification of the target. Type III primers describe arbitrarily degenerative primers with 4-6 fixed nucleotides at the 3' end (Additional files 1, [fig_ref] Table 3: Submitted genomic DNA sequences [/fig_ref] , and these primers were also successfully used to amplify the flanking regions of known fragments, with the various sizes of the product fragments relating to the given fusion primer.
The primary round of PCR in the FPNI-PCR procedure includes an optional low stringency PCR cycle (eg. 94°C for 15 s, 28°C for 1 min, ramping to 72°C over a period of 3 min, 72°C for 2.5 min), to follow the two high stringency PCR cycles [fig_ref] Figure 2: A general [/fig_ref]. We tested six different annealing temperatures in these low stringency cycles, using SP1 and various FP primers [fig_ref] Table 1: FP primers [/fig_ref] , and found an effect on the final products obtained (see Additional file 2, Figure. S1 and [fig_ref] Figure 3: The effect of annealing temperature during the low stringency PCR cycles of... [/fig_ref] for example of products obtained during cloning of the FT ortholog of Fragaria ananassa). However, all of the reactions yielded flanking regions and the patterns of amplified fragments after the secondary round of PCR [fig_ref] Figure 3: The effect of annealing temperature during the low stringency PCR cycles of... [/fig_ref] exhibited strong similarities despite the differences in annealing temperatures of the first PCR step. In many cases, lower annealing temperatures in the low stringency cycles of the first PCR step yielded fewer amplified products [fig_ref] Figure 3: The effect of annealing temperature during the low stringency PCR cycles of... [/fig_ref]. This influence of annealing temperature of the low stringency cycle was confirmed by repeated experiments for the cloning of various genes and T-DNA insertion loci. However, if the low stringency cycle of the first FPNI-PCR step was entirely omitted, successful amplification of the target product was still achieved (data not shown).
## Effectiveness and accuracy of fpni-pcr
Using the FPNI-PCR approach, we have successfully isolated 21 complete genomic sequences containing the Schematic outline of the differences in amplification of target and non-target sequences using FPNI-PCR (a) and TAIL-PCR (b). Black and dotted lines represent the target and non-target regions, respectively. Red lines represent the region of known genomic sequence (or known sequence of T-DNA or transposon elements integrated in the genomic DNA). Each red arrow denotes a specific primer designed from the known nucleotide sequence. A blue/green chimeric arrow denotes the FP primer used in the FPNI-PCR method. Each blue arrow denotes a specific primer designed from the known adaptor sequence. In the first PCR step, single stranded copies of the target template are generated in the high stringency cycles, and double stranded products are produced in the low stringency cycle (in total, involving 3-5 repeated PCR cycles); in this primary step, amplification of the target products is likely to be accompanied with other, nonspecific, products. In the secondary and tertiary PCR steps (nested PCR), the target DNA is exponentially amplified by the gene specific and adaptor specific primers, while non-target genes are not amplified because there is no corresponding gene specific primer (and/or amplification was suppressed by the stem-loop structure of the DNA).
FT, TFL1 and SOC1 gene orthologs from seven Rosaceace species, namely Spiraea cantoniensis, Pyracantha fortuneana, Photinia serrulata, Fragaria ananassa, Rosa hybrida, Prunus mume and Prunus yedoensis. In addition, we have used FPNI-PCR to isolate four MYB genes of Rosa rugosa, three promoters from the MADS-Box transcription factor family of Petunia hybrida, and four T-DNA flanking sequences in transgenic tobacco lines. All of these reactions were conducted using a total of just nine FP primers (FPs), in conjunction with specific primers designed to the conserved/partially known genomic fragments or known T-DNA sequences. Sequences have been deposited in the Genbank library, as shown in [fig_ref] Table 3: Submitted genomic DNA sequences [/fig_ref]. [fig_ref] Figure 4: Amplified products after the tertiary round of PCR in genomic walking and... [/fig_ref] shows examples of the products amplified following the tertiary round of PCR in the FPNI-PCR method. [fig_ref] Figure 4: Amplified products after the tertiary round of PCR in genomic walking and... [/fig_ref] shows the products relating to cloning of the FT-Like gene of Rosa rugosa. [fig_ref] Figure 3: The effect of annealing temperature during the low stringency PCR cycles of... [/fig_ref] show the products arising during the cloning of T-DNA flanking sequences using forward and reverse sequence-specific primers, respectively, in three different transgenic tobacco plants and using various FPs. In addition, based on known genomic sequence fragments, we used FPNI-PCR to amplify the various lengths of products from the expected flanking regions of 3 different members of the wuschel gene family of Arabidopsis (Additional file 2, S2a), and sequence from the flanking regions of Osft, Osmads1 and Ostuba1 genes of rice (Additional file 2, [fig_ref] Figure 2: A general [/fig_ref]. Even for the small DNA molecular size of the pCAMBIA2300 vector, the flanking DNA fragment was amplified successfully using FPNI-PCR (data not shown).
Comparative experiments using the same specific primers showed that, in most cases (data not shown), longer products could be obtained by FPNI-PCR than by TAIL-PCR [fig_ref] Figure 5: Comparison of PCR amplification products generated in typical genomic walking experiments using... [/fig_ref]. Furthermore, FPNI-PCR was found to be more reliable as a method of amplifying target genes and unknown flanking sequences. Thus, in a comparative test using several appropriate AD primers, TAIL-PCR frequently failed to amplify the predicted Note: * = Few cycles of thermal asymmetric interlaced PCR: target product generation; incorporation of suppression PCR inverted repeat elements in partial non specific product; ** Standard high stringency PCR: Target Sequence exponential amplification; partial nonspecific products suppression PCR; *** = Standard Nested-PCR: completely eliminating high background or nonspecific amplification in the second PCR. DNA fragment. [fig_ref] Figure 5: Comparison of PCR amplification products generated in typical genomic walking experiments using... [/fig_ref] shows the results of TAIL-PCR in the cloning of the FT ortholog from Pyracantha fortuneana using 3 gene-specific primers and the FP1-9 arbitrary degenerate primers. In this experiment, a 510 bp fragment was obtained from TAIL-PCR with just one (i.e. the fourth) AD primer. By contrast, [fig_ref] Figure 5: Comparison of PCR amplification products generated in typical genomic walking experiments using... [/fig_ref] shows the products generated by FPNI-PCR using the same primers for the same genome walking experiment. Here, four of the FPNI-PCR reactions amplified specific fragments, and the longest fragment was ca. 1.7 kp. Genomic walking for the SOC1-like gene in Pyracantha fortuneana provided another situation in which we encountered "off-target" results using TAIL-PCR, whereas, using the FPNI-PCR method we obtained two specific fragments of 2068 bp and 683 bp [fig_ref] Figure 5: Comparison of PCR amplification products generated in typical genomic walking experiments using... [/fig_ref]. Sequence analysis of 102 amplified DNA fragments resulting from the FPNI-PCR technique showed that they corresponded with 100% accuracy to the target products.
## Efficiency of fpni-pcr
FPNI-PCR facilitates the rapid identification of target genomic sequences by permitting the use of short-cuts for a number of the steps between tissue isolation and target product sequencing, thereby reducing the input of time and effort compared to other current methods of flanking sequence cloning. For example, FPNI-PCR can be employed to generate target genomic products when cell lysates are used to supply the DNA templates [fig_ref] Figure 4: Amplified products after the tertiary round of PCR in genomic walking and... [/fig_ref]. For this procedure, we extracted cell lysates from the young leaves of transgenic tobacco, petunia hybrida and Rosa hybrida using the rapid NaOH extraction method. We found no difference in the success of target product amplification from these cell lysates compared to that from the higher quality DNA templates as purified by the CTAB method. In addition, we found no differences between the results of direct sequencing of the FPNI-PCR products compared to indirect sequencing of the cloned products within the PMD18-T vector. FPNI-PCR has the advantage that, after the first step (12-15 cycles), we can proceed to the second step directly, without requiring the prior dilution of the products from the first PCR step. When using such practices, we commonly achieved > 85% positive cloning of target products following the secondary PCR reaction. Thus, we were able to amplify target products using just 42-45 PCR cycles, and this was completed in less than 3 hours from the start of tissue extraction. When proceeding through the third PCR and final sequencing steps, FPNI-PCR was found to provide 100% positive results in our working tests of genomic cloning. In total, the procedure was completed in just 54-57 cycles, taking less than 4 hours.
# Discussion
## Fpni-pcr involves a novel strategy of amplification
The FPNI-PCR strategy is based on the following technical points. (1) The arbitrary degenerate oligonucleotides forming the 3' section of the FP primers (FPs) have been designed with the intention of maximizing the level of similarity with common genomic oligo sequences throughout the various taxa. (2) During the first PCR step, the 3'-ends of the oligos forming the arbitrary degenerate portions of the FPs (which also include several fixed bases at the 3'-end) need to pair effectively with selected sites of high similarity in the single-stranded DNA template. A lower annealing temperature may be used to increase mismatch sites and, thereby, enlarge the universality of the designed fusion primers. Between 1 and 6 cycles in the first PCR step are used to ensure that the desired target products are present within the amplified mix. (3) In the secondary and tertiary PCR steps, the target DNA should be exponentially amplified via high stringency cycles, while nontarget products are not amplified by the gene-specific and FP-specific primer pairs. Off-target amplification may also be suppressed by the stem-loop structure of nonspecific DNA products. The use of high stringency PCR cycles encourages the efficient and specific amplification of the target products, while off-target PCR products are progressively diluted. This aspect of the FPNI-PCR strategy represents a key difference from the TAIL-PCR method developed by Liu and Whittier [bib_ref] Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal..., Liu [/bib_ref]. TAIL-PCR employs multiple low stringency annealing cycles and, thus, nonspecific products are readily amplified. In TAIL-PCR, enrichment of the target products relies upon the difference in amplification velocity between the target and non-target products. Thus, the TAIL-PCR protocol involves two rounds of laborious DNA dilutions and numerous (approx. 110) cycles of PCR [bib_ref] Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal..., Liu [/bib_ref]. An improved method, named hiTAIL-PCR, similarly involves two rounds of laborious DNA dilutions and multiple low and high stringency PCR cycles, thereby still providing opportunities for nonspecific product amplification [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref]. In addition, hiTAIL-PCR is uneconomical in terms of primer design, with the secondary gene-specific primers requiring the addition of > 20 nucleotides [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref].
FPNI-PCR represents a very powerful tool in genomic walking and flanking sequence cloning TAIL-PCR has been widely used for genome walking and flanking sequence cloning in plants and other organisms. Commonly, however, specific products appear in the second round of PCR only to disappear during the third round of PCR [bib_ref] A Novel and Simple PCR Walking Method for Rapid Acquisition of Long..., Luo [/bib_ref]. Liu and Chen reported that just 60% of reactions yielded specific products [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref]. Our own tests of the TAIL-PCR method found that, on some occasions, no product was observed even after three rounds of PCR. Generally, smaller and fewer products were generated by TAIL-PCR as compared to the corresponding reactions using FPNI-PCR [fig_ref] Figure 5: Comparison of PCR amplification products generated in typical genomic walking experiments using... [/fig_ref]. For inverse and ligation-mediated PCR-based genomewalking methods, successful amplification of target products relies on the restriction fragmentation of genomic DNA. When employing such techniques, we also encountered some "off-target" situations, possibly due to the location of the available restriction sites relative to the locus-specific primer [bib_ref] A highthroughput genome-walking method and its use for cloning unknown flanking sequences, Reddy [/bib_ref]. By contrast, repeated tests of the FPNI-PCR method using the nine given FP primers revealed no "off-target" cases, and more than 80% of FPNI-PCR reactions yielded DNA fragments larger than 1.0 kb. Thus, our results indicate that FPNI-PCR is a more powerful technique for genomic walking and flanking sequence cloning studies. In addition, the FPNI-PCR method makes it feasible to amplify large sections of specific DNA fragments. This can be achieved through the control of annealing temperatures during the low stringency cycles of the first PCR step. The control of this parameter can also be used to avoid the cloning of "off target" flanking sequences.
## Fpni-pcr is a simple and rapid method of genomic walking and flanking sequence cloning
Methods involving inverse PCR and ligation-mediated PCR techniques suffer from the requirement for complicated manipulations, such as restriction cleavage, ligation, or tailing before PCR amplification [bib_ref] A highthroughput genome-walking method and its use for cloning unknown flanking sequences, Reddy [/bib_ref] [bib_ref] Site Finding-PCR: a simple and efficient PCR method for chromosome walking, Tan [/bib_ref]. Consequently, the use of such methods to successfully amplify and clone the target product usually requires a time-scale exceeding one working day. Similarly, the successful application of TAIL-PCR or hiTAIL-PCR methods is also time-consuming, with the time taken to perform the necessary number of PCR cycles (approx. 110) requiring a full working day (and this is even before the time taken to perform the various dilution steps is considered) [bib_ref] Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal..., Liu [/bib_ref] [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref]. Using FPNI-PCR, a total about 42-45 cycles (requiring less than two hours) can be adequate to produce a good rate of success in target product amplification and, in our hands, a total of 54-57 cycles (completed in less than 4 hours) ensured a successful result in 100% of cases. Both TAIL-PCR and FPNI-PCR methods demand only extremely modest levels of quantity and purity of the template DNA, thus, they can work well with both cell lysates and crude DNA extracts. Using FPNI-PCR, however, we were able to choose a wider range of annealing temperatures during the low stringency cycles of the first PCR step, thereby extending the scope of target product that could be successfully amplified. The currently widely used Genome Walker Kit from Clontech employs restriction enzyme cleavage of DNA and adaptor-ligated genomic DNA construction, and also involves a nested PCR strategy. The procedures of integrated restriction cleavage, ligation and tailing make the Genome Walker Kit a substantially more complicated and time-consuming process than the first PCR step of FPNI-PCR. Thus, we believe that in comparison to the other reported methods of genomic walking and flanking sequence cloning [ [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref] [bib_ref] Characterization of the gene encoding glycoprotein C of duck enteritis virus, Liu [/bib_ref] [bib_ref] Restriction Site-dependent PCR: An Efficient Technique for Fast Cloning of New Genes..., Yu [/bib_ref] [bib_ref] Direct and efficient cloning of full-length genes from environmental DNA by RT-qPCR..., Huang [/bib_ref] , and references as above], FPNI-PCR is technically less demanding and is more time-efficient.
## Fpni-pcr is highly effective and accurate
We conducted 60 rounds of FPNI-PCR reactions using the nine given FP primers, and in each case we successfully amplified products and experienced no "off-target" results. This contrasts to the experiences when employing ligation-dependent PCR methods, when it is often a problem to find available restriction sites and obtain ligation products [bib_ref] Evaluation of PCR-based methods for isolating flanking regions of genes, Satyanarayana [/bib_ref]. In addition, sequence analysis of 102 amplification products from FPNI-PCR indicated that the accuracy of positive cloning in our experiments was 100%. By comparison, 60-80% of reactions yielded specific products according to the original TAIL-PCR technical report [bib_ref] Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal..., Liu [/bib_ref] , and a success rate of 93% was recorded for the improved hiTAIL-PCR procedure [bib_ref] High-efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences, Liu [/bib_ref]. Thus, we conclude that FPNI-PCR shows a substantially higher level of effectiveness than the established TAIL-PCR procedures. We suggest that the high rate of positive amplification results, at least partially, from the high stringency PCR cycles in the second and third PCR stages of the FPNI-PCR protocol.
# Conclusion
To date, TAIL-PCR has been widely used for genome walking and flanking sequence cloning in molecular biology research, but it is still laborious and time-consuming.
Recently, several alternative methods have been introduced [bib_ref] Restriction Site Extension PCR: A Novel Method for High-Throughput Characterization of Tagged..., Ji [/bib_ref] [bib_ref] A highthroughput genome-walking method and its use for cloning unknown flanking sequences, Reddy [/bib_ref] [bib_ref] A Novel and Simple PCR Walking Method for Rapid Acquisition of Long..., Luo [/bib_ref] , including the Genome Walker Kit from Clontech, but these methods still involve the use of restriction enzyme-mediated digestion, Phi29 DNA polymerase-mediated amplification, and a series of template DNA purification steps. Here, we report a simple and effective PCR method namely, fusion primer and nested integrated PCR (FPNI-PCR). This method is based on the use of arbitrary degenerate nucleotides reflecting natural restriction sites or site-dependent nucleotides in the genomic DNA to design universal primers. These FP primers are used to conduct the first round of PCR, which is followed by two to three rounds of nested PCR. We undertook a large number of experiments using FPNI-PCR in order to robustly demonstrate that this novel method has the capability to be more powerful, effective and accurate than the established TAIL-PCR method. Furthermore, the FPNI-PCR method is completed in a shorter time-scale than other widely-used genome-walking techniques, and is significantly less technically complex to perform.
# Methods
## Genomic dna isolation
Genomic DNA of seven Rosaceace species, namely, Pyracantha fortuneana, Prunus mume, Photinia serrulata L., Prunus persica, Spiraea cantoniensis L., Rosachinensis and Fragaria × ananassa, was extracted from young leaves according to the procedure of Yang et al [bib_ref] Establishment of AFLP technique and assessment of primer combinations for Mei Flower, Yang [/bib_ref]. In addition, genomic DNA of tobacco, Petunia hybrida, Arabidopsis and rice was isolated from young leaves according to the NaOH fast extraction method, as reported previously.
## Oligonucleotide primers
All sequence specific primers were designed using Pri-mer5 software (see Additional files 1, , and annealing (Tm) temperatures ranged from 60°C to 72°C. For the design of the arbitrary degenerate primers (ADs), AD sequences were either collected from the existing literature or were designed by us (Additional files 1, [fig_ref] Table 1: FP primers [/fig_ref]. The AD oligos were fused to the 3' end of an adaptor oligo which supplied a known sequence of 31 fixed (i.e. non-degenerate) nucleotides, thereby forming the fusion primers (FPs). The simplest form of these FPs is termed type I primers [fig_ref] Table 1: FP primers [/fig_ref]. Type II primers are designed to include a hairpin structure within the 5' end of the single-stranded adaptor regions (Additional files 1, [fig_ref] Table 2: Cycling parameters and PCR conditions for FPNI-PCR [/fig_ref]. Type III primers are arbitrary degenerative primers that contain just 4-6 fixed nucleotides at the 3' end (Additional files 1, [fig_ref] Table 3: Submitted genomic DNA sequences [/fig_ref]. illustrates the schematic relationship between the sequence-specific and arbitrary degenerate primers with the target genomic sequence or flanking sequence of a T-DNA insertion. All primers used in this study were synthesized by Invitrogen Co. Limit, and are summarized in [fig_ref] Table 1: FP primers [/fig_ref] and Additional files 1, Tables S1-S5. The annealing (Tm) temperatures of the arbitrary degenerate primers (ADP) used in this study ranged between 36°C and 48°C. Tm values were estimated by using the formula: Tm = 81.5 + 16.6 × Log 10 [Na+] + 0.41 (%GC) -600/total number of nucleotides.
## Pcr conditions, procedures and gel analysis methods
All FPNI-PCR procedures were arranged in three stages. The first stage involved a PCR step which was performed in a total volume of 20 μL, and contained 1-2 μL of the DNA products (i.e. 2 μL DNA extraction solution from the NaOH fast extraction method), 0.2 μM gene-specific primers (SP1), 1.0 μM fusion arbitrary degenerate primers and 1 U rTaq (Takara). Between 9-18 PCR cycles were performed using annealing temperatures ranging between 28°C and 52°C (full details of the conditions used are given in [fig_ref] Table 2: Cycling parameters and PCR conditions for FPNI-PCR [/fig_ref]. In the second and third PCR stages, gene-specific primers (SP2 and SP3) and adaptor-specific primers (FSP1 and FSP2) were used in place of the SP1 and fusion arbitrary degenerate primers, respectively; the schematic arrangement of these primer pairs with respect to the target genomic DNA is illustrated in [fig_ref] Figure 2: A general [/fig_ref]. Due to the exact primer-template match for the primer pairs used in the second and third PCR steps, high stringency cycles were employed in these two latter stages (details shown in [fig_ref] Table 2: Cycling parameters and PCR conditions for FPNI-PCR [/fig_ref]. For the second PCR step, a 1.0 μL aliquot of the first PCR product pool was used as the template. Following this second round of selective amplification, the secondary PCR products were diluted 50-fold with MQ water and 1 μL of the dilution was added as template to 20 μL volumes of the tertiary PCR mixtures. For visualisation of the products amplified following the second or third FPNI-PCR steps, a 5-8 μL aliquot was separated by electrophoresis in a 1.5% agarose TAE gel, then stained with ethidium bromide and observed under a UV illumination system.
## Dna sequencing
For direct sequencing, the remaining 12-15 μL of PCR products were purified through a Sephadex G-50 column and subjected to sequencing procedures using SP3 primers designed to known sections of the target sequence, or FSP2 primers designed according to the original FP primers. For indirect sequencing, bands corresponding to the largest products were excised from the agarose gels, purified and cloned into a pMD18-T TA cloning vector (TaKaRa), and then sequenced with M13 forward or reverse primers.
# Additional material
Additional file 1: [fig_ref] Table 1: FP primers [/fig_ref]. AD primer (arbitrary degenerate primers) and universal primers used in this study. [fig_ref] Table 2: Cycling parameters and PCR conditions for FPNI-PCR [/fig_ref]. FP primers (exhibiting restriction site) and the corresponding universal primers used in FPNI-PCR. [fig_ref] Table 3: Submitted genomic DNA sequences [/fig_ref]. FP primers (exhibiting hair pin structure) and the corresponding universal primers used in FPNI-PCR. . Gene specific primers used to generate the electrophoresis patterns presented in this paper. : Gene specific primers used to generate the electrophoresis patterns presented when conducting genomic walking in Arabidopsis and rice.
Additional file 2: Supporting figures employed in main text. . Illustration of the effect of annealing temperature during the low stringency PCR cycles in the primary PCR step in FT ortholog cloning of Fragaria ananassa using FPNI-PCR (amplified products after the secondary round of PCR). M: molecular marker; number in the lanes: 1-9 FP primer; number in the bottom right corner: annealing temperature. -: control. [fig_ref] Figure 2: A general [/fig_ref]. Amplified products after the tertiary round of PCR in genomic walking and T-DNA flanking sequence cloning using FPNI-PCR for the 3 genes of wuschel family from Arabidopsis (a), and Osft, Osmads1 and Ostuba1 genes from rice (b) (FP primer indicated by number in photos).
[fig] Figure 2: A general (theoretical) scheme for FPNI-PCR (PCR based method for genomic walking or tagged flanking sequence cloning). [/fig]
[fig] Figure 3: The effect of annealing temperature during the low stringency PCR cycles of the primary PCR step. Illustration of the effect of annealing temperature during the low stringency PCR cycles of the primary PCR step in FT ortholog cloning of Fragaria ananassa using FPNI-PCR (amplified products shown after the tertiary round of PCR). M: molecular marker; number in the lanes: the 1-9 FP primer; number in the bottom right corner: annealing temperature. -: control. [/fig]
[fig] Figure 4: Amplified products after the tertiary round of PCR in genomic walking and T-DNA flanking sequence cloning using FPNI-PCR. (a) Cloning of the FT-Like gene of Rosa rugosa using 3 gene-specific and 1-8 FP primers. (b) T-DNA flanking sequence cloning in three transgenic tobacco individual plants using forward primers corresponding to the known T-DNA border; various fragment sizes were obtained from the use of different FP primers (1-9) in the three tested tobacco plants. (c) T-DNA flanking sequence cloning using backward primers. Only target specific fragments were amplified by using an appropriate annealing temperature during the low stringency cycles of the first PCR step (similar T-DNA sequences were present in each of the tested transgenic tobacco plants). Note: M: molecular marker; number: 1-9 FP primers; -: control. [/fig]
[fig] Figure 5: Comparison of PCR amplification products generated in typical genomic walking experiments using TAIL-PCR and FPNI-PCR methods, respectively. (A) Products generated by 3 gene-specific primers and 1-9 arbitrary degenerate primers in FT ortholog cloning of Pyracantha fortuneana using TAIL-PCR; a 510 bp fragment was obtained using the fourth AD primer, only. (B) Products generated by the same 3 gene-specific and 1-9 FP primers in FT ortholog cloning of Pyracantha fortuneana using FPNI-PCR; four specific fragments were amplified by FPNI-PCR and the longest fragment was ca. 1.7 kp. (C) Products generated by 3 gene-specific and 1-9 arbitrary degenerate (AD) primers in SOC1 ortholog cloning of Pyracantha fortuneana using TAIL-PCR. (D) Products generated by 3 gene-specific and 1-9 FP primers in SOC1 ortholog cloning of Pyracantha fortuneana using FPNI-PCR. Note: M: molecular marker; number: 1-9 arbitrary degenerate primers (FPs); -: control. [/fig]
[table] Table 1: FP primers (no hair pin structure) and universal primers used in FPNI-PCR. [/table]
[table] Table 2: Cycling parameters and PCR conditions for FPNI-PCR. [/table]
[table] Table 3: Submitted genomic DNA sequences (genes) in varied species and T-DNA insertion flanking sequences in transgenic tobacco using FPNI-PCR. Rosa chinensis RcFT gene for flowering locus T protein, complete cds. Fragaria vesca FvFT gene for flowering locus T protein, complete cds. [/table]
|
Disruption of Beta-Cell Mitochondrial Networks by the Orphan Nuclear Receptor Nor1/Nr4a3
Nor1, the third member of the Nr4a subfamily of nuclear receptor, is garnering increased interest in view of its role in the regulation of glucose homeostasis. Our previous study highlighted a proapoptotic role of Nor1 in pancreatic beta cells and showed that Nor1 expression was increased in islets isolated from type 2 diabetic individuals, suggesting that Nor1 could mediate the deterioration of islet function in type 2 diabetes. However, the mechanism remains incompletely understood. We herein investigated the subcellular localization of Nor1 in INS832/13 cells and dispersed human beta cells. We also examined the consequences of Nor1 overexpression on mitochondrial function and morphology. Our results show that, surprisingly, Nor1 is mostly cytoplasmic in beta cells and undergoes mitochondrial translocation upon activation by proinflammatory cytokines. Mitochondrial localization of Nor1 reduced glucose oxidation, lowered ATP production rates, and inhibited glucose-stimulated insulin secretion. Western blot and microscopy images revealed that Nor1 could provoke mitochondrial fragmentation via mitophagy. Our study unveils a new mode of action for Nor1, which affects beta-cell viability and function by disrupting mitochondrial networks.
# Introduction
Type 2 diabetes is characterized by a progressive deterioration of beta-cell mass and function, leading to a relative deficit in insulin secretion. Significant reduction in beta-cell mass has been reported in individuals with type 2 diabetes, an effect that has been attributed at least in part to increased beta-cell apoptosis. However, our understanding of the molecular mechanisms contributing to beta-cell demise in the etiology of diabetes remains incomplete.
Our recent study, investigating the biological functions of Nr4a nuclear receptors in pancreatic beta cells, characterized a key role for the orphan nuclear receptor Nor1 in the regulation of functional beta-cell mass. Indeed, we reported that Nor1-knockout (KO) animals displayed increased beta-cell mass and improved glucose tolerance. In our study, genetic manipulations of Nor1 in beta-cell lines and isolated human islets demonstrated a proapoptotic role of Nor1 in beta cells. Consistently, Nor1 expression was upregulated by cytokines and siRNA-mediated knockdown of Nor1 prevented cytokine-induced beta-cell death. Moreover, Nor1 expression was robustly elevated in islets from type 2 diabetic donors, suggesting that Nor1 could participate in the deterioration of functional beta-cell mass that causes the disease. However, the precise mechanism by which Nor1 provokes beta-cell death remains unexplored. Interestingly, single-nucleotide polymorphisms (SNPs) located within the INS832/13 cells (passages 57-80, kind gift from Dr. Marc Prentki (Montreal, QC, Canada)) were grown in RPMI 1640 medium containing 11 mM glucose and supplemented with 10 mM HEPES, 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 1 mM sodium pyruvate, and 50 µM β-mercaptoethanol at 37 - C in a humidified 5% CO 2 atmosphere (as originally described in. Our cytokine treatment consisted of a combination of IL-1 (10 ng/mL) and IFNγ (10 ng/mL).
## Human islets
Human islets were purchased from the Alberta Diabetes Institute Islet Core at the University of Alberta with the assistance of the Human Organ Procurement and Exchange Program and the Trillium Gift of Life Network, who provide donor pancreata for research. All experiments were approved by the Human Research Ethics Board at the University of Alberta (protocol no. 00036707). For mitophagy quantification, islets from four different healthy (nondiabetic) donors were used. Donor characteristics are shown inbelow. Islets were cultured in 11 mM glucose RPMI medium supplemented with 10 mM HEPES, 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 1 mM sodium pyruvate, 1% penicillin-streptomycin, and 50 µM β-mercaptoethanol at 37 - C in a humidified 5% CO 2 atmosphere.
## Plasmid transfection
pCMV-Nor1-GFP, pCMV-Nor1-Flag, peGFP-LC3, and pcDNA3.1 were obtained from Origene (Rockville, MD, USA). For INS832/13 cells, plasmids were introduced at a concentration of 5 µg of Cells 2020, 3 of 15 DNA for 6 × 10 6 cells via nucleofection (Lonza, Allendale, NJ, USA). Transfection efficiency was >90% with this method. We obtain a 3-4-fold increase in Nor1 protein levels after transfection with Nor1-GFP or Nor1-Flag. For human islets, cells were dispersed using 0.05% trypsin-EDTA and transfected the following day using Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer's protocol. We obtained~50% transfection efficiency using Lipofectamine. Cells were assayed 30 h post-transfection.
## Mitochondria staining
Transfected cells were seeded onto 18 mm poly-L-lysine-coated (Millipore Sigma) coverslips (2 × 10 6 cells per coverslip). Thirty hours post-transfection, they were incubated in the presence or absence of cytokines for 1 h. Media was then changed to RPMI 1640 containing 25 nM Mitotracker ® Red CMXRos (Molecular Probes, Eugene, OR, USA) and 10 µg/mL Hoechst 33342 (BD Bioscience, San Jose, CA, USA) with or without 20 µM FCCP (Cayman Chemical, Ann Arbor, MI, USA) for 15 min. The cells were then washed once with PBS. Live cell images were acquired on a WaveFX spinning disk confocal microscope with a Hamamatsu C9100-50 EMCCD camera and analyzed using the Volocity imaging software (version 6.1.2, PerkinElmer, Waltham, MA, USA).
## Glucose oxidation
Glucose oxidation was measured as described beforein nonventilated 25 cm 2 cell culture flasks. Briefly, cells were preincubated in 2 mM glucose KRB medium for 30 min before being treated with 2, 7, or 11 mM glucose. To each flask, 5 µCi of [14C(U)]-D-glucose (Moravek Biochemicals, Brea, CA, USA) was added. All metabolic reactions were stopped after 45 min by adding H 2 SO 4 to the media. 14C-CO 2 was captured using fiberglass fiber filters and counted using a scintillation counter.
## Glucose uptake
Glucose uptake was assessed using a commercial kit (Glucose Uptake-Glo™Assay, Promega, Madison, WI, USA). Cells were treated as described above in Section 2.5 for glucose oxidation assays, except that 5% of the total glucose of each condition was 2-deoxyglucose.
## Atp production
Cells were washed with cold PBS and lysed in a buffer containing 10 mM Tris pH 7.5, 100 mM NaCl, 1 mM EDTA, and 0.01% Triton X100. ATP was quantified using the ATP Determination Kit (Molecular Probes).
## Mitochondrial membrane potential
Cells were treated in the presence or the absence of cytokines for 4 h before the assay was performed. FCCP (20 µM) was used as a negative control and 25 mM glucose as a positive control. Mitochondrial membrane potential was then evaluated using the JC-1 Mitochondrial Membrane Potential Assay Kit (Cayman chemicals, Ann Arbor, MI, USA).
## Mitochondrial ultrastructure
Transfected cells were prefixed in 4% glutaraldehyde in 0.2 M sodium cacodylate buffer, postfixed in 1% osmium tetroxide (OsO4) in 0.1 M sodium cacodylate buffer, dehydrated in an ethyl alcohol series, embedded with epoxy resin, and thermally polymerized as previously described. Ultrathin sections (70 nm) were cut by an ultramicrotome (Leica Microsystems, Richmond Hill, ON, Canada) and then stained with 4% uranyl acetate and Reynold's lead citrate. The contrasted sections were imaged using a Hitachi H-7650 transmission electron microscope at 80 kV equipped with a 16-megapixel EMCCD camera (XR111, Advanced Microscopy Technique, Woburn, MA, USA).
## Quantification of mitophagy
Fluorescence microscopy: Cells were transfected with peGFP-LC3 and with either Nor1-Flag or a control empty vector 30 h after transfection; cells were imaged as described in Section 2.4. The colocalization of peGFP-LC3 positive vacuoles and mitochondria was determined using the Imaris software (version 8.4, Bitplane, Concord, MA, USA). Electron microscopy: The number of autophagic vacuoles targeting the mitochondria was determined as described before. Briefly, we examined three grid squares per condition. In each square, images were taken at 400× to assess whole cell areas and at 12,000× to count the number of autophagic vacuoles.
## Western blot
Cells were lysed in a buffer containing 50 mM Tris (pH 8.0), 150 mM NaCl, and 1% v/v Triton X-100 and supplemented with protease inhibitors (Complete Mini, Roche Diagnostics, Mannheim, Germany). Cytosolic and mitochondrial protein extracts were obtained using a cell fractionation kit (Abcam, Toronto, ON, Canada). Protein concentrations were determined by BCA protein assay. Equal amounts of heat-denatured proteins from each treatment group were run on Novex 10% Tris-glycine gels (Thermo Fisher Scientific) and electrically transferred to nitrocellulose membranes. After blocking for 1 h at room temperature with 1% BSA, membranes were incubated overnight at 4 - C with primary antibodies. The next day, membranes were incubated with horseradish-peroxidase-linked secondary antibodies followed by exposure to Amersham ECL Western Blotting Detection Reagents (GE Healthcare, Mississauga, ON, Canada) and film development. The primary antibodies used in our studies were rabbit anti-VDAC antibody (Cell Signaling, Danvers, MA, USA), mouse monoclonal anti-tubulin (Santa Cruz Biotechnology, Dallas, TX, USA), and rabbit anti-LC3 (Novus Biologicals, Oakville, ON, Canada).
## Citrate synthase activity
Cells were harvested in a solution containing methylsulfonylmethane (MSM)/EDTA supplemented with 1% sodium cholate hydrate. Citrate synthase activity was then evaluated by measuring the conversion of 5,5 -dithiobis-(2-nitrobenzoic acid) (DTNB, 0.1 mM) into 2-nitro-5-benzoic acid (TNB), which absorbs specifically at 412 nm. The reaction was carried out in a buffer containing 0.25% Triton X100, 0.5 mM oxaloacetate, and 0.31 mM acetyl-CoA. Results were normalized to total protein content of the cells.
## High-resolution respirometry
Cellular aerobic respiration was measured using high-resolution respirometry (Oxygraph-2k, Oroboros Instruments, Innsbruck, Austria), as we have performed before. In brief, the oxygraph was calibrated at 37 - C per the manufacturer's instructions with each chamber filled with 2 mL of mitochondrial respiration medium 05 (MIR05) containing 110 mM sucrose, 60 mM K-lactobionate, 0.5 mM EGTA, 0.1% BSA, 3 mM MgCl 2 , 20 mM taurine, 10 mM KH 2 PO 4 , and 20 mM HEPES, which was magnetically stirred at 500 rpm. DatLab 4 software (Oroboros Instruments, Innsbruck, Austria) was used for data acquisition and analysis. Equal numbers of transfected cells (1 million cells per condition) were rinsed twice with MIR05 and transferred in each oxygraph chamber. After measurement of routine respiration in MIR05 and permeabilization of the cell membranes with digitonin, the following substrates and inhibitors were added (final concentration in the chamber): glutamate (10 mM), malate (5 mM), and pyruvate (5 mM) as Complex I (CI)-linked substrates; succinate (10 mM) as Complex II (CII)-linked substrates; rotenone (0.5 µM) and antimycin A (2.5 µM) as CI and CIII inhibitors; ascorbate (0.5 mM) and tetramethylphenylenediamine (TMPD, 2 mM) as CIV-linked substrates. Mitochondrial respiration was corrected for oxygen flux due to instrumental background and for residual oxygen consumption after inhibition of Complexes I and III with rotenone and antimycin A, respectively.
# Statistical analysis
Data are presented as mean ± SEM. Statistical analyses were performed with GraphPad Prism ® (GraphPad Software, San Diego, CA, USA) using ANOVA followed by Bonferroni's post hoc test. p-values < 0.05 were considered statistically significant.
# Results
## Nor1 translocates to the mitochondria in ins832/13 cells exposed to proinflammatory cytokines
Our previous study characterized the nuclear receptor Nor1 as a novel mediator of cytokine-induced beta-cell apoptosis mass. To gain insights into the mechanism, we first sought to investigate the subcellular localization of Nor1 in beta cells. Because of the dearth of reliable Nor1 antibodies, we overexpressed a Nor1-GFP protein in INS832/13 cells cultured in the absence or presence of a cocktail of cytokines comprising IL-1 (10 ng/mL) and IFNγ (10 ng/mL). Mitochondria and nuclei were stained with Mitotracker Red and DAPI, respectively. In untreated (control) cells, Nor1-GFP showed a diffuse cytosolic signal, which did not colocalize with the nucleus or the mitochondria . However, upon cytokine treatment, Nor1-GFP was found to rapidly translocate and accumulate at the mitochondria, as detected by the overlap of the green and red signals. Surprisingly, we could not detect Nor1-GFP in the nucleus in any of the time points tested (up to 24 h). Nuclear exclusion of Nor1 was unexpected since it raised questions about its presumed role as a transcription factor in beta cells. Be that as it may, the translocation of Nor1 at the mitochondria in response to cytokines prompted us to study its effects on mitochondrial function.
# Statistical analysis
Data are presented as mean ± SEM. Statistical analyses were performed with GraphPad Prism ® (GraphPad Software, San Diego, CA, USA) using ANOVA followed by Bonferroni's post hoc test. p-values < 0.05 were considered statistically significant.
# Results
## Nor1 translocates to the mitochondria in ins832/13 cells exposed to proinflammatory cytokines
Our previous study characterized the nuclear receptor Nor1 as a novel mediator of cytokine-induced beta-cell apoptosis mass. To gain insights into the mechanism, we first sought to investigate the subcellular localization of Nor1 in beta cells. Because of the dearth of reliable Nor1 antibodies, we overexpressed a Nor1-GFP protein in INS832/13 cells cultured in the absence or presence of a cocktail of cytokines comprising IL-1 (10 ng/mL) and IFNγ (10 ng/mL). Mitochondria and nuclei were stained with Mitotracker Red and DAPI, respectively. In untreated (control) cells, Nor1-GFP showed a diffuse cytosolic signal, which did not colocalize with the nucleus or the mitochondria . However, upon cytokine treatment, Nor1-GFP was found to rapidly translocate and accumulate at the mitochondria, as detected by the overlap of the green and red signals. Surprisingly, we could not detect Nor1-GFP in the nucleus in any of the time points tested (up to 24 h). Nuclear exclusion of Nor1 was unexpected since it raised questions about its presumed role as a transcription factor in beta cells. Be that as it may, the translocation of Nor1 at the mitochondria in response to cytokines prompted us to study its effects on mitochondrial function.
## Nor1 affects mitochondrial function and reduces insulin secretion in ins832/13 cells
In beta cells, mitochondrial function plays a critical role in the regulation of insulin secretion. In particular, glucose-stimulated oxidative ATP production causes a rise in the cytosolic ATP/ADP ratio, which triggers a series of electrophysiological events that provoke insulin exocytosis. We thus investigated the potential effect of Nor1 on glucose metabolism, ATP production, mitochondrial membrane potential, and insulin secretion. Nor1 significantly blunted glucose oxidation in cells exposed to intermediate or high (11 mM) glucose concentrations. This effect was not due to a reduction in glucose uptake, which remained unaffected by Nor1 overexpression. Consistently, with its action on glucose oxidation, Nor1 decreased glucose-stimulated ATP production by~30%, an effect that was not additive to the effect of cytokines. The decrease in ATP production was concomitant with an increase in lactic acid levels, suggesting that pyruvate conversion to lactic acid replaces pyruvate oxidation in the mitochondria.
shows the quantification of Nor1-GFP localization at the mitochondria as determined by the overlap of the green and red signals. Scale bar = 20 µm. Results are mean ± SEM of three independent experiments, * p < 0.05 versus control vector.
## Nor1 affects mitochondrial function and reduces insulin secretion in ins832/13 cells
In beta cells, mitochondrial function plays a critical role in the regulation of insulin secretion. In particular, glucose-stimulated oxidative ATP production causes a rise in the cytosolic ATP/ADP ratio, which triggers a series of electrophysiological events that provoke insulin exocytosis. We thus investigated the potential effect of Nor1 on glucose metabolism, ATP production, mitochondrial membrane potential, and insulin secretion. Nor1 significantly blunted glucose oxidation in cells exposed to intermediate or high (11 mM) glucose concentrations. This effect was not due to a reduction in glucose uptake, which remained unaffected by Nor1 overexpression. Consistently, with its action on glucose oxidation, Nor1 decreased glucose-stimulated ATP production by ~30%, an effect that was not additive to the effect of cytokines. The decrease in ATP production was concomitant with an increase in lactic acid levels, suggesting that pyruvate conversion to lactic acid replaces pyruvate oxidation in the mitochondria. Surprisingly, Nor1 significantly increased mitochondrial membrane potential by~30%. This is reminiscent of a modelin which hyperpolarization of the inner mitochondrial membrane in proapoptotic conditions is proposed to provoke mitochondrial matrix swelling and ruptures in the outer membrane. Nor1-induced changes in glucose oxidation and ATP production translated into inhibition of glucose-stimulated insulin secretion. Nor1 did not affect insulin secretion in response to KCl-induced depolarization, indicating that the inhibitory effects of Nor1 on insulin secretion are due to impairment of glucose oxidation and/or mitochondrial function and not through an action on the exocytotic machinery.
We next examined mitochondria morphology by TEM in INS832/13 cells transfected with Nor1 or its control vector. Nor1-expressing cells displayed mitochondrial matrix swelling and subsequent herniation through breaks in the outer membrane. These features were absent in control cells. It is known that outer mitochondrial membrane ruptures allow for the release of cytochrome c from the intermembrane space into the cytosol. Yet, we have recently shown that Nor1 does significantly increase cytochrome c release to cause apoptosis in beta cells. Surprisingly, Nor1 significantly increased mitochondrial membrane potential by ~30%. This is reminiscent of a modelin which hyperpolarization of the inner mitochondrial membrane in proapoptotic conditions is proposed to provoke mitochondrial matrix swelling and ruptures in the outer membrane. Nor1-induced changes in glucose oxidation and ATP production translated into inhibition of glucose-stimulated insulin secretion. Nor1 did not affect insulin secretion in response to KCl-induced depolarization, indicating that the inhibitory effects of Nor1 on insulin secretion are due to impairment of glucose oxidation and/or mitochondrial function and not through an action on the exocytotic machinery.
We next examined mitochondria morphology by TEM in INS832/13 cells transfected with Nor1 or its control vector. Nor1-expressing cells displayed mitochondrial matrix swelling and subsequent herniation through breaks in the outer membrane. These features were absent in control cells. It is known that outer mitochondrial membrane ruptures allow for the release of cytochrome c from the intermembrane space into the cytosol. Yet, we have recently shown that Nor1 does significantly increase cytochrome c release to cause apoptosis in beta cells.
## Nor1 induces mitochondrial fractionation in beta cells
We then performed morphological studies to confirm a potential disrupting action of Nor1 on the mitochondrial networks. Our fluorescence microscopy images showed a strong disruption of the networks in Nor1-expressing cells. Indeed, Nor1-transfected INS832/13 cellsand human islet cells
## Nor1 induces mitochondrial fractionation in beta cells
We then performed morphological studies to confirm a potential disrupting action of Nor1 on the mitochondrial networks. Our fluorescence microscopy images showed a strong disruption of the networks in Nor1-expressing cells. Indeed, Nor1-transfected INS832/13 cellsand human islet cellsdisplayed decreased mitochondrial networked areas and increased punctate staining. Moreover, mitochondrial fractionation was also observed in TEM images. Measurements of mitochondrial lengths in fluorescence microscopy images revealed that Nor1 induces a shift in the distribution of mitochondria towards smaller sizes. Indeed, there was an accumulation of smaller mitochondria (<499 nm in length) in cells transfected with Nor1 compared with control. On the contrary, large mitochondria of >1500 nm were virtually absent in Nor1-expressing cells. We next evaluated if the mitochondria were subjected to mitophagy. Cells 2020, 9, x 8 of 14 4C). Measurements of mitochondrial lengths in fluorescence microscopy images revealed that Nor1 induces a shift in the distribution of mitochondria towards smaller sizes. Indeed, there was an accumulation of smaller mitochondria (<499 nm in length) in cells transfected with Nor1 compared with control. On the contrary, large mitochondria of >1500 nm were virtually absent in Nor1-expressing cells. We next evaluated if the mitochondria were subjected to mitophagy.
## Nor1 increases mitophagy in beta cells
We first assessed mitophagy by Western blot. We investigated LC3 conversion in whole-cell extracts and mitochondrial fractions of INS832/13 cells transfected with Nor1 or a control vector. FCCP was used as a positive control. Whereas Nor1 did not significantly affect LC3 levels in whole-cell extracts, it induced an accumulation of LC3-II in mitochondrial fractions, providing evidence that Nor1 could induce mitophagy in beta cells. The effect of Nor1 was similar to that of FCCP. VDAC and tubulin were used to assess the quality of cell fractionation. We detected equal amounts of VDAC in all mitochondrial samples, while tubulin was absent. To further document Nor1 action on mitophagy, we quantified the colocalization of mitochondria (in red) with a LC3-GFP construct (green) in INS832/13 cells transfected with Nor1 or a control vector , respectively, or exposed to FCCP . We detected a significant 2-fold increase in the colocalization of the green and red signals in Nor1-expressing INS832/13 cells . Similar experiments in dispersed human islet cells revealed a 2-3-fold increase in mitochondrial LC3-GFP localization . We also quantified the number of autophagosomes targeting the mitochondria in our TEM images and detected a significant increase (~4-fold) in cells transfected with Nor1 .
## Nor1 increases mitophagy in beta cells
We first assessed mitophagy by Western blot. We investigated LC3 conversion in whole-cell extracts and mitochondrial fractions of INS832/13 cells transfected with Nor1 or a control vector. FCCP was used as a positive control. Whereas Nor1 did not significantly affect LC3 levels in whole-cell extracts, it induced an accumulation of LC3-II in mitochondrial fractions, providing evidence that Nor1 could induce mitophagy in beta cells. The effect of Nor1 was similar to that of FCCP. VDAC and tubulin were used to assess the quality of cell fractionation. We detected equal amounts of VDAC in all mitochondrial samples, while tubulin was absent. To further document Nor1 action on mitophagy, we quantified the colocalization of mitochondria (in red) with a LC3-GFP construct (green) in INS832/13 cells transfected with Nor1 or a control vector , respectively, or exposed to FCCP . We detected a significant 2-fold increase in the colocalization of the green and red signals in Nor1-expressing INS832/13 cells . Similar experiments in dispersed human islet cells revealed a 2-3-fold increase in mitochondrial LC3-GFP localization . We also quantified the number of autophagosomes targeting the mitochondria in our TEM images and detected a significant increase (~4-fold) in cells transfected with Nor1 .
## Nor1 and mitochondrial content in beta cells
Next, we sought to confirm that the increase in mitophagy induced by Nor1 translated into decreased mitochondrial content. Unfortunately, we obtained conflicting results depending on the methods used. For instance, the mitochondrial area relative to total cell area was not significantly altered in cells overexpressing Nor1 in our TEM images . However, it was significantly decreased in our immunofluorescence images . Additionally, Nor1 did not change citrate synthase activity or VDAC protein levels compared to controls . We initially expected that Nor1 would decrease mitochondrial content due to its effect on mitophagy, so these results were unexpected. It is possible that Nor1 increases the mitochondrial turnover rate in beta cells to keep the mitochondrial content unchanged despite its effect on mitophagy. Another plausible scenario is that mitophagy precedes the changes in mitochondrial content and that we were too early to detect significant changes in content. Actin, tubulin, and VDAC were used to demonstrate equal loading and assess the purity of mitochondrial fractions. The graph shows densitometry analysis of the mitochondrial LC3-I/LC3-II ratios. Results are mean ± SEM of three independent experiments, * p < 0.05 versus control vector. To test whether Nor1 affects energy metabolism by acting on specific respiratory complexes, we performed high-resolution respiratory analyses in permeabilized cells transduced by Nor1 or a control vector in the presence of inhibitors or substrates feeding electrons into specific mitochondrial respiratory complexes . Routine respiratory rates were identical between control cells and Nor1-overexpressing cells. Also, we could not detect any significant differences between the respiratory capacity of individual complexes in cells with or without Nor1 gain-of-function. Since Nor1 was found to reduce glucose oxidation and ATP production in beta cells, these results suggest that Nor1 could act predominantly by diverting pyruvate to lactate production, as shown in. respiratory complexes . Routine respiratory rates were identical between control cells and Nor1-overexpressing cells. Also, we could not detect any significant differences between the respiratory capacity of individual complexes in cells with or without Nor1 gain-of-function. Since Nor1 was found to reduce glucose oxidation and ATP production in beta cells, these results suggest that Nor1 could act predominantly by diverting pyruvate to lactate production, as shown in.
# Discussion
Nr4as are garnering considerable interest for their potential implication in metabolic diseases, in particular diabetes. However, their biological actions in pancreatic beta cells remain incompletely understood. Our recent work characterized Nor1 as a mediator of cytokine-induced beta-cell death. Indeed, Nor1 gain-of-function provoked beta-cell apoptosis whereas Nor1 loss-of-function prevented cytokine-induced apoptosis. Moreover, Nor1-KO animals displayed higher beta-cell mass and improved glucose tolerance. Finally, we also showed that Nor1 expression was increased in type 2 diabetic islets, suggesting that Nor1 could play a role in the etiology of diabetes.
We herein documented the subcellular localization of Nor1 in untreated and cytokine-treated cells. Surprisingly, we discovered that in beta cells, activated Nor1 undergoes mitochondrial translocation. In our conditions, we could not detect nuclear Nor1, suggesting that the "nuclear receptor" may not act primarily as a transcription factor in beta cells. This would confer a distinct biological role for Nor1 as compared with the other members of the Nr4a family. Indeed, Nur77 and Nurr1 have been shown previously to translocate to the nucleus and regulate transcription in response to various stressors in beta cells.
A similar honing of Nor1 to the mitochondria has been described in immune cells. In this model, Nor1 was found to interact with Bcl2 to expose its BH3 domain and trigger the intrinsic pathway of apoptosis. However, the biological outcome appears to be different in beta cells because the canonical pathway of intrinsic apoptosis involves the dissipation of the mitochondrial membrane potential.
Our data indicate that Nor1 rather increases mitochondrial membrane potential. An alternative model exists and proposes that proapoptotic stimuli could induce mitochondrial hyperpolarization, resulting in an osmotic imbalance capable of provoking outer mitochondrial membrane rupture and the release of intermembrane space proteins such as cytochrome c to induce cell death. Our results dovetail with this alternative model and suggest that Nor1 causes beta-cell death via both impairment of beta-cell function and disruption of the mitochondrial network. Of note, mitochondrial membrane potential was measured using the JC-1 ratiometric dye, the fluorescence of which may change in response to factors other than mitochondrial membrane potential. Thus, results obtained with JC-1 must be interpreted with caution.
Our results suggest that Nor1 could affect beta-cell energy metabolism and cause changes to mitochondrial function. Altogether, these changes translate into blunted glucose oxidation, reduced ATP production, and, consequently, inhibition of insulin secretion. Interestingly, Nor1 thus reiterates the effects of cytokines, which were recently shown to suppress pyruvate oxidation and inhibit insulin secretion in beta cells. In our model, Nor1-induced impairment of mitochondrial function and energy production was concomitant with the induction of morphological changes at the mitochondria, including increased mitochondrial fragmentation and the induction of mitophagy. These observations are relevant, since mitochondrial dynamics govern the maintenance of optimal mitochondrial functionand it has been previously demonstrated that disruption of the mitochondrial network leads to reduced glucose oxidation and ATP production. Moreover, mitochondrial fragmentation also actively participates in the triggering of apoptosis. Thus, our results linking Nor1 to mitochondrial fragmentation provide a molecular mechanism by which Nor1 could both stimulate beta-cell apoptosis and impair beta-cell function, two defects that participate in the development of diabetes.
Mitochondrial dynamics also dictate the removal of these organelles by autophagy. Small, rounded, fragmented mitochondria are indeed targeted by autophagosomes. We thus quantified mitophagy and detected an accumulation of double-membrane vacuoles targeting the mitochondria in Nor1-expressing cells. Interestingly, the number of autophagosomes has been shown to be increased in the beta cells of diabetic individuals compared with nondiabetic subjects . To our knowledge, the role of Nr4a members in autophagy remains unexplored. A single publication reported that Nur77 could induce autophagic cell death following its translocation to the mitochondria .
# Conclusions
In conclusion, our study demonstrates for the first time that the nuclear receptor Nor1 translocates to the mitochondria in beta cells to disrupt mitochondrial networks, a process known to regulate both beta-cell function and viability. These results are important considering that we previously detected an upregulation of Nor1 in type 2 diabetic islets. We thus believe that Nor1 constitutes a potential molecular target to improve or maintain functional beta-cell mass.
|
Multisystem inflammatory syndrome in children and Kawasaki disease: a critical comparison
[bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease: a scientific statement for..., Mccrindle Brian [/bib_ref] [bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease: a scientific statement for..., Mccrindle Brian [/bib_ref]
## Aetiology
The aetiology of Kawasaki disease is uncertain, and there is no single specific diagnostic test. The general consensus, based on results from multiple studies, is that Kawasaki disease is an immune-mediated disease triggered by infection (or infections) in patients with a genetic predisposition [bib_ref] Genetic variations in the receptor-ligand pair CCR5 and CCL3L1 are important determinants..., Burns [/bib_ref] [bib_ref] A genome-wide association study identifies novel and functionally related susceptibility loci for..., Burgner [/bib_ref] [bib_ref] Kawasaki disease: what is the epidemiology telling us about the etiology?, Burgner [/bib_ref] [bib_ref] Kawasaki disease, Son [/bib_ref] [bib_ref] The riddle of Kawasaki disease, Burns [/bib_ref]. Some epidemiological features offer clues to the pathogenesis of Kawasaki disease. It is typically noted in children between the ages of 6 months and 5 years, with an estimated incidence of 25 cases per 100,000 children younger than 5 years in North America [bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease: a scientific statement for..., Mccrindle Brian [/bib_ref] [bib_ref] Kawasaki syndrome hospitalizations in the United States, Holman [/bib_ref]. It is believed that children younger than 6 months, who have immature immune systems, are protected by passive immunity provided by the transplacental transfer of maternal antibodies, whereas children older than 5 years have developed protective antibody responses to the ubiquitous antigens that most encounter uneventfully in early childhood [bib_ref] Kawasaki disease: a maturational defect in immune responsiveness, Kuijpers [/bib_ref]. There is a male predominance (~1.5:1) in the incidence of Kawasaki disease, a feature that is shared by many common childhood infectious diseases [bib_ref] The male predominance in the incidence of infectious diseases in children: a..., Green [/bib_ref] [bib_ref] Genetic susceptibility to infectious diseases, Burgner [/bib_ref].
Seasonal variation in the incidence of Kawasaki disease has been noted, with peak incidence occurring in winter and spring in the USA and UK, and in summer in China and Korea [bib_ref] Rising incidence of Kawasaki disease in England: analysis of hospital admission data, Harnden [/bib_ref] [bib_ref] Kawasaki syndrome in the United States 1976 to 1980, Bell [/bib_ref] [bib_ref] Hospitalizations for Kawasaki disease among children in the United States, Chang [/bib_ref] [bib_ref] Epidemiologic picture of Kawasaki disease in Beijing from 1995 through 1999, Du [/bib_ref] [bib_ref] Epidemiologic study of Kawasaki disease in Korea, 1997-1999: comparison with previous studies..., Park [/bib_ref]. Seasonal variation is least evident in Japan, the country with the highest incidence of Kawasaki disease [bib_ref] Epidemiology of Kawasaki disease in Asia, Europe, and the United States, Uehara [/bib_ref] [bib_ref] Results of 12 nationwide epidemiological incidence surveys of Kawasaki disease in Japan, Yanagawa [/bib_ref]. Geographical variation and clustering in the incidence of Kawasaki disease also occurs, with the highest incidence reported in Japan, China, South Korea and Taiwan [bib_ref] Epidemiology of Kawasaki disease in Asia, Europe, and the United States, Uehara [/bib_ref] [bib_ref] Seasonality and temporal clustering of Kawasaki syndrome, Burns [/bib_ref] [bib_ref] Temporal and geographical clustering of Kawasaki disease in Japan, Nakamura [/bib_ref] [bib_ref] Nationwide epidemic of Kawasaki disease in Japan during winter of 1985-86, Yanagawa [/bib_ref]. These epidemiological features point towards a transmissible infectious agent, which tends to occur in certain regions of the world with a seasonal variation in its incidence. Evidence exists for the presence of concurrent infections (with bacteria or common respiratory viruses, including coronaviruses) in patients with Kawasaki disease [bib_ref] Viral infections associated with Kawasaki disease, Chang [/bib_ref] [bib_ref] Infections and Kawasaki disease: implications for coronary artery outcome, Benseler [/bib_ref] [bib_ref] Association between a novel human coronavirus and Kawasaki disease, Esper [/bib_ref]. Immunohistochemistry analyses have shown infiltration of IgA plasma cells indicative of the antigen-driven immune response in inflamed tissues and the presence of cytoplasmic antigens suggestive of an infectious aetiology in bronchial and vascular endothelial cells and macrophages [bib_ref] Detection of antigen in bronchial epithelium and macrophages in acute Kawasaki disease..., Rowley [/bib_ref]. However, to date no single organism has been directly proved to cause Kawasaki disease [bib_ref] Kawasaki disease at 50 years, Cohen [/bib_ref] [bib_ref] Is Kawasaki disease an infectious disorder?, Rowley [/bib_ref].
## Involvement of superantigens
The potential pathogenic role of superantigens has been evaluated, on the basis of observations of preferential expression of T cell receptor (TCR) β genes encoding variable regions Vβ2 and Vβ8.1 in the peripheral blood lymphocytes of patients with acute Kawasaki disease [bib_ref] Selective expansion of T cells expressing T-cell receptor variable regions Vβ2 and..., Abe [/bib_ref] [bib_ref] Characterization of T cell repertoire changes in acute Kawasaki disease, Abe [/bib_ref] [bib_ref] Evidence for a superantigen mediated process in Kawasaki disease, Curtis [/bib_ref]. Superantigen activity has been identified in the gut microbiota of such patients, and culture supernatants of these bacteria contain a heat shock protein (Hsp60, also known as GroEL) that induces T cell division and production of pro-inflammatory cytokines [bib_ref] Heat shock proteins and superantigenic properties of bacteria from the gastrointestinal tract..., Nagata [/bib_ref]. However, in studies using flow cytometry in large series of patients with Kawasaki disease, TCR skewing and over-presentation of the described TCR clones has not been found, and it is currently believed that Kawasaki disease is a result of T cell activation by a conventional antigen [bib_ref] TCR Vβ family repertoire and T cell activation markers in Kawasaki disease, Pietra [/bib_ref] [bib_ref] Characterization of the T-cell receptor V-β repertoire in Kawasaki disease, Mancia [/bib_ref].
## Involvement of nutritional disorders
The role of nutritional disorders, including vitamin D deficiency, in the pathogenesis of Kawasaki disease is subject to debate [bib_ref] Relationship between vitamin D levels and intravenous immunoglobulin resistance in Kawasaki disease, Jun [/bib_ref]. Vitamin D has an anti-inflammatory effect mediated through elevation of expression of IL-10 and inhibition of expression of vascular endothelial growth factor [bib_ref] 1-α,25-Dihydroxyvitamin D3 (1,25(OH)(2)D(3)) hampers the maturation of fully active immature dendritic cells..., Canning [/bib_ref] [bib_ref] 25-dihydroxyvitamin D3 (Calcitriol) inhibits hypoxia-inducible factor-1/vascular endothelial growth factor pathway in human..., Ben-Shoshan [/bib_ref]. Results from a German population-based study showed that vitamin D supplementation has a protective effect against the development of Kawasaki disease [bib_ref] Breastfeeding and vitamin D supplementation reduce the risk of Kawasaki disease in..., Meyer [/bib_ref]. Low serum concentrations of vitamin D might contribute to the development of coronary artery complications in children with [bib_ref] Severe vitamin D deficiency in patients with Kawasaki disease: a potential role..., Stagi [/bib_ref]. However, other results have
## Key points
- multisystem inflammatory syndrome in children (mIS-c) is characterized by exaggerated innate and adaptive immune responses following infection with severe acute respiratory syndrome coronavirus 2 (SArS-cov-2) in predisposed children. - clinical presentation of mIS-c involves multiple organ systems, with prominent involvement of the gastrointestinal and cardiovascular systems. - The factors that trigger the development of mIS-c in children exposed to or infected with SArS-cov-2 are not yet known. - results from epidemiological, clinical and immunological investigations have revealed that although mIS-c has phenotypic similarities to Kawasaki disease, they are different syndromes. - The approach to treatment of mIS-c aims to mute the augmented inflammatory response.
0123456789();:
identified elevation of vitamin D levels during the acute phase of Kawasaki disease in children who subsequently developed coronary arterial lesions [bib_ref] Prediction of the risk of coronary arterial lesions in Kawasaki disease by..., Chen [/bib_ref]. The contribution of other nutritional factors has also been suggested. For example, iron-deficiency anaemia is associated with development of coronary abnormalities in Kawasaki disease [bib_ref] Iron deficiency anemia as a predictor of coronary artery abnormalities in Kawasaki..., Kim [/bib_ref]. These varied results suggest the need for further investigation and research to elucidate the role of malnutrition in the pathogenesis of Kawasaki disease.
## The role of microbiota
Disturbances in the normal microbiota (dysbiosis) have been proposed to have a role in the pathogenesis of various autoimmune and inflammatory disorders, including Kawasaki disease [bib_ref] Role of the microbiota in immunity and inflammation, Belkaid [/bib_ref] [bib_ref] Our evolving understanding of Kawasaki disease pathogenesis: role of the gut microbiota, Kaneko [/bib_ref] [bib_ref] Microbiologic studies on the small intestine in Kawasaki disease, Yamashiro [/bib_ref]. Stools from children with Kawasaki disease contain higher numbers of Gram-positive bacteria from the Streptococcus, Staphylococcus, Eubacterium and Peptostreptococcus genera, as well as Hsp60-producing Gram-negative bacteria, and lower numbers of lactobacilli than stools from children with other febrile illnesses or healthy controls [bib_ref] Heat shock proteins and superantigenic properties of bacteria from the gastrointestinal tract..., Nagata [/bib_ref] [bib_ref] Microbiologic studies on the small intestine in Kawasaki disease, Yamashiro [/bib_ref] [bib_ref] Characteristic profile of intestinal microflora in Kawasaki disease, Takeshita [/bib_ref].
Dysbiosis is associated with reduction in the production of short-chain fatty acids (particularly butyrate) and is proposed to lead to aberrant immune responses that are associated with Kawasaki disease 63 .
## Genetic susceptibility
Epidemiological and genetic studies of Kawasaki disease have shed light on the role of genetic susceptibility in its development [bib_ref] Immunogenetics of Kawasaki disease, Kumrah [/bib_ref]. Kawasaki disease is prevalent in Japan, but also in children of Japanese ancestry living in Hawaii [bib_ref] Kawasaki syndrome in Hawaii, Holman [/bib_ref]. Siblings of children with Kawasaki disease have a 10-fold higher risk of development of the condition than children in the general population [bib_ref] Kawasaki disease in families, Fujita [/bib_ref]. Several candidate genes have been identified through genome-wide association studies and linkage studies. The four major groups of genes that have been studied in Kawasaki disease are those associated with T cell activation (ORAI1 and STIM1), B cell signalling (CD40, BLK and FCGR2A), apoptosis (CASP3) and transforming growth factor-β (TGFβ) signalling (TGFB2, TGFBR2, MMP and SMAD) [bib_ref] Immunogenetics of Kawasaki disease, Kumrah [/bib_ref]. CASP3 encodes caspase 3, which is an effector caspase with a vital role in the execution phase of apoptosis. A single-nucleotide polymorphism in the CASP3 gene is associated with susceptibility to Kawasaki disease 67 . TGFβ is another vital protein with a central role in immunoregulation that affects multiple populations of leukocytes. Abnormalities in TGFβ signalling resulting from genetic variation are involved in Kawasaki disease susceptibility and outcomes [bib_ref] Transforming growth factor-beta signaling pathway in patients with Kawasaki disease, Shimizu [/bib_ref].
Understanding the roles of these genetic alterations has implications for potential therapeutic approaches [bib_ref] Perspective of immunopathogenesis and immunotherapies for Kawasaki disease, Chang [/bib_ref]. In addition to these groups, mutations in ITPKC, which is involved in Ca 2+ mobilization and activation of NLRP3 inflammasomes, could result in enhancement of IL-1β and IL-18 production, disease susceptibility, coronary abnormalities and resistance to treatment with intravenous immunoglobulin (IVIG) [bib_ref] Inositol-triphosphate 3-kinase c mediates inflammasome activation and treatment response in Kawasaki disease, Alphonse [/bib_ref] [bib_ref] Molecular genetics of Kawasaki disease, Onouchi [/bib_ref]. Notably, immunosuppressive agents such as ciclosporin, a T cell inhibitor that blocks the calcineurin-NFAT pathway, have shown promise in the treatment of high-risk IVIG-resistant Kawasaki disease 72 . The observed association of HLA polymorphisms with Kawasaki disease varies; the predominant variant in a Japanese cohort was HLA-Bw54, whereas HLA-Bw51 was predominantly identified in white and Jewish populations [bib_ref] HLA antigens in mucocutaneous lymph node syndrome, Matsuda [/bib_ref] [bib_ref] HLA antigens in Kawasaki disease, Kato [/bib_ref] [bib_ref] HLA antigens in mucocutaneous lymph node syndrome in New England, Krensky [/bib_ref]. Epigenetic regulation of inflammatory and immunoregulatory genes by factors such as methylation, microRNAs and long noncoding RNAs has been identified in Kawasaki disease, and might be relevant to pathogenesis and prognosis. Additionally, single-nucleotide polymorphisms in cytokine genes, including IL1, KCNN2, TIFAB, P2RY12 and TNF, are associated with Kawasaki disease and with risk of coronary artery lesions, as well as IVIG treatment failure [bib_ref] Pro-inflammatory cytokine single nucleotide polymorphisms in Kawasaki disease, Assari [/bib_ref] [bib_ref] IL-1B polymorphism in association with initial intravenous immunoglobulin treatment failure in Taiwanese..., Weng [/bib_ref] [bib_ref] Identification of KCNN2 as a susceptibility locus for coronary artery aneurysms in..., Kim [/bib_ref] [bib_ref] Identification of the TIFAB gene as a susceptibility locus for coronary artery..., Kwon [/bib_ref] [bib_ref] P2RY12:rs7637803 TT variant genotype increases coronary artery aneurysm risk in Kawasaki disease..., Lu [/bib_ref] [bib_ref] The IL-1B gene polymorphisms rs16944 and rs1143627 contribute to an increased risk..., Fu [/bib_ref].
## Immunological aberrations
Innate and adaptive immune responses both have important roles in the development of Kawasaki disease 83 . An intense initial response driven by the innate immune system takes the form of neutrophilic leukocytosis, activation of monocytes, natural killer (NK) cells and γδ T cells, and elevation of production of acute-phase NATure revIeWS | RheuMATOlOGy reactants and cytokines, especially IL-1β, which contributes to activation of endothelial cells, inducing upregulation of expression of cell-adhesion molecules, IL6 and IL8 . Pro-inflammatory IL-17 produced by type 17 T helper (T H 17) cells could activate immune cells such as neutrophils and monocytes, leading to production of other inflammatory cytokines, such as IL-6, TNF and IL-8, thereby contributing to the pathogenesis of many inflammatory disorders [bib_ref] Increase in T helper type 17 cells in children with Kawasaki disease..., Zhang [/bib_ref] [bib_ref] Interleukin-6 is prone to be a candidate biomarker for predicting incomplete and..., Wu [/bib_ref] [bib_ref] The role of TNF-α in a murine model of Kawasaki disease arteritis..., Oharaseki [/bib_ref] [bib_ref] Serial changes of serum interleukin-6, interleukin-8, and tumor necrosis factor alpha among..., Lin [/bib_ref]. By contrast, CD4 + CD25 + regulatory T (T reg ) cells contribute to immune tolerance through suppression of the hyperactivation of both innate and adaptive immune cells, via several mutually nonexclusive mechanisms. An imbalance in these pathways could lead to immune dysregulation, which could have a role in the pathogenesis of Kawasaki disease [bib_ref] Th17-and Treg-related cytokine and mRNA expression are associated with acute and resolving..., Guo [/bib_ref] [bib_ref] Cytokines predict coronary aneurysm formation in Kawasaki disease patients, Lin [/bib_ref] [bib_ref] Interleukin 10 and and transforming growth factor β polymorphisms as risk factors..., Rahmani [/bib_ref] [bib_ref] The T helper type 17/regulatory T cell imbalance in patients with acute..., Jia [/bib_ref]. Neutrophils are activated in Kawasaki disease and release reactive oxygen species, leading to endothelial cell injury [bib_ref] Kawasaki disease: a matter of innate immunity, Hara [/bib_ref]. Release of neutrophil extracellular traps (NETs) is also implicated in the pathogenesis of Kawasaki disease 97 . Although NETs have a protective role against infections as components of the innate immune system, they also have pathogenic potential for immune dysregulation and promotion of inflammation and tissue injury [bib_ref] Neutrophil extracellular traps in immunity and disease, Papayannopoulos [/bib_ref] Autoimmune antibodies are thought to have a role in the pathogenesis of Kawasaki disease, particularly those against endothelial cell antigens [bib_ref] The role of anti-endothelial cell antibodies in Kawasaki disease -in vitro and..., Grunebaum [/bib_ref] [bib_ref] On the basis of systems-level analyses of various immune cells, cytokines and..., Consiglio [/bib_ref]. Anti-endothelial cell antibodies could cause endothelial damage, with release of pro-inflammatory cytokines and a hypercoagulable state leading to vessel-wall injury and intravascular thrombosis [bib_ref] Autoimmune aspects of Kawasaki disease, Sakurai [/bib_ref]. However, not all results have demonstrated elevation of anti-endothelial cell antibodies in patients with Kawasaki disease [bib_ref] Anti-neutrophil cytoplasmic antibodies and anti-endothelial cell antibodies are not increased in Kawasaki..., Nash [/bib_ref].
Immune complexes might have a role in the development of Kawasaki disease 106 . They appear in the first 7 days of the disease and peak in the second week before declining 107 . Elevation of circulating levels of immune complexes in Kawasaki disease is related to adverse outcomes such as coronary artery abnormalities [bib_ref] Circulation immune complexes and granulocytes chemotaxis in Kawasaki disease: the 8th conference..., Miyata [/bib_ref] [bib_ref] Impaired granulocyte chemotaxis and increased circulating immune complexes in Kawasaki disease, Ono [/bib_ref]. However, a causal relationship between immune complexes and the pathogenesis of Kawasaki disease has not been definitively established. Activation of the complement system has also been implicated in the pathogenesis of Kawasaki disease, via both the classical and the mannose-binding lectin pathways [bib_ref] Classical pathway complement activation in Kawasaki syndrome, Kohsaka [/bib_ref] [bib_ref] Polymorphisms in the mannose-binding lectin gene as determinants of age-defined risk of..., Biezeveld [/bib_ref].
Kawasaki disease is considered by some to be a form of IgA vasculitis. In children with Kawasaki disease, intestinal permeability and levels of secretory IgA in the circulation are greater than in unaffected children, and in mouse models of the disease, elevation of levels of circulating secretory IgA and IgA deposition in the vasculature are observed [bib_ref] High levels of IgA-containing circulating immune complex and secretory IgA in Kawasaki..., Ohshio [/bib_ref] [bib_ref] Kawasaki disease: pathophysiology and insights from mouse models, Rivas [/bib_ref]. Furthermore, pharmacological blockade of zonulin (a modulator of intestinal tight junctions) and administration of IVIG in these mouse models reduce intestinal permeability and cardiovascular inflammation compared with levels in untreated controls [bib_ref] Intestinal permeability and IgA provoke immune vasculitis linked to cardiovascular inflammation, Rivas [/bib_ref].
Therapeutic strategies IVIG and acetylsalicylic acid have emerged as first-line therapies for the management of Kawasaki disease 20 . IVIG therapy leads to rapid improvement in the clinical symptoms of rash, fever and conjunctival injection in most patients. Although the exact mechanism of action of IVIG is not yet known, proposals include inhibition of activation of innate immune cells and inflammatory mediators, expansion of T reg cells and suppression of T H 17 cells [bib_ref] Th17-and Treg-related cytokine and mRNA expression are associated with acute and resolving..., Guo [/bib_ref] [bib_ref] Defective FOXP3 expression in patients with acute Kawasaki disease and restoration by..., Olivito [/bib_ref] [bib_ref] Gene expression profiling of the effect of high-dose intravenous Ig in patients..., Abe [/bib_ref] [bib_ref] Increased nitric oxide production by neutrophils in early stage of Kawasaki disease, Yoshimura [/bib_ref] [bib_ref] Evaluation of intravenous immunoglobulin resistance and coronary artery lesions in relation to..., Wang [/bib_ref]. Evidence indicates that IVIG might target IL-1β + neutrophils via caspase-independent pathways [bib_ref] Immune response to intravenous immunoglobulin in patients with Kawasaki disease and MIS-C, Zhu [/bib_ref]. Single-cell RNA sequencing of peripheral blood mononuclear cells in acute Kawasaki disease before and after IVIG therapy has revealed that genes encoding inflammatory mediators (including TNF and IL1B) are highly expressed in monocytes in untreated disease, with reduction of expression following therapy, along with significant enhancement of the plasma-cell population and induction of oligoclonal expansion of T cell receptors and IgG and IgA B cell receptors [bib_ref] Single-cell RNA sequencing of peripheral blood mononuclear cells from acute Kawasaki disease..., Wang [/bib_ref]. Mining of transcriptomic data by Boolean analysis has identified that several metabolic pathways might contribute to IVIG resistance in Kawasaki disease [bib_ref] Boolean analysis of the transcriptomic data to identify novel biomarkers of IVIG..., Rambabu [/bib_ref]. In high-risk patients with acute Kawasaki disease and in those who do not respond to IVIG therapy, steroid treatment can be considered, to prevent the occurrence of coronary artery abnormalities 20 . Additional therapeutic options for IVIG-resistant Kawasaki disease include infliximab (a monoclonal antibody to TNF), ciclosporin (a calcineurin inhibitor) and anakinra (an IL-1 receptor antagonist) [bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease: a scientific statement for..., Mccrindle Brian [/bib_ref].
Overview of MIS-C Since April 2020, many reports have documented a new hyperinflammatory syndrome in children [bib_ref] An outbreak of severe Kawasaki-like disease at the Italian epicentre of the..., Verdoni [/bib_ref] [bib_ref] Hyperinflammatory shock in children during COVID-19 pandemic, Riphagen [/bib_ref] [bib_ref] Acute heart failure in multisystem inflammatory syndrome in children (MIS-C) in the..., Belhadjer [/bib_ref]. In May 2020, the US Centers for Disease Control and Prevention (CDC) issued an alert identifying MIS-C as a critical illness in children that was associated with SARS-CoV-2 infection 123 . Since then, more than 4,000 cases of MIS-C have been reported in the USA alone. In 26 studies published in 2020 and 2021, documenting 1,136 cases of MIS-C (mostly occurring in the USA and Europe), the reported median ages of the affected children were 6-11 years, with no significant gender difference [bib_ref] Multisystem inflammatory syndrome in children in New York State, Dufort [/bib_ref] [bib_ref] Multisystem inflammatory syndrome related to COVID-19 in previously healthy children and adolescents..., Cheung [/bib_ref] [bib_ref] Acute heart failure in multisystem inflammatory syndrome in children (MIS-C) in the..., Belhadjer [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children associated with severe acute respiratory syndrome coronavirus..., Kaushik [/bib_ref] [bib_ref] Intensive care admissions of children with paediatric inflammatory multisystem syndrome temporally associated..., Davies [/bib_ref] [bib_ref] Paediatric multisystem inflammatory syndrome temporally associated with SARS-CoV-2 mimicking Kawasaki disease (Kawa-COVID-19):..., Pouletty [/bib_ref] [bib_ref] Kawasaki-like multisystem inflammatory syndrome in children during the covid-19 pandemic in Paris,..., Toubiana [/bib_ref] [bib_ref] Characteristics, cardiac involvement, and outcomes of multisystem inflammatory syndrome of childhood associated..., Capone [/bib_ref] [bib_ref] Spectrum of imaging findings on chest radiographs, US, CT, and MRI images..., Hameed [/bib_ref] [bib_ref] Clinical characteristics of 58 children with a pediatric inflammatory multisystem syndrome temporally..., Whittaker [/bib_ref] [bib_ref] Multi-inflammatory syndrome in children related to SARS-CoV-2 in Spain, Moraleda [/bib_ref] [bib_ref] Epidemiological and clinical profile of pediatric inflammatory multisystem syndrome -temporally associated with..., Dhanalakshmi [/bib_ref] [bib_ref] Gastrointestinal symptoms as a major presentation component of a novel multisystem inflammatory..., Miller [/bib_ref] [bib_ref] SARS-CoV-2-related paediatric inflammatory multisystem syndrome, an epidemiological study, Belot [/bib_ref] [bib_ref] Distinct clinical and immunological features of SARS-CoV-2-induced multisystem inflammatory syndrome in children, Lee [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children related to COVID-19: a New York City..., Riollano-Cruz [/bib_ref] [bib_ref] Paediatric inflammatory multisystem syndrome: temporally associated with SARS-CoV-2 (PIMS-TS): cardiac features, management..., Ramcharan [/bib_ref] [bib_ref] Acute myocarditis and multisystem inflammatory emerging disease following SARS-CoV-2 infection in critically..., Grimaud [/bib_ref] [bib_ref] SARS-CoV-2-specific IgG1/IgG3 but not IgM in children with pediatric inflammatory multi-system syndrome, Perez-Toledo [/bib_ref] [fig_ref] Table 2 |: Demographic features of MIS-C study populations [/fig_ref].
Patients with MIS-C have symptoms that resemble those of other hyperinflammatory syndromes, such as Kawasaki disease, toxic-shock syndrome (TSS) and macrophage activation syndrome (MAS), which is a type of secondary haemophagocytic lymphohistiocytosis [bib_ref] An outbreak of severe Kawasaki-like disease at the Italian epicentre of the..., Verdoni [/bib_ref] [bib_ref] Hyperinflammatory shock in children during COVID-19 pandemic, Riphagen [/bib_ref]. To improve clarity and aid diagnosis, the CDC published a case definition, which includes age <21 years, fever, laboratory evidence of inflammation, hospital admission, multisystem (two or more) organ involvement (cardiac, renal, respiratory, haematological, gastrointestinal, dermatological or neurological), either laboratory confirmation of SARS-CoV-2 infection (by PCR with reverse transcription (RT-PCR), serology or antigen test) or known COVID-19 exposure up to 4 weeks www.nature.com/nrrheum before symptom onset, with no alternative plausible diagnosis. The WHO and the UK Royal College of Paediatrics and Child Health (RCPCH) have also published case definitions, which are largely similar to the CDC definition, except that the RCPCH does not require evidence of prior exposure to SARS-CoV-2 . Despite the broad case definitions, and the considerable overlap with primary COVID-19 and other common childhood febrile illnesses, patients with MIS-C have distinct clinical presentation and levels of biomarkers, which aids in differential diagnosis [bib_ref] Discriminating multisystem inflammatory syndrome in children requiring treatment from common febrile conditions..., Carlin [/bib_ref] [bib_ref] Severe manifestations of SARS-CoV-2 in children and adolescents: from COVID-19 pneumonia to..., García-Salido [/bib_ref].
## Aetiology
Compared with adults, primary SARS-CoV-2 infection is relatively mild in children [bib_ref] COVID-19 in 7780 pediatric patients: a systematic review, Hoang [/bib_ref]. Evidence indicates that a temporal relationship exists between SARS-CoV-2 Although 80-90% of patients with MIS-C have been found to be SARS-CoV-2 seropositive, positivity in PCR testing is only 20-40%, suggesting that the interval to the onset of MIS-C is sufficient for viral RNA levels to fall considerably [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref] [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref]. Furthermore, nasopharyngeal aspirates from patients with MIS-C have higher SARS-CoV-2 real-time RT-PCR cycle thresholds (indicating lower levels of viral RNA) than those from patients with severe COVID-19 (ref. [bib_ref] Multisystem inflammatory syndrome in children and COVID-19 are distinct presentations of SARS-CoV-2, Diorio [/bib_ref]. However, autopsy examinations for three individuals who had MIS-C identified SARS-CoV-2 in various tissues, including heart, kidneys, brain and intestine, which is consistent with multisystem organ involvement in MIS-C [bib_ref] An autopsy study of the spectrum of severe COVID-19 in children: from..., Duarte-Neto [/bib_ref].
Notably, the prolonged presence of SARS-CoV-2 in children's intestines might cause zonulin-dependent loss of tight junctions, leading to leakage of viral antigens into the circulation, and to hyperinflammation and MIS-C 152 . By contrast, single-cell RNA sequencing of peripheral blood mononuclear cells from patients with acute MIS-C have revealed low viral and bacterial signatures in the immune cells, suggesting that active viral or bacterial infectious triggers are not contributing factors [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref]. The accumulated evidence suggests that MIS-C might be the result of a combination of post-infectious immune dysregulation and virus-induced cytopathic effects and inflammation in multiple organ systems.
Paediatric patients with COVID-19 or MIS-C have strong IgG, but weak IgM antibody responses to the trimeric S glycoprotein of SARS-CoV-2, and weak responses to the nucleocapsid protein N, which is implicated in viral replication [bib_ref] SARS-CoV-2-specific IgG1/IgG3 but not IgM in children with pediatric inflammatory multi-system syndrome, Perez-Toledo [/bib_ref] [bib_ref] Mapping systemic inflammation and antibody responses in multisystem inflammatory syndrome in children..., Gruber [/bib_ref] [bib_ref] Quantitative SARS-CoV-2 serology in children with multisystem inflammatory syndrome (MIS-C), Rostad [/bib_ref] [bib_ref] Distinct antibody responses to SARS-CoV-2 in children and adults across the COVID-19..., Weisberg [/bib_ref] [bib_ref] SARS-CoV-2 antibody responses in children with MIS-C and mild and severe COVID-19, Anderson [/bib_ref] [bib_ref] This article reports that children with MIS-C maintain FcγR-binding, inflammatory monocyte/ macrophage-activating..., Bartsch [/bib_ref]. By contrast, adult COVID-19 patients have higher levels of anti-S antibodies, broader immunoglobulin response to SARS-CoV-2 with respect to specificity and isotype distribution (including IgG, IgM and IgA isotypes) and higher virus-neutralizing capacity [bib_ref] SARS-CoV-2-specific IgG1/IgG3 but not IgM in children with pediatric inflammatory multi-system syndrome, Perez-Toledo [/bib_ref] [bib_ref] Quantitative SARS-CoV-2 serology in children with multisystem inflammatory syndrome (MIS-C), Rostad [/bib_ref] [bib_ref] Distinct antibody responses to SARS-CoV-2 in children and adults across the COVID-19..., Weisberg [/bib_ref] [bib_ref] This article reports that children with MIS-C maintain FcγR-binding, inflammatory monocyte/ macrophage-activating..., Bartsch [/bib_ref]. The mild or asymptomatic nature of COVID-19 in children might be related to the extent of the antibody response. Nevertheless, IgG antibodies to S protein provide an important diagnostic criterion for MIS-C. Low IgM titres in MIS-C are consistent with its appearance several weeks after SARS-CoV-2 exposure.
Analyses from geographically diverse cohorts have demonstrated that 20-50% of people with no previous exposure to the virus have T cell reactivity against peptides corresponding to SARS-CoV-2 sequences 159 , which might be related to CD4 + T cell cross-reactivity with circulating seasonal human 'common cold' coronavirus (HCoV) [bib_ref] Targets of T cell responses to SARS-CoV-2 coronavirus in humans with COVID-19..., Grifoni [/bib_ref]. Although this phenomenon has implications for the development of herd-immunity models and vaccine candidates, it is currently unclear whether the presence of prior cross-reactive CD4 + T cells is protective or harmful in the pathogenesis of MIS-C. When tested for serological evidence of prior seasonal coronavirus infection, children with MIS-C and those hospitalized for non-COVID reasons had similar prevalence and levels of antibodies to HCoV 161 . Additionally, HCoV antibody levels did not correlate with the levels of SARS-CoV-2 antibodies, suggesting that prior HCoV infection neither provides protection nor worsens the course of paediatric SARS-CoV-2 infection or MIS-C.
## Sars-cov-2 s protein as a superantigen
SARS-CoV-2 viral S protein might behave like a superantigen, triggering a cytokine storm that results in the development of the TSS-like presentation of MIS-C 162 [fig_ref] TLRFigure 2 |: Possible mechanisms implicated in aberrant activation of immune cells in MIS-C [/fig_ref]. The S protein has a high-affinity motif for binding TCR, which is similar in structure to the staphylococcal enterotoxin B, a superantigen that mediates TSS by interacting with both TCR and MHC class II molecules. Computational modelling has shown that SARS-CoV-2 encodes a superantigen motif near the S1/S2 cleavage site, which interacts with both the TCR and CD28 (ref. [bib_ref] Superantigenic character of an insert unique to SARS-CoV-2 spike supported by skewed..., Cheng [/bib_ref]. TCR repertoire analysis of T cells in a small number of patients with MIS-C has identified skewing of TCR Vβ towards TRBV11-2 (Vβ21.3), which is associated with HLA class I alleles A02, B35 and C04 . The CDR3-independent nature of TCR Vβ skewing suggested superantigen-mediated activation of T cells in MIS-C. Further evidence supports the enrichment of TRBV11-2 among T cells [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref] [bib_ref] Polyclonal expansion of TCR Vβ 21.3 + CD4 + and CD8 +..., Moreews [/bib_ref] , although notably it has been observed in the absence of differential expression of a set of 'superantigen genes' [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref]. Also, MIS-C is usually observed several weeks after primary SARS-CoV-2 exposure, in contrast to the acute illness and cytokine storm observed in TSS [bib_ref] Toxic shock syndrome and bacterial superantigens: an update, Mccormick [/bib_ref]. In most cases, SARS-CoV-2 is undetectable in patients with MIS-C during the acute phase of inflammation. Thus, the superantigenic property of SARS-CoV-2 S protein and its implication in MIS-C is not yet confirmed. As an RNA virus, SARS-CoV-2 undergoes constant mutation, and whether any particular variant of the virus contributes to MIS-C by triggering strong inflammatory signalling in the immune cells and endothelial cells of children with COVID-19 requires further exploration. Notably, the use of in silico techniques has demonstrated that mutations in the binding region of SARS-CoV-2 S protein could influence the interaction with MHC class II molecules and TCR [bib_ref] Superantigenic character of an insert unique to SARS-CoV-2 spike supported by skewed..., Cheng [/bib_ref].
## Involvement of nutritional disorders
Nutritional factors such as vitamin D deficiency might have a role in the development of MIS-C. Adults with vitamin D deficiency were noted to have a more severe form of COVID-19 with an increased risk of death than those without this deficiency [bib_ref] Vitamin D deficiency and outcome of COVID-19 patients, Radujkovic [/bib_ref]. Vitamin D supplementation has proved to be of some benefit in infections with other viruses, such as influenza A [bib_ref] Randomized trial of vitamin D supplementation to prevent seasonal influenza A in..., Urashima [/bib_ref]. Whether a similar benefit could accrue in MIS-C has not been elucidated.
## The role of microbiota
Another contributing factor in the development of MIS-C that warrants investigation is the role of gut and respiratory tract microbiota. Gastrointestinal microbes are important regulators of the gut immune system and inflammation, and influence the balance between T H 17 cells and T reg cells [bib_ref] The T reg /T H 17 axis: a dynamic balance regulated by..., Omenetti [/bib_ref]. Adult patients with COVID-19 display alteration of gut and upper respiratory microbiomes, and gut dysbiosis persists beyond the nasal clearance of SARS-CoV-2 (refs [bib_ref] Gut microbiota composition reflects disease severity and dysfunctional immune responses in patients..., Yeoh [/bib_ref] [bib_ref] Alterations in fecal fungal microbiome of patients with COVID-19 during time of..., Zuo [/bib_ref] [bib_ref] Signatures of COVID-19 severity and immune response in the respiratory tract microbiome, Merenstein [/bib_ref] [bib_ref] Alterations in gut microbiota of patients with COVID-19 during time of hospitalization, Zuo [/bib_ref]. Notably, faecal SARS-CoV-2 load is inversely correlated with the abundance of bacteria of the Bacteroidetes phylum, which suppress ACE2 in the mouse gut [bib_ref] Alterations in gut microbiota of patients with COVID-19 during time of hospitalization, Zuo [/bib_ref] from peer reviewed and non-peer reviewed reports also suggest the persistence of microbiome dysbiosis in the upper respiratory tract and the gut in paediatric COVID-19 .
## Genetic susceptibility
The low incidence of MIS-C relative to COVID-19, and the similarity in antibody response to SARS-CoV-2 in paediatric patients with MIS-C and with COVID-19 (irrespective of the subsequent development of MIS-C) suggest that SARS-CoV-2 infection causes dysregulation of immune responses in a subgroup of predisposed children with particular genetic backgrounds [bib_ref] Mechanisms underlying genetic susceptibility to multisystem inflammatory syndrome in children (MIS-C), Chou [/bib_ref]. Specifically, predisposition might be related to mutations and polymorphisms in the genes that encode pattern recognition molecules such as Toll-like receptors, components of the signalling cascades of the immune response and
## Immunological aberrations
In general, the signatures of immune cells and inflammatory parameters of MIS-C closely overlap with those of adults with moderate-to-severe COVID-19 rather than with paediatric COVID-19, which is mostly mild or asymptomatic. Also, immune activation in MIS-C is transient and tends to reduce during recovery [bib_ref] Mapping systemic inflammation and antibody responses in multisystem inflammatory syndrome in children..., Gruber [/bib_ref] [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref] [bib_ref] Peripheral immunophenotypes in children with multisystem inflammatory syndrome associated with SARS-CoV-2 infection, Carter [/bib_ref].
## Pro-inflammatory mediators.
Elevation of levels of pro-inflammatory cytokines such as IL-6, IL-10 and IL-17A, and chemokines such as CXCL5, CXCL11, CXCL1 and CXCL6 in MIS-C distinguishes it from paediatric COVID-19 . In various cohorts, elevation of TNF, IL-1β, IFNγ, soluble IL-2R, CCL2, CCL3, CCL4, CXCL8 (IL-8) or IFNγ-induced chemokines CXCL9 and CXCL10 has been reported in the serum of patients with MIS-C relative to those with paediatric COVID-19 or healthy controls [bib_ref] Distinct clinical and immunological features of SARS-CoV-2-induced multisystem inflammatory syndrome in children, Lee [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children and COVID-19 are distinct presentations of SARS-CoV-2, Diorio [/bib_ref] [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref] [bib_ref] Polyclonal expansion of TCR Vβ 21.3 + CD4 + and CD8 +..., Moreews [/bib_ref] [bib_ref] Peripheral immunophenotypes in children with multisystem inflammatory syndrome associated with SARS-CoV-2 infection, Carter [/bib_ref] [bib_ref] Immune responses to SARS-CoV-2 infection in hospitalized pediatric and adult patients, Pierce [/bib_ref] [bib_ref] Plasmacytoid dendritic cells depletion and elevation of IFN-γ dependent chemokines CXCL9 and..., Caldarale [/bib_ref] [bib_ref] Distinct cytokine and chemokine dysregulation in hospitalized children with acute COVID-19 and..., Peart Akindele [/bib_ref].
Overall, enhancement of these pro-inflammatory molecules in the circulation indicates inflammatory responses of myeloid and lymphoid cells. Endothelial cells could also contribute innate inflammatory mediators, as E-selectin, a marker of inflamed endothelial cells, shows elevation in the serum of patients with MIS-C 153 .
The reasons for the absence of some pro-inflammatory mediators in particular cohorts of patients are not known. The mediators that were analysed could have differed from study to study, but also, the levels of inflammatory mediators might vary depending on the patients' genetic and epigenetic backgrounds, severity of the disease, geographical location and timing of the analyses. Results from a study of plasma proteomics in children with SARS-CoV-2 infection, which have not yet undergone peer review, suggest that IFNγ expression is heterogeneous among patients with MIS-C, and that patients have dysregulated response to IFNγ 190 . As the pandemic progresses, it will be important to have a consensus regarding the panel of cytokines and chemokines that should be analysed in relation to MIS-C, to facilitate our understanding of the molecular pathogenesis and heterogeneity of this complex disease, and to enable accurate prognosis and effective treatment.
## Immune-cell profiles.
Immune-cell profiling of children with MIS-C or primary COVID-19 infection reveals similarities as well as differences in their immune signatures [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref]. They have similar proportions of eosinophils, immature granulocytes, monocytes and classic dendritic cells, but patients with MIS-C have elevation of neutrophils and reduction of plasmacytoid dendritic cells [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref] [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref] [bib_ref] Plasmacytoid dendritic cells depletion and elevation of IFN-γ dependent chemokines CXCL9 and..., Caldarale [/bib_ref] , which might contribute to the low levels of IFNα that are observed in the blood of patients with MIS-C relative to those with paediatric COVID-19 (ref. [bib_ref] Plasmacytoid dendritic cells depletion and elevation of IFN-γ dependent chemokines CXCL9 and..., Caldarale [/bib_ref]. Neutrophils and monocytes are activated in patients with MIS-C 186 and show upregulation of alarmin signatures (in particular S100A genes) and reduction of expression of antigen-presenting, antigen-processing and co-stimulatory molecules [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref] [bib_ref] Polyclonal expansion of TCR Vβ 21.3 + CD4 + and CD8 +..., Moreews [/bib_ref] [bib_ref] Peripheral immunophenotypes in children with multisystem inflammatory syndrome associated with SARS-CoV-2 infection, Carter [/bib_ref]. Compared with healthy children, those with MIS-C have greater expression of cytotoxicity genes and CCL4 in NK cells, which might contribute to the occurrence of tissue damage [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref]. Preliminary results suggest that plasma levels of IFNγ correlate with levels of NCR1 and IL-2RA, which are the soluble markers of activated NK and T cells, respectively.
MIS-C could have a common pathophysiology with Kawasaki disease involving NET formation, which has been described in the sera of adults with COVID-19 and with endothelial injuries or a prothrombotic state [bib_ref] Neutrophil extracellular traps in COVID-19, Zuo [/bib_ref] [bib_ref] Neutrophil extracellular traps contribute to immunothrombosis in COVID-19 acute respiratory distress syndrome, Middleton [/bib_ref] [bib_ref] Exaggerated neutrophil extracellular trap formation in Kawasaki disease: a key phenomenon behind..., Yamashita [/bib_ref]. However, plasma levels of NETs and release of NETs from neutrophils are similar in children with mild or moderate COVID-19 or MIS-C and in healthy children [bib_ref] Blood neutrophils from children with COVID-19 exhibit both inflammatory and anti-inflammatory markers, Seery [/bib_ref]. Despite similarities between disorders associated with pathogenic NETs and MIS-C, the role of NETosis in the pathogenesis of MIS-C remains uncertain because of a lack of definitive evidence, and hence further studies are warranted.
## Www.nature.com/nrrheum
Both MIS-C and paediatric COVID-19 present with general lymphopenia (affecting cells that include mucosa-associated invariant T cells, γδ T lymphocytes and CD8 + T lymphocytes) [bib_ref] Distinct clinical and immunological features of SARS-CoV-2-induced multisystem inflammatory syndrome in children, Lee [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children and COVID-19 are distinct presentations of SARS-CoV-2, Diorio [/bib_ref] [bib_ref] Mapping systemic inflammation and antibody responses in multisystem inflammatory syndrome in children..., Gruber [/bib_ref] [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref] [bib_ref] Peripheral immunophenotypes in children with multisystem inflammatory syndrome associated with SARS-CoV-2 infection, Carter [/bib_ref] [bib_ref] Immune responses to SARS-CoV-2 infection in hospitalized pediatric and adult patients, Pierce [/bib_ref] [bib_ref] Plasmacytoid dendritic cells depletion and elevation of IFN-γ dependent chemokines CXCL9 and..., Caldarale [/bib_ref]. Compared with paediatric COVID-19, in MIS-C there is more-pronounced CD4 + T cell-biased lymphopenia, which is similar to the situation in severely ill adults with COVID-19 (ref. [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref]. However, results from single-cell RNA sequencing analysis have revealed enhanced proliferation of CD4 + T cells in patients with MIS-C compared with healthy individuals [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref] , suggesting that lymphopenia might be the result of homing of T cells to the inflamed tissues. Despite showing T cell lymphopenia, the relative distribution of various T cell subsets such as naive, central memory and effector memory cells in patients with MIS-C is similar to that in age-matched healthy individuals, indicating a pan-CD4 + /CD8 + T cell lymphopenia, rather than a subset-specific effect [bib_ref] Mapping systemic inflammation and antibody responses in multisystem inflammatory syndrome in children..., Gruber [/bib_ref] [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref].
A distinct feature of MIS-C compared with paediatric COVID-19 is the activation of CX 3 CR1 + CD8 + T cells (CD8 + T cells that express vascular endothelium-homing CX 3 CR1, also known as fractalkine receptor), which could have implications for development of vascular abnormalities and cardiovascular abnormalities [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref]. This immunological phenotype is correlated with elevation of D-dimer, reduction of platelets and with the requirement for vasoactive medication. Although not as prominent as in NK cells, CD8 + T cells in MIS-C also show increases in signatures of cytotoxicity compared with those in healthy children [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref]. RNA sequencing in blood from children with MIS-C revealed aberrant NK and CD8 + T cell regulation, with depletion of NK cells and an absence of NK cell-dependent exhaustion of effector CD8 + T cells, which can lead to sustained inflammation [bib_ref] Downregulation of exhausted cytotoxic T cells in gene expression networks of multisystem..., Beckmann [/bib_ref]. Notably, the proportion of activated CX 3 CR1 + CD8 + T cells in patients with MIS-C decreases as the clinical status improves [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref]. Thus, there seems to be a sustained activation and dysregulation of CD8 + T cells, particularly those that express CX 3 CR1.
Nonspecific activation of B cell clones and expansion of plasmablasts occurs in MIS-C [bib_ref] Multisystem inflammatory syndrome in children and COVID-19 are distinct presentations of SARS-CoV-2, Diorio [/bib_ref] [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref] [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref] [bib_ref] Peripheral immunophenotypes in children with multisystem inflammatory syndrome associated with SARS-CoV-2 infection, Carter [/bib_ref]. Plasmablast elevation also occurs in children with COVID-19 (ref. [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref]. However, the specificity of expanded B cells and plasmablasts might vary between the two conditions. Patients with MIS-C display pronounced autoreactivity signatures of plasma immunoglobulins compared with healthy children or adults and children with COVID-19 . Also, patients with MIS-C have evidence of extrafollicular responses, as indicated by high frequencies of plasmablasts expressing the T box transcription factor T-bet [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref]. Future research should aim to uncover the reasons for this B cell activation, and should compare the characteristics of expanded B cells and plasmablasts, and the specificities of immunoglobulins, in MIS-C and paediatric COVID-19. As both conditions are associated with nonspecific B cell activation and elevation of plasmablast frequencies [bib_ref] Reaction of human monoclonal antibodies to SARS-CoV-2 proteins with tissue antigens: implications..., Vojdani [/bib_ref] , molecular mimicry between self-antigens and SARS-CoV-2 antigens (as described in a paper that has not yet been peer reviewed 197 ) might not be entirely responsible for the appearance of autoreactivity in MIS-C, and instead a combination of molecular mimicry and dysfunctional immunoregulatory machinery could be involved.
Humoral features. Analysis of IgG by systems serology has identified that humoral features in patients with MIS-C, such as complement deposition and neutrophil phagocytosis, overlap with those in convalescent adults with COVID-19 (ref. [bib_ref] This article reports that children with MIS-C maintain FcγR-binding, inflammatory monocyte/ macrophage-activating..., Bartsch [/bib_ref]. However, patients with severe MIS-C have persistent levels of FcγR binding (and in particular activating FcγRIIA) and inflammatory monocyte/macrophage-activating IgG [bib_ref] This article reports that children with MIS-C maintain FcγR-binding, inflammatory monocyte/ macrophage-activating..., Bartsch [/bib_ref]. Although hypergammaglobulinaemia is not observed in patients with MIS-C, a selective expansion of the IgG repertoire to react not only to SARS-CoV-2, but also to other bacterial and viral pathogens, some of which are implicated in the triggering of Kawasaki disease, has been observed. The underlying reason for the enrichment of particular IgG specificities is not yet known, but many of the microbes have been identified in the respiratory tracts of patients with MIS-C 198 , suggesting a role for an immune-complex-driven inflammatory response in the pathogenesis of MIS-C. IgG and IgA autoantibodies occur in patients with MIS-C, and recognize gastrointestinal, mucosal, immune-cell and endothelial antigens [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref] [bib_ref] Mapping systemic inflammation and antibody responses in multisystem inflammatory syndrome in children..., Gruber [/bib_ref]. Although the functionality of these autoantibodies and their roles in the pathogenesis of MIS-C should be investigated, these results might explain at least in part the involvement of multiple organ systems in MIS-C and provide a pointer towards dysregulated activation of B lymphocytes, enhanced autoreactivity and immune-complex-mediated inflammatory responses [fig_ref] TLRFigure 2 |: Possible mechanisms implicated in aberrant activation of immune cells in MIS-C [/fig_ref]. Enhanced expression of CD64 (FcγR1), a high-affinity receptor for the Fc fragment of IgG, has been observed on neutrophils and monocytes of patients with MIS-C [bib_ref] Mapping systemic inflammation and antibody responses in multisystem inflammatory syndrome in children..., Gruber [/bib_ref] [bib_ref] Peripheral immunophenotypes in children with multisystem inflammatory syndrome associated with SARS-CoV-2 infection, Carter [/bib_ref]. Furthermore, most of these patients respond to IVIG therapy [bib_ref] An outbreak of severe Kawasaki-like disease at the Italian epicentre of the..., Verdoni [/bib_ref] [bib_ref] Multisystem inflammatory syndrome related to COVID-19 in previously healthy children and adolescents..., Cheung [/bib_ref] [bib_ref] Acute heart failure in multisystem inflammatory syndrome in children (MIS-C) in the..., Belhadjer [/bib_ref] [bib_ref] Characteristics, cardiac involvement, and outcomes of multisystem inflammatory syndrome of childhood associated..., Capone [/bib_ref] [bib_ref] Clinical characteristics of 58 children with a pediatric inflammatory multisystem syndrome temporally..., Whittaker [/bib_ref] , which provides additional indirect support for the implication of FcγR-mediated activation of innate immune cells by immune complexes formed by these IgGs.
A role for the complement system in the pathogenesis of MIS-C has been suggested. Patients with MIS-C or paediatric COVID-19 have elevated plasma levels of soluble C5b-9 compared with healthy controls [bib_ref] Multisystem inflammatory syndrome in children and COVID-19 are distinct presentations of SARS-CoV-2, Diorio [/bib_ref] [bib_ref] Evidence of thrombotic microangiopathy in children with SARS-CoV-2 across the spectrum of..., Diorio [/bib_ref]. Soluble C5b-9 is a biomarker to monitor the activity of the terminal pathway of complement, and elevated levels suggest complement activation and endothelial dysfunction. Notably, although patients with MIS-C and paediatric COVID-19 have similar levels of complement-activating IgG antibodies to S protein of SARS-CoV-2 , those with MIS-C have enhanced autoreactive signatures of IgG [bib_ref] On the basis of systems-level analyses of various immune cells, cytokines and..., Consiglio [/bib_ref] [bib_ref] Immune dysregulation and autoreactivity correlate with disease severity in SARS-CoV-2-associated multisystem inflammatory..., Ramaswamy [/bib_ref] [bib_ref] Mapping systemic inflammation and antibody responses in multisystem inflammatory syndrome in children..., Gruber [/bib_ref]. As patients with MIS-C typically have minimal or no SARS-CoV-2 at the time of development of the disease, enhanced autoreactivity and immune-complex formation might contribute to the elevated levels of C5b-9. Consistent with complement activation, MIS-C is associated with clinical criteria for complement-mediated thrombotic microangiopathy, such as microangiopathic haemolytic anaemia, hypertension, thrombocytopenia, proteinuria and evidence of organ damage on the basis of lactate dehydrogenase elevation [bib_ref] Evidence of thrombotic microangiopathy in children with SARS-CoV-2 across the spectrum of..., Diorio [/bib_ref]. Compared with paediatric COVID-19, patients with MIS-C have higher incidence of thrombotic events [bib_ref] Rate of thrombosis in children and adolescents hospitalized with COVID-19 or MIS-C, Whitworth [/bib_ref]. Results from proteomics analyses of plasma samples, which have not yet been peer reviewed, suggest that phospholipase A2 (PLA2G2A) could be a biomarker for diagnosis of thrombotic microangiopathy in MIS-C 190 . The lectin complement pathway might also have an important role in the pathogenesis of diseases associated with SARS-CoV-2, as a result of the carbohydrate-residue-rich surface structures of the virus [bib_ref] Lectin complement system and pattern recognition, Endo [/bib_ref] [bib_ref] Site-specific glycan analysis of the SARS-CoV-2 spike, Watanabe [/bib_ref] [bib_ref] Does the lectin complement pathway link Kawasaki disease and SARS-CoV-2? Front, Polycarpou [/bib_ref].
## Therapeutic strategies
Treatment approaches to MIS-C aim to mute the exaggerated inflammatory response. Multiple approaches, borrowed from Kawasaki disease and other hyperinflammatory syndromes, have been considered, ranging from IVIG to glucocorticoids and immunotherapy [bib_ref] Multisystem inflammatory syndrome in children: survey of protocols for early hospital evaluation..., Dove [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children associated with coronavirus disease 2019 in a..., Jonat [/bib_ref]. MIS-C treatment regimens described in 24 studies, involving 1,020 individuals, are summarized in [fig_ref] Table 3 |: Treatment of MIS-C [/fig_ref] , highlighting the many variations on the theme of attempting to calm overactive inflammatory responses [bib_ref] The incidence of congenital heart disease, Hoffman [/bib_ref] [bib_ref] Multi-system inflammatory syndrome in children in association with COVID-19, Simpson [/bib_ref] [bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease: a scientific statement for..., Mccrindle Brian [/bib_ref] [bib_ref] Cytokines predict coronary aneurysm formation in Kawasaki disease patients, Lin [/bib_ref] [bib_ref] Netting neutrophils in autoimmune small-vessel vasculitis, Kessenbrock [/bib_ref] [bib_ref] NETosis as source of autoantigens in rheumatoid arthritis, Corsiero [/bib_ref] [bib_ref] The role of anti-endothelial cell antibodies in Kawasaki disease -in vitro and..., Grunebaum [/bib_ref] [bib_ref] On the basis of systems-level analyses of various immune cells, cytokines and..., Consiglio [/bib_ref] [bib_ref] Autoimmune aspects of Kawasaki disease, Sakurai [/bib_ref] [bib_ref] Anti-neutrophil cytoplasmic antibodies and anti-endothelial cell antibodies are not increased in Kawasaki..., Nash [/bib_ref] [bib_ref] Immunoglobulin G subclasses in serum and circulating immune complexes in patients with..., Li [/bib_ref] [bib_ref] Circulation immune complexes and granulocytes chemotaxis in Kawasaki disease: the 8th conference..., Miyata [/bib_ref] [bib_ref] Impaired granulocyte chemotaxis and increased circulating immune complexes in Kawasaki disease, Ono [/bib_ref] [bib_ref] Classical pathway complement activation in Kawasaki syndrome, Kohsaka [/bib_ref] [bib_ref] Polymorphisms in the mannose-binding lectin gene as determinants of age-defined risk of..., Biezeveld [/bib_ref] [bib_ref] High levels of IgA-containing circulating immune complex and secretory IgA in Kawasaki..., Ohshio [/bib_ref] [bib_ref] Kawasaki disease: pathophysiology and insights from mouse models, Rivas [/bib_ref] [bib_ref] Intestinal permeability and IgA provoke immune vasculitis linked to cardiovascular inflammation, Rivas [/bib_ref] [bib_ref] Evidence that SARS-CoV-2 spike protein acts as a superantigen leading to HLA..., Porritt [/bib_ref]. In most studies, most (70-100%) of the patients were treated with IVIG as the first-line agent, with satisfactory results. Steroids were the second most common treatment employed for patients with MIS-C.
Shock and cardiovascular manifestations comprise a predominant mode of presentation of MIS-C, and high-dose glucocorticoids have been advocated for, and used successfully in, patients with shock. Widely followed guidance from the ACR recommends IVIG as first-line therapy in hospitalized patients with MIS-C, with addition of glucocorticoids in the presence of shock, organ-threatening disease or refractory disease [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref]. In a study of 181 children with suspected MIS-C, IVIG alone had a higher failure rate than the use of IVIG with methylprednisolone (OR 0.25; 95% CI 0.09-0.70) [bib_ref] Association of intravenous immunoglobulins plus methylprednisolone vs immunoglobulins alone with course of..., Ouldali [/bib_ref]. By contrast, results from a multinational observational cohort study that involved 615 children with suspected MIS-C identified no difference in acute outcomes between primary treatment with IVIG alone, IVIG with steroids or steroids alone [bib_ref] Treatment of multisystem inflammatory syndrome in children, Mcardle [/bib_ref]. In view of the apparently important role of IL-1β in the pathogenesis of MIS-C, anakinra (an IL-1 receptor antagonist) has been used in MIS-C that is refractory to therapy with IVIG or steroids, extrapolating from its success in small groups of patients with IVIG-resistant Kawasaki disease [bib_ref] Acute heart failure in multisystem inflammatory syndrome in children (MIS-C) in the..., Belhadjer [/bib_ref] [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref] [bib_ref] Usefulness and safety of anakinra in refractory Kawasaki disease complicated by coronary..., Guillaume [/bib_ref] [bib_ref] The use of interleukin 1 receptor antagonist (anakinra) in Kawasaki disease: a..., Kone-Paut [/bib_ref].
Zonulin-dependent loss of intestinal mucosal permeability is implicated in mediation of the hyperinflammation observed in MIS-C, and accordingly, a patient who did not respond to anti-inflammatory therapies was treated with the zonulin antagonist larazotide, with a satisfactory outcome [bib_ref] Multisystem inflammatory syndrome in children is driven by zonulin-dependent loss of gut..., Yonker [/bib_ref]. In a pooled meta-analysis, D-dimer was found to be elevated in 92% of patients (330 out of 356) [bib_ref] COVID-19 and multisystem inflammatory syndrome in children and adolescents, Jiang [/bib_ref]. Because of the associated risk of hypercoagulability, and extrapolating from the management of Kawasaki disease, the use of anticoagulants such as acetylsalicylic acid and/or enoxaparin has been reported [bib_ref] Characteristics and outcomes of US children and adolescents with multisystem inflammatory syndrome..., Feldstein [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children related to COVID-19: a systematic review, Hoste [/bib_ref].
MIS-C: distinct from Kawasaki disease? Both Kawasaki disease and MIS-C have temporal associations with infectious diseases and are associated with immune-system alteration, systemic inflammation and cytokine storm. Myocardial dysfunction, which is seen in both pathologies, might be a consequence of systemic inflammation [bib_ref] Evaluation of left ventricular systolic strain in children with Kawasaki disease, Xu [/bib_ref]. An artificial intelligence computational analysis based on viral pandemics and disease-severity gene signatures, and in particular induction of IL15-IL15RA genes, has placed Kawasaki disease and MIS-C on the same host-immune-response continuum (although these results have not yet been peer reviewed). Consistently, patients with MIS-C have significantly higher levels of IL-15 than paediatric patients with COVID-19 (ref. [bib_ref] Distinct cytokine and chemokine dysregulation in hospitalized children with acute COVID-19 and..., Peart Akindele [/bib_ref]. However, the intensity of the immune response is high in MIS-C, which places it further along the severity spectrum than Kawasaki disease. A quarter to half of patients with MIS-C meet the full criteria for diagnosis of Kawasaki disease [bib_ref] Multi-system inflammatory syndrome in children in association with COVID-19, Simpson [/bib_ref] [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref] [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children related to COVID-19: a systematic review, Hoste [/bib_ref] [bib_ref] A single intravenous infusion of gamma globulin as compared with four infusions..., Newburger [/bib_ref]. Without evidence of prior SARS-CoV-2 exposure in these patients, it might not be possible to differentiate them from those with classic Kawasaki disease. Commonly reported clinical features of MIS-C include fever, mucocutaneous findings, myocardial dysfunction with cardiogenic or vasoplegic shock, gastrointestinal symptoms and neurological features including headache and altered mental status [fig_ref] Table 4 |: Reported clinical features of multisystem inflammatory syndrome in children [/fig_ref]. Like Kawasaki disease, these clinical manifestations are not specific to MIS-C, and they could occur in other infectious or inflammatory conditions 20 .
## Epidemiological and clinical differences
Despite the apparent similarities between MIS-C and Kawasaki disease, there are important epidemiological and clinical differences [bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease: a scientific statement for..., Mccrindle Brian [/bib_ref] [bib_ref] Acute heart failure in multisystem inflammatory syndrome in children (MIS-C) in the..., Belhadjer [/bib_ref] [bib_ref] Paediatric multisystem inflammatory syndrome temporally associated with SARS-CoV-2 mimicking Kawasaki disease (Kawa-COVID-19):..., Pouletty [/bib_ref]. Kawasaki disease is typically a disease of young children <5 years old, whereas MIS-C has been reported in a wide age range from 1.6 to 20 years, with a median age of 6-11 years [bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease: a scientific statement for..., Mccrindle Brian [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children associated with severe acute respiratory syndrome coronavirus..., Abrams [/bib_ref] [bib_ref] Emergence of Kawasaki disease related to SARS-CoV-2 infection in an epicentre of..., Ouldali [/bib_ref] [fig_ref] Table 2 |: Demographic features of MIS-C study populations [/fig_ref]. In sharp contrast to Kawasaki disease, there is a surprising lack of reports of MIS-C from Japan and East Asian countries [bib_ref] Surveillance of COVID-19-associated multisystem inflammatory syndrome in children, Choe [/bib_ref] [bib_ref] Why multisystem inflammatory syndrome in children has been less commonly described in..., Li [/bib_ref]. In fact, published data from the USA and Europe suggest that MIS-C is most commonly encountered in children of African and Hispanic heritage [bib_ref] Trends in geographic and temporal distribution of US children with multisystem inflammatory..., Belay [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children (MIS-C) and the coronavirus pandemic: current knowledge..., Rafferty [/bib_ref]. These epidemiological differences suggest that although MIS-C has phenotypic similarities to Kawasaki disease, they are essentially distinct syndromes.
Cardiac involvement is more prevalent and severe in MIS-C than in Kawasaki disease. Although a quarter of untreated patients with Kawasaki disease will develop coronary artery abnormalities, in the current era with a high level of clinical suspicion as well as early diagnosis and treatment, the incidence of coronary artery abnormalities in Kawasaki disease is <10% [bib_ref] A single intravenous infusion of gamma globulin as compared with four infusions..., Newburger [/bib_ref] [bib_ref] Treatment of Kawasaki disease: analysis of 27 US pediatric hospitals from, Son [/bib_ref] [bib_ref] Paediatric overdiagnosis modelled by coronary abnormality trends in Kawasaki disease, Coon [/bib_ref] [bib_ref] Can coronary artery involvement in Kawasaki disease be predicted, Ghelani [/bib_ref] [bib_ref] Arthritis presenting during the acute phase of Kawasaki disease, Gong [/bib_ref]. By contrast, our understanding of coronary artery dilation in MIS-C is still evolving, and incidence rates of 14-48% have been reported in various patient populations [bib_ref] Multisystem inflammatory syndrome in children (MIS-C) and the coronavirus pandemic: current knowledge..., Rafferty [/bib_ref] [bib_ref] Neutrophil extracellular traps in COVID-19, Zuo [/bib_ref] [bib_ref] Neutrophil extracellular traps contribute to immunothrombosis in COVID-19 acute respiratory distress syndrome, Middleton [/bib_ref] [fig_ref] Figure 3 |: Comparative incidence of clinical signs in MIS-C and Kawasaki disease [/fig_ref]. However, the adoption of standardized MIS-C management protocols has begun to reduce the rate of coronary artery involvement [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref]. Cardiac MRI in MIS-C has demonstrated high signal intensity on T1-weighted and T2-weighted imaging, consistent with diffuse myocardial oedema, with no enhancement on late gadolinium imaging to suggest fibrosis [bib_ref] Cardiac MRI in children with multisystem inflammatory syndrome associated with COVID-19, Blondiaux [/bib_ref]. Results from echocardiographic studies have demonstrated that global left ventricular longitudinal strain is significantly lower in individuals with MIS-C than in those with Kawasaki disease 226 . A longitudinal, single-centre study involving 15 children with MIS-C has demonstrated significant improvement towards normalization of both ventricular function and coronary artery size over a 30-day follow-up period [bib_ref] Longitudinal echocardiographic assessment of coronary arteries and left ventricular function following multisystem..., Jhaveri [/bib_ref].
Fewer than 10% of cases of Kawasaki disease manifest as Kawasaki disease-shock syndrome (KDSS), which requires the use of intravascular fluid resuscitation and vasoactive medication [bib_ref] Kawasaki disease shock syndrome: unique and severe subtype of Kawasaki disease, Gamez-Gonzalez [/bib_ref] [bib_ref] Changes in causes of sudden deaths by decade in patients with coronary..., Tsuda [/bib_ref] [bib_ref] Kawasaki disease: an unexpected etiology of shock and multiple organ dysfunction syndrome, Gatterre [/bib_ref]. Patients with KDSS tend to be older, have longer duration of fever and higher levels of inflammatory markers, and have a higher incidence of IVIG resistance as well as coronary abnormalities than those without KDSS [bib_ref] Kawasaki disease shock syndrome: clinical characteristics and possible use of IL-6, IL-10..., Li [/bib_ref] [bib_ref] Clinical manifestations of Kawasaki disease shock syndrome, Ma [/bib_ref]. By contrast, shock and depressed left ventricular systolic function are more frequent with MIS-C, for which reports indicate that 40-80% of patients present with shock 210,234-236 [fig_ref] Figure 3 |: Comparative incidence of clinical signs in MIS-C and Kawasaki disease [/fig_ref]. In a retrospective comparison of a cohort of patients with KDSS with published data relating to MIS-C, individuals with KDSS were more likely than those with MIS-C to fulfil the diagnostic criteria for complete Kawasaki disease, with higher incidence of coronary artery aneurysms [bib_ref] Kawasaki disease shock syndrome in Japan and comparison with multisystem inflammatory syndrome..., Suzuki [/bib_ref] Kawasaki disease has been reported to occur with MAS [bib_ref] Kawasaki disease followed by haemophagocytic syndrome, Kaneko [/bib_ref] [bib_ref] Kawasaki disease associated with reactive hemophagocytic lymphohistiocytosis, Cummings [/bib_ref]. In a retrospective analysis of 638 patients with Kawasaki disease, the incidence of MAS was <2% 240 . However, this figure is likely to be an underestimation of the true incidence, as a result of an absence of sensitive diagnostic criteria and a lack of awareness among health-care providers [bib_ref] Is macrophage activation syndrome in Kawasaki disease underrecognized?, Natoli [/bib_ref]. Patients with Kawasaki disease and MAS tend to have elevation of levels of IFNγ, TNF, serum neopterin, IL-18 and sTNFR-II 242 . A retrospective comparison of patients with MAS (as a complication of systemic-onset juvenile idiopathic arthritis) and MIS-C revealed that MAS was associated with lower levels of haemoglobin and fibrinogen, and higher ferritin and lactate dehydrogenase, whereas patients with MIS-C tended to have signs of shock and need of intensive care management [bib_ref] Comparison of baseline laboratory findings of macrophage activation syndrome complicating systemic juvenile..., Aydın [/bib_ref]. Results that have not yet been peer reviewed, based on analyses of IFNγ and CXCL9 signalling characteristics, suggest that >50% of patients with MIS-C have a MAS-like cytokine phenotype, along with elevation of CD163, IL-2RA and ferritin (during the early period) in the plasma. However, although MAS has an association with neutropenia, patients with MIS-C, including those who meet the criteria for MAS, display neutrophilia. Thus, although KDSS and Kawasaki disease with MAS have overlapping clinical features with MIS-C, there are subtle differences that are likely to reflect the different cytokine profiles in these conditions. Gastrointestinal and neurological symptoms are also more commonly encountered in MIS-C than in Kawasaki disease [bib_ref] Gastrointestinal symptoms as a major presentation component of a novel multisystem inflammatory..., Miller [/bib_ref] [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref] [bib_ref] Acute surgical abdomen as presenting manifestation of Kawasaki disease, Zulian [/bib_ref] [bib_ref] Small bowel pseudo-obstruction in Kawasaki disease, Miyake [/bib_ref] [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] [bib_ref] Diagnosis and management of Kawasaki disease, Saguil [/bib_ref] [bib_ref] COVID-19 and multisystem inflammatory syndrome in children and adolescents, Jiang [/bib_ref] [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] , with rare presentations that resemble appendicitis requiring surgical exploration [bib_ref] Gastrointestinal involvements in children with COVID-related multisystem inflammatory syndrome, Chen [/bib_ref]. In a national US registry consisting of 1,695 children and adolescents with active COVID-19 infections including Cardiac arrhythmias 2 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Abnormal ST-or T-wave segment 22 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Prolonged QT interval 2 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Pericardial effusion 13-28 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] Decreased LVEF by echo 31-58 [bib_ref] Review of cardiac involvement in multisystem inflammatory syndrome in children, Alsaied [/bib_ref] [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Increased troponin 68-95 [bib_ref] Review of cardiac involvement in multisystem inflammatory syndrome in children, Alsaied [/bib_ref] Ascites 21 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Ileitis 9 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Colitis 4 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Ophthalmological Conjunctivitis 32-83 [bib_ref] Review of cardiac involvement in multisystem inflammatory syndrome in children, Alsaied [/bib_ref] [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref] [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] [bib_ref] Mucocutaneous manifestations of multisystem inflammatory syndrome in children during the COVID-19 pandemic, Young [/bib_ref] Periorbital erythema and oedema 20 [bib_ref] Mucocutaneous manifestations of multisystem inflammatory syndrome in children during the COVID-19 pandemic, Young [/bib_ref] Nervous system Neurological symptoms 13-35 [bib_ref] COVID-19 and multisystem inflammatory syndrome in children and adolescents, Jiang [/bib_ref] [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref] [bib_ref] Neurologic involvement in children and adolescents hospitalized in the United States for..., Larovere [/bib_ref] Severe symptoms, including encephalopathy, stroke, central nervous system infection/demyelination, Guillain-Barré syndrome and acute cerebral oedema 3 [bib_ref] Neurologic involvement in children and adolescents hospitalized in the United States for..., Larovere [/bib_ref] Integumentary Rash 50-70 [bib_ref] Review of cardiac involvement in multisystem inflammatory syndrome in children, Alsaied [/bib_ref] [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref] [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] Erythematous skin rash 62 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Hyperaemia, oedema or desquamation of extremities 26-51 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] [bib_ref] Mucocutaneous manifestations of multisystem inflammatory syndrome in children during the COVID-19 pandemic, Young [/bib_ref] Malar erythema 17 [bib_ref] Mucocutaneous manifestations of multisystem inflammatory syndrome in children during the COVID-19 pandemic, Young [/bib_ref] Skin eruptions 9-14 [bib_ref] Mucocutaneous manifestations of multisystem inflammatory syndrome in children during the COVID-19 pandemic, Young [/bib_ref] Desquamation in groin 26 [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref]
## Respiratory
Upper respiratory tract infection 34 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Lower respiratory tract infection 22 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Pleural effusion on CT 20 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Lung involvement on CT (bilateral pulmonary consolidation and ground-glass opacity) 13 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Mucosal Oral mucosa hyperaemia 41 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] Red and/or cracked lips 37-49 [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] [bib_ref] Mucocutaneous manifestations of multisystem inflammatory syndrome in children during the COVID-19 pandemic, Young [/bib_ref] Strawberry tongue 11-23 [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] [bib_ref] Mucocutaneous manifestations of multisystem inflammatory syndrome in children during the COVID-19 pandemic, Young [/bib_ref] Lips and oral-cavity changes 74 [bib_ref] Review of cardiac involvement in multisystem inflammatory syndrome in children, Alsaied [/bib_ref] [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref] Other Lymphadenopathy (cervical) 19-61 [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref] [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] Extremity changes 8-52 [bib_ref] Review of cardiac involvement in multisystem inflammatory syndrome in children, Alsaied [/bib_ref] [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref] [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] LVEF, left ventricular ejection fraction.
www.nature.com/nrrheum MIS-C (n = 616), neurological symptoms were noted in 22% of the patient population (n = 365), with most of those affected having transient symptoms [bib_ref] Neurologic involvement in children and adolescents hospitalized in the United States for..., Larovere [/bib_ref]. Among these 365 patients, 126 met the criteria for MIS-C. In the patients with neurological involvement (n = 365), 43 (12%) had life-threatening neurological involvement (including encephalopathy, stroke, central nervous system infection and/or demyelination, Guillain-Barré syndrome and acute cerebral oedema) and among which 20 (47%) met the criteria for MIS-C. In a study of 286 children with MIS-C located in 55 centres in 17 European countries, neurological involvement was identified in 43 individuals (15%) [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref]. In a pooled meta-analysis of data from 370 children with MIS-C, 133 (35.9%) had neurological symptoms [bib_ref] COVID-19 and multisystem inflammatory syndrome in children and adolescents, Jiang [/bib_ref] [fig_ref] Table 4 |: Reported clinical features of multisystem inflammatory syndrome in children [/fig_ref]. Neurological involvement in Kawasaki disease is variable, reportedly affecting 5-39% of patients [bib_ref] Cerebrospinal fluid profile in patients with acute Kawasaki disease, Dengler [/bib_ref] [bib_ref] Neurological involvement in Kawasaki disease: a retrospective study, Liu [/bib_ref].
In contrast to those with Kawasaki disease, patients with MIS-C tend to have a worse acute clinical course and multisystem involvement, as illustrated by an increased requirement forintensive care management. A large study of >1,000 patients with Kawasaki disease revealed that 2.4% of these children required intensive care [bib_ref] Characteristics of children with Kawasaki disease requiring intensive care: 10 years' experience..., Kuo [/bib_ref]. In stark contrast, an analysis of 783 cases of MIS-C determined that 68% of patients required intensive care admission, 63% needed inotropic support, 28% had some form of respiratory support and 4% of patients required extra-corporeal membrane oxygenation [bib_ref] Multi-system inflammatory syndrome in children & adolescents (MIS-C): a systematic review of..., Radia [/bib_ref]. Among 1,080 patients with MIS-C, intensive care admission was more likely in children aged >5 years old than in younger children, and in non-Hispanic Black patients than in non-Hispanic white patients, and coronary artery abnormalities were more common in boys than in girls [bib_ref] Factors linked to severe outcomes in multisystem inflammatory syndrome in children (MIS-C)..., Abrams [/bib_ref]. Elevated acute-phase inflammatory markers, troponin, B-type natriuretic peptide and D-dimer levels also identified patents at risk of severe disease [bib_ref] Factors linked to severe outcomes in multisystem inflammatory syndrome in children (MIS-C)..., Abrams [/bib_ref]. In a study evaluating 29 children with MIS-C in France, severe disease occurred in 52% of them and was associated with high persistent fever and high levels of inflammatory markers [bib_ref] Factors associated with severe SARS-CoV-2 infection, Ouldali [/bib_ref]. Although the reason for a more critical illness in the acute phase of MIS-C than in Kawasaki disease is unclear, it is thought to be linked to the cytokine storm in MIS-C [bib_ref] Pediatric Crohn disease and multisystem inflammatory syndrome in children (MIS-C) and COVID-19..., Dolinger [/bib_ref].
## Immunological differences
In several small studies, comparative immune profiling of children with MIS-C and Kawasaki disease has been performed, to differentiate between these two disease entities. MIS-C is associated with lymphopenia, lower white blood cell and naive CD4 + T cell counts, and increased central and effector memory T cell subpopulations, compared with Kawasaki disease 103 . IL-17 is a mediator of inflammation in Kawasaki disease, but is less prominent in MIS-C [bib_ref] On the basis of systems-level analyses of various immune cells, cytokines and..., Consiglio [/bib_ref]. In a comparison of cytokine profiles, levels of circulating IFNγ were significantly higher in patients with severe forms of MIS-C than in those with milder MIS-C or Kawasaki disease [bib_ref] Similarities and differences between the immunopathogenesis of COVID-19-related pediatric multisystem inflammatory syndrome..., Esteve-Sole [/bib_ref].
In a study of the immunological profiles of paediatric patients, 75% of those with MIS-C, but none with Kawasaki disease, TSS or COVID-19, displayed non-HLA-biased, SARS-CoV-2 non-reactive, polyclonal expansion of TCR Vβ 21.3 + activated CD4 + and CD8 + T cells [bib_ref] Polyclonal expansion of TCR Vβ 21.3 + CD4 + and CD8 +..., Moreews [/bib_ref]. Notably, these Vβ 21.3 + T cells had high expression of CX 3 CR1, a marker previously identified on the activated CD8 + T cells of patients with MIS-C [bib_ref] Deep immune profiling of MIS-C demonstrates marked but transient immune activation compared..., Vella [/bib_ref]. The remarkable specificity of Vβ 21.3 + T cell subset expansion noted in MIS-C is consistent with superantigen-mediated activation of the immune system [bib_ref] Superantigenic character of an insert unique to SARS-CoV-2 spike supported by skewed..., Cheng [/bib_ref] , whereas in Kawasaki disease, evidence of a role of superantigens in pathogenesis is lacking [bib_ref] TCR Vβ family repertoire and T cell activation markers in Kawasaki disease, Pietra [/bib_ref].
Autoantibody profiles have been compared in patients with MIS-C and Kawasaki disease 103 . Levels of antibodies to some vascular endothelial cell proteins, such as endoglin, were higher in both groups of patients than in healthy controls, whereas some autoantibodies (such as that to EGF-like repeat and discoidin I-like domain-containing protein 3) were overexpressed in Kawasaki disease compared with MIS-C. To confound matters, plasma levels of endoglin were elevated in both sets of patients compared with healthy children, raising the possibility that antibodies to endothelial cells were the result, rather than the cause, of vascular damage. Another possibility is that the S protein superantigen of SARS-CoV-2 might cause aberrant activation of B cells [bib_ref] COVID-19-associated multisystem inflammatory syndrome in children (MIS-C): a novel disease that mimics..., Rivas [/bib_ref].
Some laboratory parameters are important differentiators between Kawasaki disease and MIS-C. Although both syndromes involve a diffuse hyperinflammatory response, patients with MIS-C tend to have a lower platelet count, lower absolute lymphocyte count and higher levels of C-reactive protein, N-terminal pro-B-type natriuretic peptide, troponin and ferritin [bib_ref] Autoimmune and inflammatory diseases following COVID-19, Galeotti [/bib_ref] [bib_ref] Multi-system inflammatory syndrome in children in association with COVID-19, Simpson [/bib_ref] [bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease: a scientific statement for..., Mccrindle Brian [/bib_ref] [bib_ref] Cytokines predict coronary aneurysm formation in Kawasaki disease patients, Lin [/bib_ref] [bib_ref] Netting neutrophils in autoimmune small-vessel vasculitis, Kessenbrock [/bib_ref] [bib_ref] NETosis as source of autoantigens in rheumatoid arthritis, Corsiero [/bib_ref] [bib_ref] The role of anti-endothelial cell antibodies in Kawasaki disease -in vitro and..., Grunebaum [/bib_ref] [bib_ref] On the basis of systems-level analyses of various immune cells, cytokines and..., Consiglio [/bib_ref] [bib_ref] Autoimmune aspects of Kawasaki disease, Sakurai [/bib_ref] [bib_ref] Anti-neutrophil cytoplasmic antibodies and anti-endothelial cell antibodies are not increased in Kawasaki..., Nash [/bib_ref] [bib_ref] Immunoglobulin G subclasses in serum and circulating immune complexes in patients with..., Li [/bib_ref] [bib_ref] Circulation immune complexes and granulocytes chemotaxis in Kawasaki disease: the 8th conference..., Miyata [/bib_ref] [bib_ref] Impaired granulocyte chemotaxis and increased circulating immune complexes in Kawasaki disease, Ono [/bib_ref] [bib_ref] Classical pathway complement activation in Kawasaki syndrome, Kohsaka [/bib_ref] [bib_ref] Polymorphisms in the mannose-binding lectin gene as determinants of age-defined risk of..., Biezeveld [/bib_ref] [bib_ref] High levels of IgA-containing circulating immune complex and secretory IgA in Kawasaki..., Ohshio [/bib_ref] [bib_ref] Kawasaki disease: pathophysiology and insights from mouse models, Rivas [/bib_ref] [bib_ref] Intestinal permeability and IgA provoke immune vasculitis linked to cardiovascular inflammation, Rivas [/bib_ref] [bib_ref] Evidence that SARS-CoV-2 spike protein acts as a superantigen leading to HLA..., Porritt [/bib_ref]. Additionally, coagulation abnormalities are common, including elevation of D-dimer and fibrinogen levels [bib_ref] An outbreak of severe Kawasaki-like disease at the Italian epicentre of the..., Verdoni [/bib_ref] [bib_ref] Hyperinflammatory shock in children during COVID-19 pandemic, Riphagen [/bib_ref] [bib_ref] Pediatric Crohn disease and multisystem inflammatory syndrome in children (MIS-C) and COVID-19..., Dolinger [/bib_ref] [bib_ref] Multisystem inflammatory syndrome in children (MIS-C) and the prothrombotic state: coagulation profiles..., Al-Ghafry [/bib_ref]. Hyponatraemia is another common laboratory finding in patients with MIS-C [bib_ref] Paediatric multisystem inflammatory syndrome temporally associated with SARS-CoV-2 mimicking Kawasaki disease (Kawa-COVID-19):..., Pouletty [/bib_ref] [bib_ref] COVID-19 and multisystem inflammatory syndrome in children and adolescents, Jiang [/bib_ref] [bib_ref] Kawasaki disease shock syndrome: clinical characteristics and possible use of IL-6, IL-10..., Li [/bib_ref] [bib_ref] Review of cardiac involvement in multisystem inflammatory syndrome in children, Alsaied [/bib_ref] [bib_ref] Acute cardiovascular manifestations in 286 children with multisystem inflammatory syndrome associated with..., Valverde [/bib_ref] [bib_ref] Distinctive features of Kawasaki disease following SARS-CoV-2 infection: a controlled study, Toubiana [/bib_ref] [bib_ref] Oral manifestations of COVID-2019-related multisystem inflammatory syndrome in children: a review of..., Halepas [/bib_ref] [bib_ref] Neurologic involvement in children and adolescents hospitalized in the United States for..., Larovere [/bib_ref] [bib_ref] Cerebrospinal fluid profile in patients with acute Kawasaki disease, Dengler [/bib_ref] [bib_ref] Mucocutaneous manifestations of multisystem inflammatory syndrome in children during the COVID-19 pandemic, Young [/bib_ref] [bib_ref] Diagnosis, treatment, and long-term management of Kawasaki disease, Newburger Jane [/bib_ref] [bib_ref] Clinical and epidemiological characteristics of children with Kawasaki disease in Turkey, Ozdemir [/bib_ref] [bib_ref] Incidence and clinical features of incomplete Kawasaki disease, Fukushige [/bib_ref] [bib_ref] Lack of association of cervical lymphadenopathy and coronary artery complications in Kawasaki..., Sung [/bib_ref] [bib_ref] Kawasaki syndrome and the eye, Smith [/bib_ref] [bib_ref] Associated symptoms in the ten days before diagnosis of Kawasaki disease, Baker [/bib_ref]. Although some clinical signs, such as fever and cervical lymphadenopathy are equally prevalent in both MIS-C and Kawasaki disease, the incidence of other symptoms, including shock, coronary artery involvement and gastrointestinal symptoms (vomiting, diarrhoea or abdominal pain), are characteristic of MIS-C. a 'Conjunctival injection' refers to bilateral non-exudative conjunctivitis in Kawasaki disease. b 'Rash' refers to polymorphous rash in Kawasaki disease. c 'Coronary artery dilation of aneurysm' refers to incidence in untreated cases of Kawasaki disease.
of burr cells and neutrophils with toxic granulation can also discriminate MIS-C from severe COVID-19 (ref. [bib_ref] Multisystem inflammatory syndrome in children and COVID-19 are distinct presentations of SARS-CoV-2, Diorio [/bib_ref]. Finally, evidence of recent SARS-CoV-2 infection, particularly by positive serology, is a diagnostic indicator of MIS-C [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref] [bib_ref] American College of Rheumatology clinical guidance for multisystem inflammatory syndrome in children..., Henderson [/bib_ref].
# Conclusion
Epidemiological and clinical differences reveal that although MIS-C has phenotypic similarities to Kawasaki disease, they are different syndromes. They have varying degrees of hyperinflammation and dysregulated immune responses [bib_ref] Clinical characteristics of 58 children with a pediatric inflammatory multisystem syndrome temporally..., Whittaker [/bib_ref]. Children with MIS-C are, in general, more critically ill, with prominent gastrointestinal symptoms, cardiac involvement with shock, haematological abnormalities and elevated acute-phase reactants. They have positive SARS-CoV-2 serology, suggesting a link to prior clinical or subclinical infection or exposure. It is likely that a combination of pathogen and host factors is involved in the genesis of an intense aberrant activation of both innate and adaptive immune responses and subsequent cytokine storm [bib_ref] Autoimmune and inflammatory diseases following COVID-19, Galeotti [/bib_ref] [bib_ref] Childhood multisystem inflammatory syndrome -a new challenge in the pandemic, Levin [/bib_ref]. Some of the pathogen-related factors include antigen mimicry of host antigens, and superantigen properties of viral proteins. Potential host factors include age and immune-system immaturity, altered intestinal microbiota, nutritional deficiencies and genetic (including inborn errors of immunity) and epigenetic predisposition [bib_ref] Multisystem inflammatory syndrome in U.S. children and adolescents, Feldstein [/bib_ref]. However, to what extent each of these factors contributes, and how they interact to cause the clinical syndrome, are relative unknowns that need further exploration. Various lines of evidence based on inflammatory parameters, clinical signs or gene analyses have evoked the possibility that MIS-C is a heterogeneous complex disorder. Current data on the immune signatures in patients with MIS-C are based on small sample size, non-homogeneous cohorts. Hence, analysis of dynamic changes in the immune signatures of patients with MIS-C and their comparison with those with Kawasaki disease in a large homogeneous cohort is needed, to accurately determine the similarities and distinct features of the two disease entities. Nevertheless, with the evidence of elevation of signatures of autoimmunity in MIS-C, and reports of various post-COVID-19 conditions in adults, long-term follow-up of patients with MIS-C might be advisable, because of the possibility of relapse. Notably, however, relapse is rare in Kawasaki disease, which might suggest that recurrence is also unlikely in MIS-C.
While researchers and clinicians navigate the possibilities and evaluate the best treatment options for patients affected by COVID-19-related illnesses, it is imperative to establish registries and dedicated multidisciplinary research teams to investigate the pathogenesis and specific therapeutic strategies in MIS-C. An important step towards this end was the workshop that was convened by the NIH in June 2020, which aimed to bring together the experts on the subject, to initiate dialogue leading to future studies [bib_ref] The immune roadmap for understanding multi-system inflammatory syndrome in children: opportunities and..., Martinez [/bib_ref]. In conclusion, MIS-C has important epidemiological, clinical and immunological differences from Kawasaki disease, enabling its classification as a separate syndrome. Study of MIS-C will continue to enhance our understanding of these conditions that are related by their association with the cytokine storm phenomenon.
Published online 29 October 2021
[fig] 263 IQR: interquartile range; MIS-C, multisystem inflammatory syndrome in children. NATure revIeWS | RheuMATOlOGy exposure and development of MIS-C, as a spike in MIS-C cases occurs 3-6 weeks after the peak of SARS-CoV-2 transmission in a community 13,129,147 (fig. 1). Median intervals of 21 and 25 days have been observed between the occurrence of COVID-19 symptoms and the onset of MIS-C 13,14 . [/fig]
[fig] Figure 1 |: The temporal relationship between SARS-CoV-2 infection and development of MIS-C. Evidence suggests that a relationship exists between the timing of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and development of multisystem inflammatory syndrome in children (MIS-C). Cases of MIS-C tend to be seen 3-6 weeks after the peak of SARS-CoV-2 transmission in a community. Because of this time lag, MIS-C is associated with a strong anti-spike protein IgG response, but a weak IgM response. It should be noted that implication of SARS-CoV-2 as a triggering factor for the development of MIS-C has yet to be firmly established.www.nature.com/nrrheum 0123456789();: [/fig]
[fig] TLRFigure 2 |: Possible mechanisms implicated in aberrant activation of immune cells in MIS-C.Clinical signs of multisystem inflammatory syndrome in children (MIS-C) mostly appear several weeks after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. MIS-C might be triggered by dysregulation of immune responses following viral infection. Aberrant activation of immune cells in patients with MIS-C could result from several factors. Infection with particular variants of SARS-CoV-2 might trigger hyperinflammatory responses. Genetic predisposition resulting from variants in the genes that encode pattern recognition receptors, Fcγ receptors and components of the signalling cascades of immune response, as well as mutations in genes such as SOCS1, which regulate inflammatory responses, could all contribute to enhancement of inflammatory responses to infection. Dysregulated activation of lymphocytes, with production of IgG corresponding to microbial pathogens or autoantigens, could cause immune-complex-mediated innate-cell activation by signalling via Fcγ receptors. Production of autoantibodies could also lead to complement activation and autoantibody-mediated endothelial damage. SARS-CoV-2 spike (S) protein might function as a superantigen, contributing to activation of T cells. SOCS1, suppressor of cytokine signalling 1; TLR, Toll-like receptor. [/fig]
[fig] Figure 3 |: Comparative incidence of clinical signs in MIS-C and Kawasaki disease. Percentage incidence of particular symptoms in patients with multisystem inflammatory syndrome in children (MIS-C) or Kawasaki disease is shown, with the values derived from published reports [/fig]
[table] Table 1 |: Comparison of Kawasaki disease and MIS-C [/table]
[table] Table 2 |: Demographic features of MIS-C study populations [/table]
[table] Table 3 |: Treatment of MIS-C [/table]
[table] Table 4 |: Reported clinical features of multisystem inflammatory syndrome in children [/table]
|
Identification and functional characterization of the CYP51 gene from the yeast Xanthophyllomyces dendrorhous that is involved in ergosterol biosynthesis
Background: Xanthophyllomyces dendrorhous is a basidiomycetous yeast that synthesizes astaxanthin, a carotenoid with great biotechnological impact. The ergosterol and carotenoid synthetic pathways derive from the mevalonate pathway and involve cytochrome P450 enzymes. Among these enzymes, the CYP51 family, which is involved in ergosterol biosynthesis, is one of the most remarkable that has C14-demethylase activity. Results: In this study, the CYP51 gene from X. dendrorhous was isolated and its function was analyzed. The gene is composed of ten exons and encodes a predicted 550 amino acid polypeptide that exhibits conserved cytochrome P450 structural characteristics and shares significant identity with the sterol C14-demethylase from other fungi. The functionality of this gene was confirmed by heterologous complementation in S. cerevisiae. Furthermore, a CYP51 gene mutation in X. dendrorhous reduced sterol production by approximately 40% and enhanced total carotenoid production by approximately 90% compared to the wild-type strain after 48 and 120 h of culture, respectively. Additionally, the CYP51 gene mutation in X. dendrorhous increased HMGR (hydroxy-methylglutaryl-CoA reductase, involved in the mevalonate pathway) and crtR (cytochrome P450 reductase) transcript levels, which could be associated with reduced ergosterol production. Conclusions: These results suggest that the CYP51 gene identified in X. dendrorhous encodes a functional sterol C14-demethylase that is involved in ergosterol biosynthesis.
# Background
Xanthophyllomyces dendrorhous is a basidiomycete yeast that has been mostly studied for its ability to synthesize xanthophyll astaxanthin as its main carotenoid. Astaxanthin (3,3'-dihydroxy-β,β-carotene-4-4'-dione) takes third place in the global market for carotenoids, which reached $226 million in the year 2010 and is expected to exceed $250 million by the year 2018. This carotenoid is currently used in aquaculture for salmon flesh pigmentation and as a supplement for the diets of farm chickens to produce a stronger yolk and flesh pigmentation [bib_ref] Biotechnological production of astaxanthin with Phaffia rhodozyma/ Xanthophyllomyces dendrorhous, Schmidt [/bib_ref]. Additionally, astaxanthin has strong antioxidant properties and has been proposed to play a protective role against oxidative stress in yeast [bib_ref] Antioxidant role of carotenoids in Phaffia rhodozyma, Schroeder [/bib_ref] [bib_ref] Carotenoids protect Phaffia rhodozyma against singlet oxygen damage, Schroeder [/bib_ref] [bib_ref] Singlet oxygen and peroxyl radicals regulate carotenoid biosynthesis in Phaffia rhodozyma, Schroeder [/bib_ref]. Recent findings suggested that astaxanthin supplementation could be beneficial as a treatment for several degenerative diseases in addition to other potential benefits for human health [bib_ref] Astaxanthin-rich algal meal and vitamin C inhibit Helicobacterpylori infection in BALB/cA mice, Wang [/bib_ref] [bib_ref] Astaxanthin: a review of its chemistry and applications, Higuera-Ciapara [/bib_ref] [bib_ref] Astaxanthin decreased oxidative stress and inflammation and enhanced immune response in humans, Park [/bib_ref] [bib_ref] Dietary astaxanthin inhibits colitis and colitis-associated colon carcinogenesis in mice via modulation..., Yasui [/bib_ref]. To date, X. dendrorhous is the only known organism that synthesizes astaxanthin from β-carotene through a cytochrome P450 enzyme system [bib_ref] The crtS gene of Xanthophyllomyces dendrorhous encodes a novel cytochrome-P450 hydroxylase involved..., Alvarez [/bib_ref] [bib_ref] Cloning of the astaxanthin synthase gene from Xanthophyllomyces dendrorhous (Phaffia rhodozyma) and..., Ojima [/bib_ref] [bib_ref] Cloning of the cytochrome p450 reductase (crtR) gene and its involvement in..., Alcaino [/bib_ref] [bib_ref] Expression of Xanthophyllomyces dendrorhous cytochrome-P450 hydroxylase and reductase in Muco rcircinelloides, Csernetics [/bib_ref] , suggesting that this yeast has developed a unique P450 system specialized for the production of this carotenoid.
Cytochrome P450 proteins (P450s or CYPs) represent a large superfamily of heme-containing monooxygenases that are distributed throughout three domains of life [bib_ref] Cytochrome b5 increases the rate of product formation by cytochrome P450 2B4..., Zhang [/bib_ref] and play significant roles in the oxidative metabolism of a wide range of exogenous and endogenous substrates [bib_ref] Molecular evolution of P450 superfamily and P450-containing monooxygenase systems, Degtyarenko [/bib_ref]. CYP enzymes are involved in the biosynthesis of many physiologically important compounds, such as sterols, steroid hormones, fatty acids and vitamins [bib_ref] Cytochromes P450 as versatile biocatalysts, Bernhardt [/bib_ref] , and a vast array of secondary metabolites in plants, insects and fungi [bib_ref] A passion for P450s (remembrances of the early history of research on..., Estabrook [/bib_ref]. Additionally, these enzymes are the main catalysts involved in the activation and detoxification of different xenobiotics, such as drugs, pesticides, carcinogens and environmentally contaminating chemicals [bib_ref] Cytochrome P-450. Multiplicity of isoforms, substrates, and catalytic and regulatory mechanisms, Porter [/bib_ref]. CYP enzymes require the reducing equivalents from NADPH or NADH, which are generally transferred to CYPs through a redox partner to catalyze the general reaction RH + O 2 + 2e − + 2H + → ROH + H 2 O [bib_ref] Electron transfer between cytochrome P450cin and its FMN-containing redox partner, cindoxin, Kimmich [/bib_ref].
Only two CYP enzymes have been described in X. dendrorhous: the astaxanthin synthase that is involved in astaxanthin biosynthesis and CYP61, which participates in ergosterol biosynthesis [bib_ref] Enhancement of carotenoid production by disrupting the C22-sterol desaturase gene (CYP61) in..., Loto [/bib_ref]. Considering the biological relevance of this protein family and the involvement of a unique system in X. dendrorhous, in this work we describe a third CYP in this yeast: CYP51. CYP51 is one of the first modifying enzymes involved in sterol synthesis. It catalyzes the C14 demethylation of lanosterol in yeasts during ergosterol biosynthesis and is considered to be an ancestral CYP from which other CYPs evolved [bib_ref] Microbial cytochromes P450: biodiversity and biotechnology. Where do cytochromes P450 come from,..., Kelly [/bib_ref].
# Results and discussion
Cloning and sequence analysis of the X. dendrorhous CYP51 gene
Using bioinformatic analyses of the genomic and transcriptomic data of X. dendrorhous available in our laboratory with the S. cerevisiae ERG11 gene as a query [GenBank: NM_001179137], we identified a putative X. dendrorhous CYP51 gene. This gene was uploaded into the Genbank database [GenBank: KP317478]. The primers CYP51 up.F and CYP51dw.R were designed based on this gene sequence and used to amplify genomic DNA from strain UCD 67-385. The obtained PCR-product of approximately 4,300 bp was inserted into the EcoRV site of the pBluescript SK-plasmid, generating plasmid pBS-gCYP51. Similarly, the cDNA version (ORF: 1,653 bp) of the CYP51 gene was obtained by RT-PCR from X. dendrorhous total RNA using the primer set cCYP51.F + cCYP51.R. This cDNA was ligated into the EcoRV site of the pBluescript SK-plasmid to generate plasmid pBS-cCYP51. The sequences of the genomic and cDNA versions were determined in both strands and compared, revealing that the CYP51 gene from X. dendrorhous consists of [bib_ref] The crtS gene of Xanthophyllomyces dendrorhous encodes a novel cytochrome-P450 hydroxylase involved..., Alvarez [/bib_ref] , which belongs to the cytochrome P450 protein family and is involved in the third step of ergosterol biosynthesis from squalene: the conversion of lanosterol to 4,4-dimethylcholesta-8,14,24-trienol [bib_ref] Yeast cytochrome P-450 catalyzing lanosterol 14 alpha-demethylation. II. Lanosterol metabolism by purified..., Aoyama [/bib_ref] [bib_ref] Primary structure of the P450 lanosterol demethylase gene from Saccharomyces cerevisiae, Kalb [/bib_ref].
Amino acid sequences are highly diverse among the cytochrome P450 protein family; however, their structural fold is highly conserved [bib_ref] Cytochromes P450: a success story, Werck-Reichhart [/bib_ref]. In the deduced CYP51 protein from X. dendrorhous, several cytochrome P450 secondary structural elements were predicted using the CYP450 Engineering database, including alpha helices A, B, C, E, F, G, H, I, J, K, K' and L, beta-sheets 1-1,1-2, 1-5, 3-1, 1-4, 2-1, 2-2, 1-3, 3-3, 4-1, 4-2 and 3-2, the meander loop involved in the stabilization of the tertiary structure and heme binding, and the Cys pocket that contains the conserved cysteine involved in heme binding [bib_ref] The Cytochrome P450 Engineering Database: a navigation and prediction tool for the..., Fischer [/bib_ref]. Identification of P450s across biological kingdoms depends largely on the identification of two P450 signature motifs: ExxR at the K-helix and FxxGxRxCxG (also known as the CxG motif ) at the Cys-pocket. These motifs contain three totally conserved amino acids: the glutamic acid and arginine of the ExxR motif and the cysteine in the Cys pocket that serves as a fifth ligand for the heme iron [bib_ref] The cytochrome P450 engineering database: Integration of biochemical properties, Sirim [/bib_ref]. Site-directed mutagenesis of these three invariant residues in most CYPs results in inactive and misfolded P450 isoforms, suggesting their importance in maintaining the P450 structure [bib_ref] Probing the role of lysines and arginines in the catalytic function of..., Shimizu [/bib_ref] [bib_ref] Site-directed mutagenesis of human prostacyclin synthase: Alteration of Cys441 of the Cys-pocket,..., Hatae [/bib_ref]. All of these motifs were recognized in the X. dendrorhous deduced CYP51 protein, including the three invariant P450 residues. A threonine and a leucine were found at the second and third positions of the ExxR motif; these are the preferred residues at these positions in P450s, including the CYP51 family [bib_ref] Comparative analysis of P450 signature motifs EXXR and CXG in the large..., Syed [/bib_ref]. Moreover, the CxG motif of the X. dendrorhous CYP51 protein (FGAGRHRCIG) maintains the preferred residues described for the CYP51 family [bib_ref] Comparative analysis of P450 signature motifs EXXR and CXG in the large..., Syed [/bib_ref]. This finding is interesting because instead of the conserved arginine at the sixth position of this motif, CYP51 contains a histidine surrounded by arginines. Additionally, a putative hydrophobic trans-membrane segment at the CYP51 amino terminus was predicted. This finding represents an important feature that allows class II P450 enzymes to anchor to the endoplasmic reticulum [bib_ref] Cytochrome P450 enzyme systems in fungi, Van Den Brink [/bib_ref]. These observations strongly suggest that the identified X. dendrorhous gene encodes a cytochrome P450 enzyme, specifically CYP51 with lanosterol demethylase activity.
The deduced X. dendrorhous CYP51 protein was modeled [fig_ref] Figure 2: Three-dimensional model and docking of the X [/fig_ref] with the homology modeling technique using the SwissModel web server. The S. cerevisiae ERG11 protein [PDB: 4 lxj.1], which shares 48% sequence identity, was used as a template. Additionally, possible protein-ligand binding interactions between the CYP51 protein model and the potential lanosterol and itraconazole ligands were predicted with the protein docking technique using the AutoDock4 program [bib_ref] AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility, Morris [/bib_ref] [fig_ref] Figure 2: Three-dimensional model and docking of the X [/fig_ref]. According to the results, the azole antifungal fits in a similar way to the substrate lanosterol in a hydrophobic groove and interacts with Met-315 and the P-methyl group of Thr-322. Previous studies have also reported these residues as binding sites of azole antifungals in cytochrome P450 proteins [bib_ref] Triazoles in Antifungal Therapy: A Review, Sharma [/bib_ref]. Therefore, the itraconazole azole binds to the CYP51 protein in the substrate-binding pocket of the enzyme, thus blocking its enzymatic catalysis.
## Functional analysis of the x. dendrorhous cyp51 gene
The functionality of the X. dendrorhous CYP51 gene was evaluated by two different approaches: heterologous complementation in S. cerevisiae and gene mutation in X. dendrorhous.
The heterologous complementation analyses were performed using a diploid S. cerevisiae that is heterologous for the ERG11 gene (Sc-erg11 +/− strain), because a null mutant is not viable [bib_ref] Functional characterization of the S. cerevisiae genome by gene deletion and parallel..., Winzeler [/bib_ref]. The S. cerevisiae ERG11 gene and the cDNA of the X. dendrorhous CYP51 gene were ligated into the S. cerevisiae expression vector YEpNP, generating plasmids YEpNP-gERG11 and YEpNP-cCYP51; Sequence alignment between X. dendrorhous CYP51 and S. cerevisiae ERG11 proteins and prediction of structurally conserved motifs in CYP51. Amino acid alignment between the deduced CYP51 sequence from X. dendrorhous (Xd) strain UCD 67-385 and the S. cerevisiae (Sc) strain S288c ERG11 (lanosterol C14-demethylase) protein [Swiss Protein: P10614]. Amino acid differences with the same properties are denoted with a plus (+). Structural elements are highlighted with the name of the corresponding feature above them: possible transmembrane helix (underlined and italics), alpha helices (red and italics), beta-sheets (blue and italics), meander loop and Cys pocket (white highlighted in black). The asterisks (*) indicate the three totally conserved amino acids among cytochromes P450. Secondary structural elements were predicted with the CYP450 Engineering database [bib_ref] The cytochrome P450 engineering database: Integration of biochemical properties, Sirim [/bib_ref] , and the potential transmembrane region was predicted with TMpred [bib_ref] TMbase-A database of membrane spanning protein segments, Hofmann [/bib_ref].
then, these plasmids were used to transform the Sc-erg11 +/− strain. Several transformants were recovered in each case, and at least three transformants of each type were randomly chosen to confirm that they contained the respective plasmid. Given that all of the analyzed transformants showed a positive result in this analysis, one of each (named Sc-gERG11Sc and Sc-cCYP51Xd, respectively) was chosen for sporulation to obtain haploid strains. After 6 days of incubation in sporulation media agar plates, the formation of asci was confirmed by optical microscopy. The asci were broken and seeded onto SD agar plates supplemented with G418 to confirm the presence of the S. cerevisiae ERG11 mutant allele, and with uracil, histidine, lysine and methionine to sustain all possible auxotrophies in the resulting haploid strains. Although the haploid cells were enriched by diethyl ether treatment [bib_ref] Selective killing of vegetative cells in sporulated yeast cultures by exposure to..., Dawes [/bib_ref] , it is also possible to obtain diploid strains among the G418-resistant strains. Because of this, haploid strains were first selected according to their auxotrophy for methionine and lysine, both of which are heterozygous markers in the original diploid parental strain. Subsequently, the haploid condition of the ERG11 mutant allele and its complementation of the expression vector were confirmed by PCR analyses using comprehensive sets of primers [fig_ref] Figure 3: PCR-based analyses of S [/fig_ref]. After these analyses, a single strain that fulfilled the requirements for heterologous expression studies carrying the YEpNP-gERG11 or the YEpNP-cCYP51 vector was chosen; these strains were named Sc-hERG11 and Sc-hCYP51, respectively.
The strains Sc-hERG11 and Sc-hCYP51 did not show significant differences in their growth curves when cultured in YM medium at 22°C with constant agitation (Additional file 1: . Additionally, sterols were extracted after 10, 24 and 56 h of cultivation [fig_ref] Table 1: Sterols obtained from S [/fig_ref]. When the sterol compositions were analyzed by RP-HPLC, all strains showed the same sterol pattern, with one main sterol produced with 13 min of retention time. This sterol was confirmed to correspond to ergosterol after co-injecting each sample with standard ergosterol. In general, strain Sc-hCYP51 showed a lower sterol content than strains Sc-erg11 +/− and Sc-hERG11 at all analyzed time points. However, the fact that strain Sc-hCYP51 was viable and able to produce ergosterol indicates that the CYP51 gene from X. dendrorhous complements the erg11 null mutation in S. cerevisiae. Therefore, this gene encodes a lanosterol C14-demethylase that is able to couple with the endogenous S. cerevisiae cytochrome P450 reductase in vivo.
To evaluate the functionality of CYP51 in X. dendrorhous, cyp51 mutants were created by replacing the CYP51 locus with an antibiotic resistance module through homologous recombination as previously described [bib_ref] Functional characterization of the Xanthophyllomyces dendrorhous farnesyl pyrophosphate synthase and geranylgeranyl pyrophosphate..., Alcaino [/bib_ref]. For this purpose, the wild-type strain CBS 6938 was chosen because it has been shown to be aneuploid, and hemizygous mutants can be obtained after only one transformation event [bib_ref] Cloning of the cytochrome p450 reductase (crtR) gene and its involvement in..., Alcaino [/bib_ref]. Before transformation, the CYP51 gene from strain CBS 6938 was sequenced, uploaded to the Genbank database [Genbank: KP317479] and compared to the sequence from strain UCD 67-385. Minimal differences were found between these two strains, which possessed 99.707% identity in the region from the start to the stop translation codons. After CBS 6938 transformation with linear plasmid pBS-cyp51/hph, a hygromycin B-resistant transformant was obtained. Although it was confirmed that the CYP51 locus was indeed replaced by the resistance module in this strain, the CYP51 gene could still be amplified using a comprehensive set of primers, suggesting that this strain was heterozygous for the CYP51 locus. Therefore, this strain was named CBS-CYP51 +/− . This strain was still a useful model to evaluate the effect of the CYP51 mutation in X. dendrorhous; because the deletion of this gene has been shown to be lethal to the cells of other yeast strains, a cyp51 null mutant would most likely not be viable [bib_ref] Sterol 14alpha-demethylase cytochrome P450 (CYP51), a P450 in all biological kingdoms, Lepesheva [/bib_ref]. Interestingly, the mutant strain had a more intense reddish color that was discernable to the naked eye than the wild type strain, suggesting that it produces more carotenoids than the parental strain. To confirm this observation and to evaluate phenotypic variations between these two strains, growth curves were constructed and sterols and carotenoids were extracted and quantified after 48 and 120 h of cultivation.
The growth curves of both strains did not show significant differences at the lag and exponential phases of growth; however, strain CBS-CYP51 +/− reached a lower OD 600 during the stationary phase of growth (P-value < 0.05, Student's t test) (Additional file 1: . After (B) The AutoDock4 program [bib_ref] AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility, Morris [/bib_ref] was used to perform automated docking to predict the binding sites of lanosterol (red) and itraconazole (blue) in the obtained model of the CYP51 protein from X. dendrorhous.
48 h of cultivation, the wild-type strain accumulated approximately 70% more sterols than the mutant strain; however, after 120 h of culture, there were no statistically significant differences between the strains [fig_ref] Table 2: Sterols [/fig_ref]. This outcome may be the result of a regulation mechanism that modulates sterol biosynthesis according to the metabolic state of the yeast, which is reduced during the stationary phase of growth. Similarly, previous studies in filamentous fungi and in yeast have reported reduced ergosterol accumulation during the stationary phase of growth [bib_ref] Enhancement of carotenoid production by disrupting the C22-sterol desaturase gene (CYP61) in..., Loto [/bib_ref] [bib_ref] Determination of ergosterol as a measure of fungal growth using Si 60..., Zill [/bib_ref]. Moreover, both strains showed the same sterol composition, with ergosterol representing close to 100% of the identified sterols; this result was confirmed by co-injection of the samples with standard ergosterol in RP-HPLC analysis. An additional two unidentified sterols could be detected in both strains with retention times of 12.5 and 15 min; however, these sterols were found in a significantly lower proportion. The mutant strain contained a higher amount of carotenoids (P-value < 0.05, Student's t test), with an approximately 1.9-fold increase after 120 h of culture. Previous studies on carotenogenesis in X. dendrorhous have shown that this process is induced by the end of the exponential and early stationary phases of growth, coinciding with the point at which glucose is completely exhausted [bib_ref] Study of the expression of carotenoid biosynthesis genes in wild-type and deregulated..., Lodato [/bib_ref] [bib_ref] Differential carotenoid production and gene expression in Xanthophyllomyces dendrorhous grown in a..., Wozniak [/bib_ref] [bib_ref] Glucose and ethanol-dependent transcriptional regulation of the astaxanthin biosynthesis pathway in Xanthophyllomyces..., Marcoleta [/bib_ref]. Therefore, it is expected that the major differences in carotenoid content can be better appreciated in older cultures. However, the CBS-CYP51 +/− strain also had a higher carotenoid content after 48 h of culture, suggesting that carotenogenesis starts earlier in this strain. Moreover, carotenoid composition also differed between both strains, with a reduced astaxanthin proportion in the CBS-CYP51 +/− strain [fig_ref] Table 2: Sterols [/fig_ref]. Interestingly, changes in the carotenoid composition showed similarity to cultures of X. dendrorhous in which the crtI gene (encoding the phytoene desaturase) was over-expressed [bib_ref] Metabolic engineering of the carotenoid biosynthetic pathway in the yeast Xanthophyllomyces dendrorhous..., Verdoes [/bib_ref] , resulting in an increased monocyclic carotenoid proportion in CBS-CYP51 +/− cells. Moreover, β-carotene and xanthophyll intermediaries such as phoenicoxanthin and canthaxathin are increased during the synthesis of astaxanthin from β-carotene in CBS-CYP51 +/− , suggesting that the cytochrome P450 system involved in these steps may be weakened. It is likely that this change is the consequence of the reduced CYP51 gene dose in CBS-CYP51 +/− , which could favor the activity of the cytochrome P450 systems involved in ergosterol biosynthesis at the expense of astaxanthin synthesis. This idea is supported by the higher cytochrome P450 reductase transcript levels found in the CBS-CYP51 +/− strain after 24 h of cultivation (see below) that can sustain cytochrome P450 activity during ergosterol biosynthesis.
Similar to our findings, several studies have observed increased carotenoid production in X. dendrorhous when ergosterol levels are reduced. For example, when Phaffia rhodozyma (the anamorphic state of X. dendrorhous) was cultured in the presence of the antifungal ergosterol biosynthesis inhibitor fluconazole, enhanced astaxanthin production was observed [bib_ref] Astaxanthin biosynthesis is enhanced by high carotenogenic gene expression and decrease of..., Miao [/bib_ref]. One possible explanation is that ergosterol down-regulates its own synthesis via a negative-feedback mechanism. Therefore, a reduced ergosterol content favors the availability of mevalonate pathway products, which are common to ergosterol and carotenoid syntheses. This reasoning is supported by a previous study in which X. dendrorhous mutants that are unable to synthesize ergosterol due to the cyp61 mutation had increased carotenoid content compared to their parental wild type strains. Moreover, transcript levels of at least one gene of the mevalonate pathway (HMGR that encodes the hydroxymethylglutaryl CoA-reductase) were significantly increased [bib_ref] Enhancement of carotenoid production by disrupting the C22-sterol desaturase gene (CYP61) in..., Loto [/bib_ref]. In this sense, the supplementation of mevalonate in P. rhodozyma cultures resulted in an increase in carotenoid production [bib_ref] Mevalonic acid increases trans-astaxanthin and carotenoid biosynthesis in Phaffia rhodozyma, Calo [/bib_ref]. Furthermore, deletion of the gene encoding squalene synthase (ERG9, the first step of sterol biosynthesis per se) in combination with the overexpression of the catalytic domain of the hydroxymethylglutaryl CoA-reductase in a recombinant Candida utilis strain caused an increase in lycopene production [bib_ref] Increased carotenoid production by the food yeast Candida utilis through metabolic engineering..., Shimada [/bib_ref]. Based on this reasoning, we evaluated whether HMGR transcript levels were enhanced in strain CBS-CYP51 +/− . To accomplish this, total RNA was extracted from the parental and CBS-CYP51 +/− strains after 24 and 48 h of cultivation in YM media with constant agitation at 22°C, and the transcript levels were analyzed by RT-qPCR. The results revealed that the HMGR transcript level was indeed higher in the mutant strain compared to the wild-type strain, indicating that the elimination of one allele of the CYP51 gene of X. dendrorhous affected the expression of HMGR [fig_ref] Figure 4: RT-qPCR analysis of the HMGR, crtR and CYP51 genes in the wild-type... [/fig_ref]. The HMGR gene controls one of the key regulatory steps in the mevalonate pathway, and, in turn, carotenoid and ergosterol biosynthesis through substrate availability [bib_ref] Physiological implications of sterol biosynthesis in yeast, Parks [/bib_ref] ; therefore, an increase in the HMGR transcript level could partly explain the increased carotenoid content in the mutant strain CBS-CYP51 +/− . Similar to our findings, there was an increase in carotenoid content that was also related to a higher HMGR transcript levels following the treatment of the fungus Blaskelea trispora with ketoconazole, a specific inhibitor of CYP51 [bib_ref] Effects of an ergosterol synthesis inhibitor on gene transcription of terpenoid biosynthesis..., Tang [/bib_ref].
Because cytochrome P450 enzymes are involved in both astaxanthin and ergosterol biosynthesis, we also evaluated the crtR transcript levels [bib_ref] Cloning of the cytochrome p450 reductase (crtR) gene and its involvement in..., Alcaino [/bib_ref]. crtR encodes for a cytochrome P450 reductase, which is the main electron donor in these systems. Interestingly, the crtR transcript level was also higher in CBS-CYP51 +/− after 24 h of cultivation; however, this difference was not statistically significant after 48 h of growth [fig_ref] Figure 4: RT-qPCR analysis of the HMGR, crtR and CYP51 genes in the wild-type... [/fig_ref]. Although several different genes may encode cytochrome P450 enzymes in an organism, usually there is only one gene encoding the enzyme cytochrome P450 reductase. Consequently, a complex mechanism of regulation of the expression of crtR is required to set the levels of cytochrome P450 reductase activity in relation to the various cytochrome P450 proteins in this organism. To date, three cytochrome P450-encoding genes have been described in X. dendrorhous: crtS [bib_ref] The crtS gene of Xanthophyllomyces dendrorhous encodes a novel cytochrome-P450 hydroxylase involved..., Alvarez [/bib_ref] , CYP61 [bib_ref] Enhancement of carotenoid production by disrupting the C22-sterol desaturase gene (CYP61) in..., Loto [/bib_ref] and CYP51 [this current work]). Thus, the cytochrome P450 reductase would act as a cytochrome P450 redox partner at least in the carotenoid and ergosterol synthesis pathways. Finally, we also analyzed whether the elimination of one CYP51 allele indeed affected its own transcript levels. After 24 h of cultivation, no significant differences were observed between the mutant and the parental strain; however, after 48 h of cultivation, differences could be appreciated [fig_ref] Figure 4: RT-qPCR analysis of the HMGR, crtR and CYP51 genes in the wild-type... [/fig_ref] , confirming that the X. dendrorhous mutant generated in this work indeed affected CYP51 transcript levels.
# Conclusions
In this work, the CYP51 gene from X. dendrorhous was identified and its function was characterized. This gene encodes a putative protein that contains all of the cytochrome P450 conserved motifs and is consistent with CYP51 protein family characteristics. The heterologous complementation analysis in S. cerevisiae strongly suggests that this gene in fact encodes a functional sterol C14demethylase. Moreover, a CYP51 mutation in X. dendrorhous affected sterol and carotenoid production, supporting the finding that reduced sterol production is related to higher carotenoid production.
# Methods
## Biological material and microorganism culture conditions
Restriction enzymes, the Klenow polymerase and the M-MLV reverse transcriptase were purchased from Promega, and the Pfu DNA polymerase was purchased from Invitrogen. The plasmids and microbial strains that were used and constructed for this work are included in [fig_ref] Table 3: Plasmids and strains used and constructed in this study [/fig_ref]. The wild type UCD 67-385 (ATCC 24230) X. dendrorhous strain was used for genome and transcriptome sequencing and for genomic and cDNA CYP51 gene amplification and isolation. For the CYP51 gene mutation, the wild type CBS 6938 (ATCC 96594) X. dendrorhous strain was used. Escherichia coli DH-5α was used as a host for plasmid propagation. Saccharomyces cerevisiae Meyen ex E.C Hansen YHR007C BY4743 (ATCC 26604) was used for CYP51 heterologous complementation analysis because it is a diploid and heterozygous mutant for the ERG11 gene (erg11 (+/−) ), with one copy of ERG11 replaced by a geneticin resistance cassette. Most plasmid constructs derived from pBluescript SK-. Plasmid YEpNP was modified from YEpACT4 [bib_ref] Heterologous Expression of a Candida molischiana Anthocyanin-beta-glucosidase in a Wine Yeast Strain, Sanchez-Torres [/bib_ref] and used for S. cerevisiae heterologous complementation. X. dendrorhous was cultured in YM rich medium (1% glucose, 0.3% yeast extract, 0.3% malt extract and 0.5% peptone) with constant agitation at 22°C. For transformant selection, the yeast was grown on 1.5% agar YM
## This work
YEp-Act4 pBR322 and 2 micron ori, Amp R , LEU2 and containing the ACT promoter to regulate gene expression.
[48]
YEpNP YEp-Act4 bearing the S. cerevisiae TDH3 terminator. This work YEpNP-cCYP51 YEpNP bearing the X. dendrorhous CYP51 cDNA (1,653 bp) under the regulation of the ACT promoter and the TDH3 terminator.
## This work
YEpNP-gERG11 YEpNP bearing the S. cerevisiae ERG11 gene (1,593 bp) under the regulation of the ACT promoter and the TDH3 terminator.
## This work
Strains:
## E. coli:
DH-5α F-φ80d lacZΔM15Δ (lacZY-argF) U169 deoR recA1 endA1 hsdR17(rk-mk+) phoA supE44 l-thi-1 gyrA96 relA1S. cerevisiae:
## S288c
MATα, SUC2, gal2, mal, mel, flo1, flo8-1, hap1, ho, bio1, bio6. [bib_ref] Submit your next manuscript to BioMed Central and take full advantage of:..., Mortimer [/bib_ref] Meyen ex E.C Hansen YHR007C BY4743 (Sc-erg11 +/− ) This work plates supplemented with 10 μg/ml hygromycin B (US Biological). E. coli was cultured at 37°C with constant agitation in Luria-Bertani (LB, 1% tryptone, 0.5% yeast extract, and 0.5% NaCl) medium and on 1.5% agar LB plates supplemented with 100 μg/ml ampicillin for plasmid selection and 20 μg/ml of X-gal (5-bromo-4chloro-3-indolyl-β-Dgalactopyranoside) for recombinant clone selection by blue-white screening. For selection, recombinant clones bearing the constructed plasmids were analyzed by direct colony PCR with an adequate set of primers.
[formula] MATa/MATα, [/formula]
S. cerevisiae strains were cultured at 30°C in YM or YEP (YM: 0.3% yeast extract, 0.3% malt extract and 0.5% peptone; YEP: 2% glucose, 1% yeast extract, and 2% peptone) rich media. The strain Meyen ex E.C Hansen YHR007C BY4743 was cultured at 22°C in SD medium (0.67% yeast nitrogen base without amino acids and 2% glucose) supplemented with metabolites to sustain the strain auxotrophy (0.002% uracil, 1% histidine and 1% leucine) and 200 μg/ml of geneticin. Transformants derived from this strain were cultured in SD medium supplemented with uracil, histidine and geneticin, because the vector YEpNP complements the parental strain's leucine auxotrophy.
Nucleic acid extraction, DNA amplification and sequence analysis X. dendrorhous DNA extraction was performed from protoplasts according to [bib_ref] Genetics and electrophoretic karyotyping of wild-type and astaxanthin mutant strains of Phaffia..., Cifuentes [/bib_ref] , and total RNA extraction was performed according to a modified protocol of Chomczynski and Sacchi [bib_ref] Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction, Chomczynski [/bib_ref] for X. dendrorhous [bib_ref] Study of the expression of carotenoid biosynthesis genes in wild-type and deregulated..., Lodato [/bib_ref]. Total RNA was quantified spectrophotometrically at 260 nm according toin a V-630 UV-vis Spectrophotometer (JASCO). S. cerevisiae genomic DNA was obtained by mechanical cell disruption using 0.5 mm glass beads (BioSpec) and shaking in a mini bead beater-16 (BioSpec). Plasmid DNA was obtained from E. coli strains using the AxyPrep Plasmid Miniprep kit (Axygene).
Oligonucleotides designed and used in this work were synthesized by Integrated DNA Technologies and are listed in Additional file 2: [fig_ref] Table 1: Sterols obtained from S [/fig_ref]. PCR reactions were performed in a final volume of 25 μl with 2 U of Taq DNA polymerase, 2.5 μl of 10X Taq buffer, 0.5 μl of 10 mM dNTPs, 1 μl of 50 mM MgCl 2 , 1 μl of each primer (25 μM) and 10 to 20 ng of template DNA in an Applied Biosystems 2720 thermal cycler. The general amplification protocol was as follows: initial denaturation at 95°C for 3 min; 35 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and synthesis at 72°C for 3 min; and a final extension step at 72°C for 10 min. Samples were stored at 4°C until analyzed by 0.8% agarose gel electrophoresis in TAE buffer stained with 0.5 μg/ml ethidium bromide. DNA for sequencing or plasmid construction was purified from gels by the glass milk method [bib_ref] An inexpensive alternative to glassmilk for DNA purification, Boyle [/bib_ref].
DNA sequencing was performed in an ABI 3100 Avant genetic analyzer using the BigDye terminator v3.1 kit (Applied Biosystems) and analyzed with Vector NTI Suite 10 (Informax), CLUSTAL W 1.8 and programs available at the NCBI web site. Protein sequence analyses were performed with programs available at www.ch.embnet.org/software/TMPRED_form.html [bib_ref] TMbase-A database of membrane spanning protein segments, Hofmann [/bib_ref] , www.ebi.ac.uk/InterProScan/ [bib_ref] InterProScan-an integration platform for the signature-recognition methods in InterPro, Zdobnov [/bib_ref] and www.cyped.unistuttgart.de/ [bib_ref] The Cytochrome P450 Engineering Database: a navigation and prediction tool for the..., Fischer [/bib_ref].
Protein modeling was performed using programs available at www.swissmodel.expasy.org/ [bib_ref] SWISS-MODEL: modelling protein tertiary and quaternary structure using evolutionary information, Biasini [/bib_ref] , http:// autodock.scripps.edu/wiki/AutoDock4 [bib_ref] AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility, Morris [/bib_ref] and www. ks.uiuc.edu/Research/vmd/ [bib_ref] VMD: Visual Molecular Dynamics, Humphrey [/bib_ref].
## Isolation of the x. dendrorhous cyp51 gene and plasmid construction
To identify the CYP51 gene from X. dendrorhous, BLAST analyses using homologous CYP51 nucleotide and amino acid sequences were performed with the yeast genomic and transcriptomic databases available in our laboratory [bib_ref] Enhancement of carotenoid production by disrupting the C22-sterol desaturase gene (CYP61) in..., Loto [/bib_ref]. Specific primers were designed from the identified sequences (Additional file 2: [fig_ref] Table 1: Sterols obtained from S [/fig_ref].
For the S. cerevisiae heterologous complementation experiments, the CYP51 cDNA from X. dendrorhous and the ERG11 gene from S. cerevisiae (homologous to CYP51) were ligated into the yeast expression plasmid YEpNP [fig_ref] Table 3: Plasmids and strains used and constructed in this study [/fig_ref]. First, plasmid pBS-cCyp51, which contained the cDNA of the CYP51 gene, was constructed by inserting a 1,653 bp RT-PCR product into the EcoRV site of plasmid pBluescript SK-, which was obtained using the primer set cCYP51.F + cCYP51.R (Additional file 2: [fig_ref] Table 1: Sterols obtained from S [/fig_ref]. Plasmid pBS-cyp51/Hyg was used to eliminate an allele of the CYP51 gene in X. dendrorhous. This plasmid was obtained by sequentially joining three DNA fragments: i) 700 bp (amplified with primers CYP51_del.Fw + CYP51_del-HpaI.Rv) upstream of the CYP51 gene encoding sequence; ii) the hygromycin B resistance module [bib_ref] Genomic organization of the structural genes controlling the astaxanthin biosynthesis pathway of..., Niklitschek [/bib_ref] (amplified with primers H.F and H.R); and iii) 600 bp (amplified with primers CYP51_del-HpaI.Fw + CYP51_del.Rv) downstream of the CYP51 gene encoding sequence. These sequences were inserted into the EcoRV site of plasmid pBluescript SK-. For X. dendrorhous transformation, the insert of plasmid pBS-cyp51/Hyg was released using the NotI and XhoI endonucleases.
## Yeast transformation
S. cerevisiae electrocompetent cells were prepared from 60 ml of exponentially growing cultures in YEP medium at 22°C with constant agitation. The cells were washed three times with chilled sterile distilled water, one time with sorbitol 1 M and finally suspended in 0.2 ml of 1 M sorbitol. The cells were kept at 4°C, and 40 μl aliquots were used for electroporation with a BioRad gene pulser × cell with PC and CE modules under the following conditions: 1.5 kV, 25 μF and 200 Ω. X. dendrorhous transformation was also performed by electroporation according to [bib_ref] Transformation of the astaxanthin-producing yeast Phaffia rhodozyma, Adrio [/bib_ref] and [bib_ref] Cloning of the ribosomal protein L41 gene of Phaffia rhodozyma and its..., Kim [/bib_ref] using 1 to 5 μg of linear donor DNA. Electrocompetent cells were prepared from exponential cultures, grown in YM medium and electroporated using the following program: 125 mF, 600 Ω, 0.45 kV.
## S. cerevisiae heterologous complementation assay
The erg11 (+/−) S. cerevisiae strain, which is diploid and heterozygous for the ERG11 gene, was used for the gene complementation assays. The S. cerevisiae ERG11 gene and the X. dendrorhous CYP51 gene were ligated into plasmid YEpNP and used to independently transform the original diploid yeast strain. Transformants were selected in SD agar plates supplemented with 200 μg/ml geneticin, 0.002% uracil and 1% histidine according to the parental strain requirements. Transformant strains were randomly chosen and incubated in pre-sporulation medium plates (0.8% yeast extract, 0.3% peptone, 10% glucose and 2% agar), incubated at 22°C for 1 to 2 days and then transferred to minimal sporulation medium plates (1% potassium acetate and 2% agar) for incubation at 22°C for 3 to 6 days. Supplements according to the parental strain genotype were added to the sporulation media. Asci formation was confirmed by optical microscopy. To enrich the asci fraction, cells were collected from the sporulation plates, treated with diethyl ether for 20 min at 22°C [bib_ref] Selective killing of vegetative cells in sporulated yeast cultures by exposure to..., Dawes [/bib_ref] and the ascospores were released by treatment with zimoliase-120 T (US Biological) at 30°C for 2 h. Afterwards, the mixture was shaken in a mini bead beater-16 (BioSpec) with 0.5 mm glass beads (BioSpec) for 2 min [bib_ref] Getting started with yeast, Sherman [/bib_ref]. Finally, the cells were spread on SD agar plates supplemented with histidine, uracil, lysine, methionine and geneticin to support all possible combinations of auxotrophies in the resulting haploid cells.
To select the desired haploid cells to study gene complementation, the obtained colonies were streaked in replicates on SD agar plates supplemented with geneticin and: i) uracil, histidine and lysine, ii) uracil, histidine and methionine and iii) uracil, histidine, methionine and lysine. Colonies that developed on the plates supplemented with the four metabolites but not when lysine or methionine were omitted were chosen for further analyses.
## Sterol and carotenoid extraction and analysis
Sterols were extracted from both the S. cerevisiae and X. dendrorhous strains, while carotenoids were extracted only from X. dendrorhous. Metabolites were spectrophotometrically quantified and normalized to the dry weight of the yeast.
Sterols were extracted according to the protocol of [bib_ref] Effect of nitrogen limitation on the ergosterol production by fed-batch culture of..., Shang [/bib_ref] by mixing cell pellets with 4 g of KOH and 16 ml of 60% (v/v) ethanol/water and incubating at 80 ± 2°C for 2 h. Then, non-saponificable sterols were extracted with 10 ml of petroleum ether and quantified spectrophotometrically at 280 nm. Carotenoids were extracted from cellular pellets by the acetone method [bib_ref] Isolation of Phaffia rhodozyma mutants with increased astaxanthin content, An [/bib_ref] and quantified spectrophotometrically at 465 nm. Sterol and carotenoid compositions were analyzed by RP-HPLC with an RP-18 Lichrocart125-4 (Merck) column using methanol: water (97: 3, v/v) or acetonitrile: methanol: isopropanol (85: 10: 5, v/v) as the mobile phase with a 1 ml/min flux under isocratic conditions to separate the sterols and carotenoids, respectively. The elution spectra were recovered using a diode array detector, and the metabolites were identified according to their spectra and retention time in comparison to standards.
## Single-strand dna synthesis and rt-qpcr
Single-stranded DNA was synthesized according to the M-MLV reverse transcriptase (Invitrogen) manufacturer's instructions using 5 μg of total RNA in a final volume of 20 μl. Relative gene expression levels were obtained in an Mx3000P quantitative PCR system (Stratagene) using 1 μl of the reverse transcription reaction, 0.25 μM of each primer and 10 μl of the SensiMix SYBR Green I (Quantace) kit in a final volume of 20 μl. The Ct values obtained were normalized to the corresponding value for the beta-actin encoding gene [Genbank: X89898.1] [bib_ref] Expression of the carotenoid biosynthesis genes in Xanthophyllomyces dendrorhous, Lodato [/bib_ref] , and later expressed as a function of the control conditions using the ΔΔCt algorithm [bib_ref] Analysis of relative gene expression data using real-time quantitative PCR and the..., Livak [/bib_ref].
[fig] Figure 2: Three-dimensional model and docking of the X. dendrorhous CYP51 deduced protein. (A) The model was generated by the SwissModel web server [54] using the S. cerevisiae ERG11 protein with lanosterol as the template ligand (model code. 4 lxj.1). [/fig]
[fig] Figure 3: PCR-based analyses of S. cerevisiae erg11 − strains carrying the YEpNP-gERG11 and YEpNP-cCYP51 vectors. PCR analyses to confirm the presence of the kanMX gene at the ERG11 locus (panel A: primers CYP51ScExt.F and KanMX4.R2 and panel B: primers KanMX4.F2 and CYP51ScExt.R), the presence of the X. dendrorhous CYP51 gene (panel C: primers RTCYP51.F and CYP51.Rb), the presence of the ERG11 gene (panel D: gERG11.F and gERG11.R) and the absence of the ERG11 gene in the S. cerevisiae genome (panel E: gERG11.F and CYP51ScExt.R). Template DNA in each lane: S. cerevisiae parental diploid strain erg11 +/− (Lane 1), S. cerevisiae S288c strain (Lane 2), S. cerevisiae Sc-hCYP51 strain (Lane 3), S. cerevisiae Sc-hERG11 strain (Lane 4), X. dendrorhous UCD 67-385 strain (Lane 5), negative control without DNA (Lane 6). Molecular size marker Lambda/Hind III (Lane M: 23.1, 9.4, 6.6, 4.4, 2.3, 2.0 and 0.6 kb). A schematic diagram is included to represent the primer sets (shown in arrows) that were used. The UP and DOWN regions correspond to regions located 300 bp upstream and downstream of the S. cerevisiae ERG11 gene, respectively. Region KanMX4 corresponds to the geneticin (G418) resistance module and the pACT4 and tTDH3 regions correspond to the S. cerevisiae promoter and terminator region in YEpNP. [/fig]
[fig] Figure 4: RT-qPCR analysis of the HMGR, crtR and CYP51 genes in the wild-type and CBS-CYP51 +/− strains. The HMGR (A), crtR (B) and CYP51 (C) transcript levels were determined by RT-qPCR after 24 and 48 h of cultivation of the wild-type and the CBS-CYP51 +/− strains. Each transcript level was normalized with respect to the transcript level of the actin gene and then with respect to the wild-type strain: CBS 6938 (=1, black bars) and CBS-CYP51 +/− (gray bars). Values are the mean ± standard error of three independent experiments (* p ≤ 0.05; Student's t test). [/fig]
[table] 87: and 90 bp. The CYP51 gene was predicted to encode a 550 amino acid protein with an expected molecular weight of 61.8 kDa and pI of 6.64. This protein shares 48% identity and 65% similarity at 91% coverage with the S. cerevisiae C14-demethylase [ERG11, Swiss-Prot: P10614] [/table]
[table] Table 1: Sterols obtained from S. cerevisiae strains used in this study (mg per g of dry yeast) ND: Not detected.Table valuescorrespond to the average result from three independent cultures ± standard deviations. The RP-HPLC retention time is indicated in parenthesis. [/table]
[table] Table 2: Sterols (mg per g of dry yeast) and carotenoids (μg per g of dry yeast) obtained from the X. dendrorhous used in this study ND: Not detected. Table values correspond to the average result from three independent cultures ± standard deviations. Percentage relative to total carotenoids is indicated in parenthesis. Monocyclic carotenoids include: γ-carotene, keto-γ-carotene, hydroxy-keto-γ-carotene, torulene and hydroxy-keto-torulene. [/table]
[table] Table 3: Plasmids and strains used and constructed in this study [/table]
|
RNA silencing can explain chlorotic infection patterns on plant leaves
Background: RNA silencing has been implicated in virus symptom development in plants. One common infection symptom in plants is the formation of chlorotic tissue in leaves. Chlorotic and healthy tissue co-occur on a single leaf and form patterns. It has been shown that virus levels in chlorotic tissue are high, while they are low in healthy tissue. Additionally, the presence of siRNAs is confined to the chlorotic spots and the boundaries between healthy and infected tissue. These results strongly indicate that the interaction between virus growth and RNA silencing plays a role in the formation of infection patterns on leaves. However, how RNA silencing leads to the intricate patterns is not known.Results:Here we elucidate the mechanisms leading to infection patterns and the conditions which lead to the various patterns observed. We present a modeling approach in which we combine intraand inter-cellular dynamics of RNA silencing and viral growth. We observe that, due to the spread of viruses and the RNA silencing response, parts of the tissue become infected while other parts remain healthy. As is observed in experiments high virus levels coincide with high levels of siRNAs, and siRNAs are also present in the boundaries between infected and healthy tissue. We study how single-and double-stranded cleavage by Dicer and amplification by RNA-dependent RNA polymerase can affect the patterns formed.Conclusion: This work shows that RNA silencing and virus growth within a cell, and the local spread of virions and siRNAs between cells can explain the heterogeneous spread of virus in leaf tissue, and therewith the observed infection patterns in plants.
# Background
RNA silencing is an evolutionary conserved mechanism in eukaryotes that has a major role in gene regulation, development, transposon control and defense against viruses. Antiviral RNA silencing is induced by virus doublestranded RNA (dsRNA) or by specific single-stranded RNA (ssRNA) structures. Double-or single-stranded RNA is cleaved into small interfering RNA (siRNA) by RNase IIIlike enzymes such as Dicer and Dicer-like [bib_ref] Induction and suppression of RNA silencing: insights from viral infections, Voinnet [/bib_ref] [bib_ref] Plant virus-derived small interfering RNAs originate predominantly from highly structured single-stranded viral..., Molnar [/bib_ref]. siRNAs associate with the RNA-induced silencing complex (RISC) and guide the complex to complementary sequences that are then destroyed. In addition to the primary response, siRNAs can be produced through a secondary pathway that involves synthesis of dsRNA or siRNA by host encoded RNA-dependent RNA polymerase (RDR) [bib_ref] Nomenclature and functions of RNA-directed RNA polymerases, Wassenegger [/bib_ref] [bib_ref] Cellular RNA-dependent RNA polymerase involved in post-transcriptional gene silencing has two distinct..., Makeyev [/bib_ref] [bib_ref] Amplified silencing, Baulcombe [/bib_ref].
The antiviral role of RNA silencing is extensively studied in plants [bib_ref] Induction and suppression of RNA silencing: insights from viral infections, Voinnet [/bib_ref] [bib_ref] Antiviral immunity directed by small RNAs, Ding [/bib_ref] [bib_ref] A call to arms: coevolution of animal viruses and host innate immune..., Marques [/bib_ref]. Virus spread through the plant results in diverse symptoms, for example leaf curling, abnormal leaf or flower development, and patterns on infected leaves. These patterns consist of both chlorotic or necrotic tissue in combination with healthy looking tissue. Different types of patterns that occur are concentric circles or rings, mosaic patterns, vein clearing and spots. Interestingly, virus levels are high in yellow, chlorotic tissue and low in the green, healthy tissue [bib_ref] On the origin of dark green tissue in tobacco leaves infected with..., Atkinson [/bib_ref]. This means that virus accumulation varies from cell to cell. It has been hypothesized that RNA silencing may play a role in the development of leaf patterns resulting from virus infections [bib_ref] Transcriptional and posttranscriptional plant gene silencing in response to a pathogen, Al-Kaff [/bib_ref] [bib_ref] Dark green islands in plant virus infection are the result of posttranscriptional..., Moore [/bib_ref] [bib_ref] An RNA-based information superhighway in plants, Jorgensen [/bib_ref]. Recent observations by Hirai et al. [bib_ref] Antiviral RNA silencing is restricted to the marginal region of the dark..., Hirai [/bib_ref] on mosaic patterns support this hypothesis. They have shown that RNA silencing activity is confined to the yellow spots and the marginal regions of the green spots. Reduced expression of RDR, which is part of the secondary pathway of the silencing response, resulted in smaller or no green tissue. These results strongly suggest that RNA silencing plays a major role in the development of plant symptoms.
Previously we developed a mathematical model of RNA silencing and its interaction with viral growth within a cell [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref]. We found that depending on the strength of the silencing response the virus equilibrium can be almost unaffected, oscillations can occur, or the virus can be cleared. Additionally, we found that a change in Dicer cleavage rate is representative for a general change in silencing strength. For a low Dicer cleavage rate the equilibrium amount of virus is slightly decreased and the virus grows slower than without silencing. For high Dicer cleavage rate the virus is not able to grow and is cleared directly after introduction. For intermediate Dicer cleavage rates oscillations in virus levels can occur. When a secondary response is added these oscillations can be enlarged to such extend that the virus is cleared after a single growth peak.
We here study how RNA silencing can explain the development of leaf patterns resulting from viral infection. To this end we use a detailed modeling approach in which we combine an intra-cellular model of viral growth and the RNA silencing pathway with inter-cellular tissue dynamics.
We observe that RNA silencing and virus growth on a tissue can result in a discontinuous spread of the virus: the virus reaches high levels in some cells, while it is suppressed in other cells. We study the conditions for different type of patterns. These patterns could be the basis of plant symptom development. We elucidate the mechanisms leading to these patterns and how increased silencing efficiency, siRNA movement and the occurrence of a secondary response relate to the pattern formation.
# Methods
To investigate the formation of infection patterns in plants we model an area of plant tissue on a grid. Each grid point represents a plant cell. Within each cell we calculate virus levels and levels of RNA silencing proteins with a detailed model. Virions and siRNAs can move from cell-to-cell. A schematic representation of the model is shown in .
## Intracellular model
The number of molecules in each cell is calculated with our previously described model of the antiviral RNA silencing pathway and a replicating plus-strand RNA virus [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref].
The intracellular model consists of coupled differential equations, representing viral plus-and minus-strand RNA, dsRNA, virions, RdRP, siRNA targeting plus or minus-stranded RNA, and free and active RISC. The virus replication cycle starts with the translation of plusstranded RNA into a poly-protein. After auto-cleavage one of the products is RNA-dependent RNA polymerase (RdRP) that associates with plus-strand RNA to synthesize a complementary strand. The formed dsRNA separates into a plus-and minus-strand that can both associate with RdRP again. We assume that the minus-strand is the preferred template for dsRNA synthesis. Semi-conservative synthesis of multiple plus-strands from a single minusstrand template is incorporated in the model, resulting in a biased plus-to-minus ratio. The virus produces virions that consist of plus-stranded RNA and coat proteins. We simplify here by using the number of plus-stranded RNA instead of modeling a separate coat protein.
Viral double-or single-stranded RNA is degraded by host encoded Dicer into siRNAs, that have a plus-or minusstrand polarity. siRNAs cleaved from dsRNA have a 50% chance of targeting either the plus or the minus strand. siRNAs cleaved from a ssRNA hairpin automatically target strands with the opposite polarity. Via RISC the siRNAs cause degradation of either plus-or minus-stranded viral RNA.
Secondary siRNAs can be synthesized through the amplification pathway that involves synthesis of dsRNA by host encoded RDR. We implement unprimed, primed and guided amplification. Each type of amplification can be studied separately.
All equations are integrated using a timestep of 3.6 seconds. Simulations run for 300 hours, unless indicated otherwise. The equations can be found in the Appendix.
## Tissue level model
To study RNA silencing and viral infection in a tissue we use a spatial model. Each grid point represents a cell for which the intracellular dynamics are calculated with the model described above.
Viruses encode movement proteins that enable the movement of virions from cell-to-cell through plasmodesmata. We implement the movement of virions to the four neighboring cells in our model. Virions can be unpacked in each cell into naked plus-strand RNA. We chose a 4 neighborhood because cells share almost no surface area with the diagonal neighbors.
For movement we shift to a particle based system, because we do not want incomplete particles to trigger a reaction in a neighboring cell. Movement of particles occurs every timestep (3.6 seconds). A fraction of the total number of virions in the cell is evenly distributed among the four neighbors, and excess virions are distributed randomly among the neighbors. When the number of moving virions is smaller than one, we draw a random number to decide if one virion moves to a random neighbor. With this method we underestimate the heterogeneity of viral spread as compared to Brownian motion. This method is therefore a good worst-case scenario for the study of heterogeneous spread of virus particles.
The silencing response is able to spread from cell-to-cell with a short range silencing signal, most likely siRNAs [bib_ref] Transitivity-dependent and -independent cell-to-cell movement of RNA silencing, Himber [/bib_ref]. We implement the spread of siRNAs in the same way virions move. There is also a long-range silencing signal [bib_ref] Carroll BJ: Nuclear gene silencing directs reception of long-distance mRNA silencing in..., Brosnan [/bib_ref] [bib_ref] Mixing and matching: the essence of plant systemic silencing?, Dunoyer [/bib_ref]. Because of the elusive nature of the long range silencing signal, and because we here take only a tissue or leaf area into account we do not include a long range silencing signal.
# Results and discussion
We vary Dicer cleavage rates as representative for silencing strength to study the effect of RNA silencing on the spread of virus particles over the tissue.
Within the cell three different behaviors can be observed [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref]. High Dicer cleavage rate results in fast clearance of the virus. Low Dicer cleavage rate delays viral growth but hardly decreases the virus levels in equilibrium. Intermediate cleavage rate results in oscillating virus levels.
## Infection patterns without rna silencing
We first study virus spread without silencing present. We initialize the tissue with healthy cells and infect one cell in the center with 10 viral plus strands. After initiation the Schematic representation of the model Schematic representation of the model. In each grid point the dynamics of a replicating plus-strand RNA virus and antiviral silencing are calculated. Viral plus-and minus-strand RNA is replicated by RdRP. The formed complex dissociates into ssRNA. Virus single-strand and part of the double-stranded RNA can be cleaved into siRNA by Dicer. siRNA associates with RISC and cleaves the target RNA. siRNAs can guide or prime amplification of the response through host encoded RDR, or viral ssRNA is amplified by RDR in an unprimed manner. The full intracellular model can be found in the Appendix. siRNAs and virions produced in each cell can move from cell to cell on the grid.
virus starts to produce virions that spread from cell-to-cell. We fix the fraction of virions leaving a cell to 1% per hour.
The virus spreads rapidly over the entire area and 75 hours post infection (hpi) a circular area of the grid has become infected with the virus. In [fig_ref] Figure 2: Virus spread over plant tissue [/fig_ref] we show the number of virions, the number of siRNAs and the number of plus strands in separate screen-shots. When virions, virus plusstranded RNA or siRNAs are absent from the cell, it is shown in green. Cells that have virions, plus-stranded RNA or siRNAs are shown in a color ranging from black to yellow (via red): the color ramps are shown in [fig_ref] Figure 2: Virus spread over plant tissue [/fig_ref]. Also shown is a space-time plot, that is, horizontal crosssections of the grid every hour post infection (1 hpi top row, 75 hpi bottom row). In [fig_ref] Figure 2: Virus spread over plant tissue [/fig_ref] we show timeseries of plus-strand accumulation in adjacent cells. The curve starting at 0 hpi is an initially infected cell. The other cells in turn each become infected by the neighboring cell(s) and in each cell the virus expands to the equilibrium.
## Infection patterns with silencing, without sirna movement
When RNA silencing is active, we initialize the cells with RISC and Dicer. We first assume that Dicer is only capable of cleaving siRNA from viral dsRNA, and that siRNAs cannot move from cell to cell. Even without siRNA movement RNA silencing slows down virus spread over the area [fig_ref] Figure 2: Virus spread over plant tissue [/fig_ref]. In the space-timeplot can be seen that the infec-tion advances much slower than without the presence of the silencing response. A stronger silencing response slows down the spread of the virus more than a weaker response. We here show results for an intermediate Dicer cleavage rate, that results in oscillatory behavior within the cells [fig_ref] Figure 2: Virus spread over plant tissue [/fig_ref]. For high Dicer cleavage rate the virus is cleared immediately after introduction, and is not able to spread from cell to cell, because it is eliminated before virions could be produced.
## Infection patterns with sirna movement
As shown above, without siRNA movement the infection spreads homogeneously over the area. When siRNAs do move from cell to cell, they can limit virus growth in neighboring cells, resulting in viral growth in some cells and suppression in others. This results in patterns that can spread over the entire area or stay localized to the area around the inoculated site. We observe somewhat different patterns for low and intermediate Dicer cleavage rates and we will discuss results from both possibilities.
## Low dsrna cleavage rate by dicer
We first study the effects of a low Dicer cleavage rate for all possible siRNA movement rates. Varying the rate of siRNA movement we observe that the virus does not spread uniformly over the area and that patterns are formed.
Low siRNA movement results in a circular pattern that is shown in . The virus reaches a high equilibrium Virus spread over plant tissue in the cells close to the initiation site. The siRNAs produced by these cells inhibit viral growth in a ring of cells around the center. Virions are unpacked in these cells, but the viral RNA is silenced with the siRNAs from the neighbors. At sufficient distance from the siRNA producing cells, the virus will be able to grow, and these cells will in their turn inhibit viral growth in the next ring. Once a ring is formed it remains stable.
In the time-series it can be seen that the virus can expand to an equilibrium with high RNA levels in some cells, while in other cells the virus decreases to low RNA levels. We will refer to these cells as "silenced cells". The silenced equilibrium is maintained by siRNA movement: the virus grows to the high equilibrium when siRNA movement is stopped. In silenced cells the siRNA influx suppresses virus replication completely: no dsRNA is formed, and the observed plus-strand RNA levels result from the virion influx from neighboring cells.
For a higher siRNA movement rate a different pattern occurs . The circular pattern breaks and protrusive waves occur that resemble mosaic-like symptoms. We can distinguish three cell types in this pattern: cells with high virus levels, silenced cells and healthy cells. The healthy spots between the infected patches are surrounded by the silenced cells. Apart from low virus levels in the silenced cells at the edge of the spots there are no virions or other virus particles present in the healthy spots. The siRNAs present in the edge cells seem to "guard" the edges of the healthy spots as described by Hirai et al. [bib_ref] Antiviral RNA silencing is restricted to the marginal region of the dark..., Hirai [/bib_ref]. These siRNAs, however, are not produced by the silenced cells in the edges, but move there from the infected neighboring cells.
For the maximum siRNA movement rate (10% of the available siRNAs in a cell) the infection is confined to the area around the inoculated site .
Concluding, siRNA movement creates silenced cells in which the virus is suppressed, and cells in which the virus grows to high values. Increasing siRNA movement increases the number of silenced cells, rather than decreasing virus load in all cells.
## Intermediate dsrna cleavage rate by dicer
When Dicer cleavage rate is intermediate, similar patterns can be observed . However, a lower siRNA movement rate is sufficient to generate them. For low siRNA movement rate a speckle pattern occurs. Initially the virus is able to expand considerably in all cells . However, some time after infection virus levels drop in some cells and reach the silenced equilibrium. After 300 hours the pattern is almost completely stable, either cells are at the high or at the low equilibrium. This speckle pattern distinguishes itself from the others by the high initial growth of the virus in all cells: in other patterns virus levels never reach this high levels before declining to the silenced state.
When siRNA movement increases concentric circles are formed as is the case for the lower silencing efficiency . The number of silenced cells increases compared to the speckle pattern. A relatively low siRNA movement rate results in broad bands of virus infection, a higher rate results in very thin bands.
With still higher siRNA movement the thin bands in which the virus reaches high levels break, and a growing ice crystal-like pattern is observed . The pattern consists of protrusive waves, and in only a small number of cells the virus reaches the high equilibrium. Because the infected spots are much smaller compared to the similar mosaic pattern for low silencing strength there are not enough siRNAs to create truly healthy spots.
With even higher siRNA movement rate the protrusive waves are reduced to a local spot, with the possibility of one or two small outbreaks . For low silencing efficiency we also observed a spot pattern . Low silencing efficiency results in a completely infected spot, while higher silencing efficiency results in a small ring-like pattern.
## Alternative equilibria depend on influx of sirnas
As shown in the previous section, virus levels in the spatial model can reach two different equilibria, a high and a low one, while in the intracellular ODE model only a single equilibrium exists. The patterns disappear when siRNA movement is stopped, therefore siRNA movement maintains the silenced equilibrium. To analyze the influence of the tissue dynamics on the cellular dynamics mathematically we add in-and efflux of siRNAs and virions to the intracellular model. Efflux is fixed to the movement parameters used in the tissue model. We measure average virion and siRNA influx in the equilibrium for 4 cells from the center of the tissue and use these as influx values in the intracellular model.
As an example case we take the parameters and measurements for the mosaic-like pattern for low and intermediate silencing strength shown in [fig_ref] Figure 2: Virus spread over plant tissue [/fig_ref].
equilibria occur, a high and a low one, corresponding to the equilibria in the spatial model . A low siRNA influx results in virus growth to the high equilibrium, high siRNA influx results in growth to the silenced equilibrium.
We calculate bifurcation diagrams as a function of the siRNA influx. In we show the bifurcation diagram for low silencing strength. Each line in the bifurcation diagram is calculated with measurements from different patterns. For all measurements only the high equilibrium exists for low siRNA influx, while for higher siRNA influx there are three equilibria: two stable states separated by an unstable state . Depending on the initial conditions and the siRNA influx the virus can either grow to the high or the low equilibrium. The number of virus RNAs in the high and low equilibrium are similar for the bifurcation lines calculated for different siRNA movement rates. However, the siRNA influx needed to reach the low equilibrium (that is, the bifurcation point) shifts to lower siRNA influx values for lower siRNA movement rate. Low siRNA movement rate increases the effect of silencing within the cell, because less siRNAs leave it and a lower siRNA influx is sufficient to reach the silenced state.
In the spatial model both virion and siRNA influx change over time. Therefore not the final siRNA influx but the combined virion and siRNA influx during the growth phase of the virus determine which equilibrium is reached. Once a stable state is reached a change in siRNA influx has no effect on equilibrium values. Only when siRNA influx is dramatically lowered or when a very large amount of virions would be introduced it would be possible to pass the bifurcation point and move from the silenced to the high state.
For intermediate silencing strength the bifurcation diagram is similar, however due to the lower siRNA efflux and the stronger silencing response within the cell, it is shifted toward lower siRNA influx. The high and the low equilibrium are connected, meaning that an increase in siRNA influx results in a shift from the high equilibrium to the low one. This is the case in the speckle pattern: all cells initially expand to the high equilibrium, but some cells decline to the silenced state later in infection.
To indicate the measured siRNA influx from the patterns shown in , we placed circles in the bifurcation diagrams for two high and two low cells . The resulting pattern is at a delicate balance: cells that reach the high equilibrium suppress virus growth in neighboring cells, and the silenced cells cause a low enough siRNA influx into the cell to stay at the high state.
Concluding, the effect of siRNA influx from neighboring cells is two-fold. Firstly, siRNA influx creates two coexisting equilibria; secondly, siRNA influx during the initial growth phase of the virus largely determines which equilibrium is reached.
## Combined single-and double-stranded rna cleavage by dicer
It has been proposed that Dicer can cleave ssRNA in addition to dsRNA [bib_ref] Plant virus-derived small interfering RNAs originate predominantly from highly structured single-stranded viral..., Molnar [/bib_ref]. On the cellular level combined singleand double-stranded cleavage by Dicer results in the diverse skewed siRNA ratios that have been observed [bib_ref] Plant virus-derived small interfering RNAs originate predominantly from highly structured single-stranded viral..., Molnar [/bib_ref] [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref] [bib_ref] Molecular bases of viral RNA targeting by viral small interfering RNA-programmed RISC, Pantaleo [/bib_ref] [bib_ref] A simplified method for cloning of short interfering RNAs from Brassica juncea..., Ho [/bib_ref]. Although combined Dicer cleavage of single-and double-stranded RNA can clear the virus for lower Dicer cleavage rates, it can also be less efficient. For the same total Dicer cleavage rate dsRNA cleavage results in somewhat lower plus-strand RNA levels then combined single-and double-strand cleavage [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref]. This is due to the skewed siRNA ratio: many siRNAs are produced that target the minus-strand, but few siRNAs are produced that target the plus-strand. This bias results in slightly less efficient silencing, however when the minus strand is completely degraded, the virus is cleared despite the less efficient response.
The complex feedbacks between the intracellular and the spatial model result in an unexpected effect of ssRNA cleavage in the spatial model. While keeping total Dicer cleavage rate constant, increasing ssRNA cleavage by Dicer shifts the silenced equilibrium to higher RNA levels. An increase of virus levels in the silenced equilibrium results in less pronounced patterns and it can even result in disappearance of the silenced state.
In -F we show patterns for increasing rates of ssRNA cleavage by Dicer 300 hpi. The patterns change from circles to mosaic-like, and the silenced cells change from a dark to a red color, indicating higher virus levels in silenced cells. This effect becomes more clear later in infection . Moreover, the infection pattern shown in fades out (at 450 hpi, shown in , and eventually the entire tissue becomes fully infected.
In -C we show how the equilibria change when changing ratio's of double-to single-stranded RNA cleavage by Dicer. We keep total Dicer cleavage rate constant at 10 cleavages per Dicer per hour. To explain the observed siRNA ratios approximately 0-20% of Dicer cleavages should take place on ssRNA [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref]. In 5% of Dicer cleavages is on ssRNA and in 15%. The silenced state clearly shifts to higher virus levels when ssRNA is cleaved. When more ssRNA is cleaved the equilibrium shifts further upwards until it disappears. In we show the behavior for 19.5% ssRNA cleavage by Dicer, at 20% ssRNA cleavage the silenced state has completely disappeared. The upwards shift of viral RNA levels implies that virus replication is not completely silenced. Indeed dsRNA is produced in the silenced cells.
These results are explained further in the bifurcation diagram in . Shown are the equilibrium values of plus-strand RNA as a function of siRNA influx. The different lines represent a varying fraction of ssRNA cleavage by Dicer. The virion influx increases for increasing ssRNA cleavage by Dicer and the silenced state shifts upwards. Additionally, an increasingly higher siRNA influx is needed to reach this equilibrium. The dots indicate the location of cells from the spatial model that reach the high and the low equilibrium. Due to increased siRNA levels the siRNA influx is increased when ssRNA is cleaved by Dicer. However, due to the less efficient silencing within cells, an increasingly higher siRNA influx is needed to reach the silenced state.
This means that, although combined single-and doublestranded RNA cleavage by Dicer can clear the virus at lower Dicer cleavage rate [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref] , the slight increase in virus equilibrium is disadvantageous at the tissue level as RNA silencing can no longer fully suppress virus replication in silenced cells. However, an advantage of single strand cleavage by Dicer is the slower rate of spread over the tissue.
## Effect of amplification
Amplification of silencing through RDR can decrease viral levels in plants [bib_ref] RDR6 has a broad-spectrum but temperature-dependent antiviral defense role in Nicotiana benthamiana, Qu [/bib_ref] [bib_ref] Arabidopsis SGS2 and SGS3 genes are required for posttranscriptional gene silencing and..., Mourrain [/bib_ref] [bib_ref] Analysis of the involvement of an inducible Arabidopsis RNA-dependent RNA polymerase in..., Yu [/bib_ref]. It also affects symptoms observed: plants without a functional RDR become fully infected or die, while plants with RDR show mild chlorosis or mosaic [bib_ref] A natural variant of a host RNA-dependent RNA polymerase is associated with..., Yang [/bib_ref] [bib_ref] An important role of an inducible RNA-dependent RNA polymerase in plant antiviral..., Xie [/bib_ref].
To study the effect of the secondary response on pattern formation and virus levels we add amplification to the model. At the cellular level RNA silencing with unprimed amplification can clear the virus for much lower silencing strength. It has to be noted however, that unprimed amplification can be triggered by any RNA, and will lead to responses against host RNA. Therefore, a mechanism has to be included to protect the host against auto-immunity [bib_ref] The RNA silencing pathway: the bits and pieces that matter, Groenenboom [/bib_ref]. Primed and guided amplification can increase oscillations and create a new region of behavior, in which the virus is degraded after an initial growth peak. Primed amplification can only be beneficial when there is a net siRNA gain [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref]. At low Dicer cleavage rate the virus equilibrium is slightly decreased on the cellular level. In the spatial model the addition of amplification can result in a change of patterns . Without amplification almost the complete area becomes infected with the virus, with amplification circles or mosaic-like patterns can occur.
Concluding, for low silencing strength amplification has a similar effect to increasing silencing strength or siRNA movement.
New patterns can be observed when the virus is cleared after an initial growth peak at the cellular level. Because the virus can expand and produce virions initially, it is able to spread over the tissue. However, in each cell the virus will be completely degraded by the RNA silencing response. We therefore observe transient patterns that leave healthy tissue behind. These patterns are shown in and 6C. Without siRNA movement a single growing ring occurs, low siRNA movement can lead to a disappearing spot near the infected site. These results indicate that a strong feedback can result in transient infection patterns. When siRNA movement is higher, the feedback becomes smaller due to the efflux of siRNAs, and infection patterns as shown in and 6A occur.
# Conclusion
We have shown that RNA silencing causes local differences in virus accumulation that can be the basis of different virus symptoms developed in plants. We have studied all qualitatively different patterns that can occur for different parameter values.
When siRNAs spread from cell to cell, we observe patterns that can spread over the entire tissue and consist of alternating healthy and infected tissue. When siRNAs are able to spread quickly we observe localized spots around the infection site. The presence of a secondary response can result in transient patterns that leave healthy tissue behind. In accordance with our results, it has been shown that in absence of a secondary response tissue can get fully infected, while with secondary response patterns develop [bib_ref] A natural variant of a host RNA-dependent RNA polymerase is associated with..., Yang [/bib_ref] [bib_ref] An important role of an inducible RNA-dependent RNA polymerase in plant antiviral..., Xie [/bib_ref]. The initial appearance of patterns that slowly fade until the entire tissue is infected occurs only when ssRNA is cleaved by Dicer.
In plants a variety of chlorotic patterns caused by viral infection have been observed. Chlorotic tissue contains high virus levels [bib_ref] Tobacco mosaic virus virulence and avirulence, Dawson [/bib_ref]. Some of the patterns of our model resemble infection patterns in plants. When silencing is weak the entire leaf area becomes infected, corresponding to chlorosis of the entire leaf.
When silencing is stronger we observe a mosaic-like pattern that resembles symptoms observed for many mosaic viruses. A difference with the experimentally observed patterns and our simulated patterns is the scale. We only observe larger scale mosaic when silencing strength is low. When silencing strength is high or when the secondary response is active we observe a small scale mosaic pattern. Our model, however, does not include growth of the tissue, and it has been shown that the healthy spots in mosaic-like patterns come about from cell division of single cells or a cluster of a few cells (reviewed in [bib_ref] Tobacco mosaic virus virulence and avirulence, Dawson [/bib_ref]. Development and growth of our small scale mosaic-like pattern could result in the experimentally observed large scale patterns.
In our model local spots occur when siRNA movement and silencing strength are sufficiently high. The virus infection stays localized to the area around the initially infected cell. We observe two types of spots: spots that are completely infected and ring spots that have a circle of low virus levels in between high levels. Lesions and ring spots are very common local symptoms after inoculation.
Ring patterns similar to our concentric rings have been observed for Tomato ringspot virus [bib_ref] Recovery of Nicotiana benthamiana plants from a necrotic response induced by a..., Jovel [/bib_ref]. Ringspot viruses can also give rise to a large single ring-like spot that resembles the single growing ring that we observed. Tissue at the leading edge of virus infection is infected, tissue in the center of the ring is healthy. We observe this pattern only when there is a very strong feedback, as for example a secondary response. In contrast to our rings, the observed ring patterns do not spread over the entire tissue. This could be due to a long range silencing signal that we have not included in our model. The effect of such signal could be tested by adding an influx of siRNA or dsRNA at a specific site on the grid. This site represents the presence of the vascular system of the plant through which the silencing signal spreads. In this way the effect of different candidates for the long-range silencing signal could be tested.
We have used a detailed model of both intra-and intercellular dynamics of virus replication and RNA silencing.
Nevertheless we were able to analyze mathematically (by bifurcation diagram) the intracellular dynamics that lead to alternative equilibria underlying the formation of infection patterns in plants. We have shown that siRNA movement is the driving force behind the pattern formation observed. Data of Hirai et al. [bib_ref] Antiviral RNA silencing is restricted to the marginal region of the dark..., Hirai [/bib_ref] on siRNA location strongly support our model. Further experiments and parameter validation is necessary to study specific cases, and we hope to inspire researchers to further investigate how the chlorotic patterns on plant leaves relate to virus and siRNA levels. Additionally, it would be very interesting to study the development of these patterns and local virus and siRNA levels in time-series.
In conclusion, we have shown that the interplay of RNA silencing and virus growth within a cell, and the spread of virions and siRNAs between cells can explain the variety of viral infection patterns observed in plants.
# Authors' contributions
MG and PH conceived and designed the models, and wrote the paper. MG performed the numerical computations. All authors read and approved the final manuscript.
# Appendix
## Intracellular dynamics
The entire intracellular model:
The biological meaning of the variables is mentioned to the left of the equations. Multiple RdRPs can bind to minus-strand RNA, we refer to these as 'active RdRPs'. +RISC and -RISC are RISC loaded with siRNA with a plusor minus-strand polarity. The abbreviation 'sec.' stands for secondary and is used to indicate siRNA that is produced through a secondary amplification pathway. Secondary RISC is loaded with secondary siRNA.
All parameter values can be found in [fig_ref] Table 1: Parameter values used [/fig_ref] , as well as a short description of each parameter. The other terms , and are functions for the complex formation between RdRP and RNA strands, Dicer cleavage and amplification, respectively.
The complex formation ( ) between RdRP and RNA strands is saturated for both viral RNA and RdRP (the Beddington-DeAngelis functional response [bib_ref] Mutual Interference Between Parasites or Predators and its Effect on Searching Efficiency, Beddington [/bib_ref] [bib_ref] A model for trophic interaction, Deangelis [/bib_ref] :
Dicer can cleave double-stranded and single stranded RNA and is saturated for D e , D p , D m , P and M. The Dicer cleavage functions, one for cleaving dsRNA and one for ssRNA, are saturated according to the ratio between single-and double-stranded RNA cleavage (q) by Dicer:
The amplification terms are:
Where is unprimed amplification, is primed amplification and is guided amplification. We study the amplification pathways separately. In the case of guided amplification, the siRNAs are not removed when they guide amplification, in contrast to primed amplification. Amplification produces dsRNA that is not used for virus replication (D e ). This dsRNA is degraded into secondary siRNAs with a plus-or minus-strand polarity; Sis p and Sis m respectively.
## Parameters
For the intracellular dynamics we use the parameters previously described [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref]. Parameter values can be found in [fig_ref] Table 1: Parameter values used [/fig_ref]. The only intracellular parameter that we vary in this study is the cleavage rate by Dicer (0 to 15 cleavages per Dicer per hour), because it is representative for a general increase in strength of the silencing response [bib_ref] The dynamics and efficacy of antiviral RNA silencing: a model study, Groenenboom [/bib_ref].
The extra parameters of the spatial model are the percentages of siRNAs and virions that move out of a cell (0 to 10% per hour), and virion unpacking rate.
u p g
[fig] Figure 2: Virus spread over plant tissue. (A) and (C) are screen-shots 75 hours post infection showing the number of virions, siR-NAs and +RNA in the cells. (B) and (D) show the intracellular +RNA levels for a row of cells through time. (A) and (B) show results without silencing and (C) and (D) with silencing. Virions move from cell-to-cell, but siRNAs cannot spread. [/fig]
[table] Table 1: Parameter values used. #mol is number of molecules. [/table]
|
Trends of in-hospital cardiac arrests in a single tertiary hospital with a mature rapid response system
BackgroundMost studies on rapid response system (RRS) have simply focused on its role and effectiveness in reducing in-hospital cardiac arrests (IHCAs) or hospital mortality, regardless of the predictability of IHCA. This study aimed to identify the characteristics of IHCAs including predictability of the IHCAs as our RRS matures for 10 years, to determine the best measure for RRS evaluation.MethodsData on all consecutive adult patients who experienced IHCA and received cardiopulmonary resuscitation in general wards between January 2010 and December 2019 were reviewed. IHCAs were classified into three groups: preventable IHCA (P-IHCA), non-preventable IHCA (NP-IHCA), and inevitable IHCA (I-IHCA). The annual changes of three groups of IHCAs were analyzed with Poisson regression models.ResultsOf a total of 800 IHCA patients, 149 (18.6%) had P-IHCA, 465 (58.1%) had NP-IHCA, and 186 (23.2%) had I-IHCA. The number of the RRS activations increased significantly from 1,164 in 2010 to 1,560 in 2019 (P = 0.009), and in-hospital mortality rate was significantly decreased from 9.20/1,000 patients in 2010 to 7.23/1000 patients in 2019 (P = 0.009). The trend for the overall IHCA rate was stable, from 0.77/1,000 patients in 2010 to 1.06/1,000 patients in 2019 (P = 0.929). However, while the incidence of NP-IHCA (P = 0.927) and I-IHCA (P = 0.421) was relatively unchanged over time, the incidence of P-IHCA decreased from 0.19/1,000 patients in 2010 to 0.12/1,000 patients in 2019 (P = 0.025).
[formula] a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 [/formula]
# Introduction
Despite the advances in the management of in-hospital cardiac arrest (IHCA) over the past decade, IHCA remains associated with poor outcomes. However, while it is frequently preceded by a more gradual, possibly treatable, decline, many cases of IHCAs are considered preventable based on retrospective reviews. To reduce the incidence of IHCA, a rapid response system (RRS) was designed to identify early signs of clinical deterioration and activate a specialized team of caregivers. Most recently, the International Society of RRS recommends that hospitals collect data on IHCAs and their potential predictability.
A cardiac arrest is treated with cardiopulmonary resuscitation (CPR), which is lifesaving for patients who have a history of acute, potentially reversible illness. However, this is not effective if cardiac arrest occurs in patients in fatal conditions with terminal illness. "Do not attempt CPR" (DNACPR) decisions allow resuscitation to be withheld when the chance of success is little or when the burdens of CPR outweigh the benefits. Nevertheless, physicians are occasionally faced with patients requesting full resuscitation against medical advice. More commonly, neither patients nor their family members make such a request, but physicians simply presume that providing CPR comports with the patient's wishes. Therefore, performing CPR on all IHCAs, regardless of the severity of the underlying illness and end-oflife medical decision, may be inappropriate. In contrast, the implementation of RRS also increases the likelihood of DNACPR, which could be partially attributed to the reduced IHCA cases after RRS implementation but with a lesser impact on hospital mortality.
However, most studies on RRS have simply focused on its role and effectiveness in reducing IHCA or hospital mortality, regardless of the type of IHCA. As a result, it has been difficult to determine the overall rate of potentially avoidable IHCA and if this rate is changing with implementation and maturity of RRS. Therefore, this study aimed to identify the rates and characteristics of IHCAs including predictability of the IHCAs as our RRS matures for 10 years, to determine the best measure for RRS evaluation.
# Methods
## Study design, setting, and participants
This retrospective observational study included all consecutive patients who experienced IHCA and received CPR in general wards at Samsung Medical Center, Seoul, South Korea between January 2010 and December 2019; this university affiliated tertiary referral hospital has a 1,989-bed capacity with a hospital-wide medical emergency team (MET) for the RRS. To address the primary research question of whether characteristics of IHCAs in hospitalized adult patients is associated with maturity of our MET over 10 years, we reviewed the clinical data of all treated IHCAs through the electronic medical records. This study was approved by the institutional review board of the Samsung Medical Center and performed in compliance with Helsinki declaration. The institutional review board waived the requirement for informed consent due to the observational nature of the research. Additionally, the patients' information was anonymized and de-identified prior to analysis.
## Plos one
## Operation of the rrs
The hospital-wide MET at the Samsung Medical Center was introduced at the beginning of March 2009, consisted of either fellows that were training in critical care or senior residents in internal medicine. Since March 2013, the MET consists of dedicated intensivist physicians, including critical care fellows and attending intensivists, which provide round-the-clock coverage. All hospital medical personnel were presented with information about the MET and educated to prevent inadequate clinical assessments of patient deterioration that cause delays in the MET activation. Before implementation of the automated system, physicians and nurses directly contacted the MET using a dedicated phone number when a patient met any single criterion. Activation was also allowed when the medical staff was concerned about changes in their patient's clinical condition, even in the absence of physiological disorders that meet the criteria. In August 2016, the MET initiated an automated alert and activation system for all ward patients using a modified early warning score (MEWS). The MEWS was automatically calculated using five physiological parameters (systolic blood pressure, heart rate, respiratory rate, body temperature, and level of consciousness) when nurses records the patient's vital signs on the electronic medical record. Patient vital signs were recorded at the bedside immediately after measurement using a laptop or portable device whenever possible. MEWS was automatically updated with each new vital sign recorded. The frequency of measuring vital signs was made according to the order of the attending physician, but vital signs were usually measured at least four times a day and more often when the patient's clinical condition changed. If the MEWS was 7 or higher, an automated alert was sent to MET as a text message in real-time, 24-hours a day, 7 days a week, and MET was automatically activated. Calls for MET activation were available for all patients regardless of do-not-resuscitate status during the study period. When activated, the MET is expected to arrive within 10 min, complete patient assessments within 30 min, and order diagnostic tests and therapeutic treatments relevant to the patient's condition. In certain clinical problems requiring specialized expertise, other teams such as acute myocardial infarction team, acute stroke team, and acute care surgery team also can be activated. Following assessment and initial treatment, an individual treatment plan is created for each patient, and a joint decision is made about whether to transfer the patient to the intensive care unit (ICU). The issue of limitation on medical intervention and end-of-life care can also be discussed at this point. After assessment and therapeutic interventions by the MET, patients who are considered to require treatment and monitoring that cannot be provided outside of the ICU are transferred to the ICU, while patients in a stable condition remain on the general ward. For patients who are not fully stabilized and require intensive monitoring but are able to manage with lower levels of care than in the ICU, admission to the unit is decided on a case-by-case basis. The MET determined the patient's disposition and shared information about the advance care plan with the primary care team. Discussion about end-of-life care and ceiling of care were also included in certain patient populations. The decisions about completion of intervention and disposition were left to the judgment of each member of the MET without specific criteria, but the decision making generally followed international guidelines . Details of all MET calls were recorded as soon as possible after the event by a member of the team and were entered into a registry. This recorded patient demographics, reasons for the MET activation based on calling criteria, time of the first documented physiological disorder, modified early warning score, time of the MET activation and deactivation, vital signs at the time of the MET activation and deactivation, interventions delivered by the MET, and the final outcomes including the patient's disposition after the clinical episode. These data were supplemented on the next day with a retrospective review of hospital medical records before registration for quality control of registry data.
## Definitions
All IHCAs were classified into three groups: Preventable IHCA (P-IHCA) was defined as a cardiac arrest with preexisting signs of acute physiologic disturbance that fulfilled the MET activation criteria from 8 hours to 30 minutes before arrest. Non-preventable IHCA (NP-IHCA) was defined as a cardiac arrest that occurred within 8 hours after admission, or without any record of vital signs within 8 hours before arrest, or within 30 minutes after drug administration or procedures, or from unexpected lethal arrhythmia; this includes cardiac arrest that occurred within 30 minutes after MET activation. Inevitable IHCA (I-IHCA) was defined as IHCA in patients who had already requested a DNACPR order or were in terminal health conditions. Cases that were difficult to be classified were resolved by consensus of the MET team.
# Statistical analysis
The MET dose was calculated by the number of MET calls per annum divided by the total number of discharged patients per year, represented as cases per 1,000 patients. In addition, each rates of IHCAs according to the classification or in-hospital mortality was calculated by the number of IHCA or in-hospital mortality per annum divided and represented as cases per 1,000 patients. Data were presented as number and percentages. The annual changes of the numbers and rates were analyzed with Spearman correlation analysis and Poisson regression models, respectively. All data were analyzed using SPSS version 22 (IBM Corp., Armonk, NY).
# Results
During the 10-years study period, 843,180 patients were admitted to Samsung Medical Center and a total of 824 consecutive IHCAs were recorded for adult patients. After excluding duplicated CPRs for the same IHCAs (n = 24), a total of 800 treated IHCAs with CPR on the general ward were retrieved for the primary analysis (0.95/1,000 patients). The baseline characteristics of 800 IHCA patients are given in. There were 467 (58.4%) male, and the median age was 64.5 (IQR, 53.0-74.0) years. Malignant disease (47.8%) and cardiovascular disease (26.8%) were the most frequent comorbidities. The median hospital admission day before arrest was 9.7 (IQR 4.0-22.7) days. The most IHCA was occurred during weekdays (68.9%) and half of events occurred at daytime. Common location was general ward (58.3%), followed by monitoring room (36.4%). Cardiovascular (29.5%) was common cause of arrest, and 50.6% initially presented with pulseless electrical activity rhythm. Extracorporeal cardiopulmonary resuscitation was applied to 22 (7.2%) patients.
As MET matured, the number of the MET activations increased significantly from 1,164 in 2010 to 1,560 in 2019 (P = 0.002). In addition, the median time from derangement to MET activation decreased from 66 minutes in 2010 to 40 minutes in 2019 (P < 0.001).
Of the 800 IHCA patients, 149 (18.6%) had P-IHCA, 465 (58.1%) had NP-IHCA, and 186 (23.2%) had I-IHCA.
Among 252 patients (31.5%) survived at discharge, only 121 (15.1%) patients were managed by our MET within 24 hours before the IHCA. However, the MET dose was relatively unchanged over time (P = 0.531), since the number of admitted patients increased from 72,468 in 2010 to 100,788 in 2019 (P < 0.001). Finally, in-hospital mortality rate was significantly decreased from 9.20/1,000 patients in 2010 to 7.23/1,000 patients in 2019 (P = 0.004).
The trend for the overall IHCA rate was stable from 0.77/1,000 patients in 2010 to 1.06/ 1,000 patients in 2019 (P for trend = 0.720) . However, the incidence of NP-IHCA (P for trend = 0.382) and I-IHCA (P for trend = 0.054) was relatively unchanged over time, while that of P-IHCA decreased from 0.19/1,000 patients in 2010 to 0.12/1,000 patients in 2019 (P for trend = 0.006) .
# Discussion
This study investigated the change of overall and various type of IHCAs following RRS implementation and maturation over 10 years. The major finding is that the RRS call had been increased significantly as the RRS had been maturated, and the P-IHCA and in-hospital mortality were decreased significantly. However, the overall rate of IHCAs did not change significantly during the study period.
## Plos one
Trends of IHCAs with a mature RRS Although there is no standard measure for evaluating the maturity of RRS from the existing literature on RRS maturity, several studies have revealed the correlation of the number of activation with the maturity of RRS. In a long-term observational study, Herod R et al. found that progressively increased number of RRS activations concurred with lower hospital mortality. Moriarty JP et al. also found that the reduction in rescue failure rates was associated with a substantial increase in the number of RRS activation. In addition, timeless response to patient deterioration has been recommended as quality metrics of RRS process, since delayed activation of RRS is associated with higher in-hospital mortality. In the present study, the increased number of the RRS activations and decreased the time from derangement to RRS activation concurred with lowered in-hospital mortality over 10 years, although no causality could be concluded. Therefore, it might be considered that our RRS has matured over the past decade.
Several previous studies have shown reduced incidence of IHCAs after the implementation and maturity of RRS. However, these studies simply focused on reducing IHCAs regardless of the predictability of IHCA, although the overall rate of IHCAs might be limited for the evaluation of RRS. Therefore, potentially preventable IHCAs, rather than total number of IHCAs, is recommended as a quality metric for the evaluation of RRS by the International Society of RRS. In this study, the trend for the overall IHCA rate was stable with maturity of our RRS over 10 years. The lack of change in the overall rate of IHCAs might be associated with the number of I-IHCAs that were not suitable for resuscitation, which contributed to the overall rate of IHCAs. Therefore, 23.2% of IHCAs might receive inappropriate CPR despite the futile situation reflecting both a low chance of survival and likely poorer quality of life afterward if spontaneous circulation is returned. This highlights a potential problem with using the overall rate of IHCAs as an outcome measure for RRS, which could explain the failure of previous studies to demonstrate consistently the efficacy of RRS in decreasing total hospital mortality.
Our results indicate that the most common circumstances of P-IHCAs were sudden critical illness in under-monitored patients and delays in initiating RRS response for monitored patients who met crisis criteria, which are consistent with previous reports. Therefore, more IHCAs might be preventable by closer monitoring on floors and by preventing delays in addressing deterioration in patient condition. Effective risk management necessitates that preventable IHCA is minimized. Therefore, the effort for reducing preventable IHCA, rather than overall IHCA, could be a more appropriate quality metric to measure the clinical outcomes of RRS implementation and maturation; however, inconsistent definitions have limited its generalizability in a wide range of healthcare settings.
Although this study provides additional information on a more appropriate quality metric for RRS implementation and maturation using simple methods that are reproducible within the existing resources of most hospitals, there are several limitations that should be acknowledged. First, the study was limited by its inherent retrospective observational nature. However, data on treated IHCAs were prospectively collected from consecutive patients received CPRs. Therefore, our cohort is more likely to reflect the patients encountered in routine practice and thus can be readily applicable in similar settings. Second, the present study was conducted at a single institution with physician-based MET. Accordingly, our findings may have limited generalizability in other RRS. Finally, an automated alert and activation system was integrated into the original RRS activation process in August 2016. However, this change of activation process could be itself a sign of the maturation of our RRS (S1 .
# Conclusion
In conclusion, the incidence of P-IHCA could be a more appropriate quality metric to measure the clinical outcomes of RRS implementation and maturation than overall IHCA. Supporting information S1 Comparison of MET activation, incidence of IHCAs, and in-hospital mortality before and after implementing the automated alert and activation system in August 2016.
|
EWSR1—The Most Common Rearranged Gene in Soft Tissue Lesions, Which Also Occurs in Different Bone Lesions: An Updated Review
Citation: Flucke, U.; van Noesel, M.M.; Siozopoulou, V.; Creytens, D.; Tops, B.B.J.; van Gorp, J.M.; Hiemcke-Jiwa, L.S. EWSR1-The Most Common
# Introduction
Ewing sarcoma was molecularly defined by Delattre et al. in 1992 upon the identification of the Ewing sarcoma breakpoint region 1 (EWSR1) located on chromosome 22q12.2 and the term for this gene was coined. EWSR1 is a multifunctional protein ubiquitously expressed in most cell types, indicating diverse roles in physiological cellular processes, including organ development and aging. Genetic and epigenetic pathways are modulated by EWSR1 but the exact mechanisms are still poorly understood. EWSR1 belongs to the FET (also known as TET) family of RNA-binding proteins that also includes Fused in Sarcoma (FUS), and TATA-box binding protein Associated Factor 15 (TAF15). As a consequence of the multifunctional role of EWSR1 leading to a high frequency of transcription of the chromosomal region where the gene is located, EWSR1 is exposed to aberrations such as rearrangements. Consecutive binding to other genes leads to chimeric proteins inducing oncogenesis. These various somatic genetic rearrangements involving EWSR1 result in a fusion of its N-terminal coding region to the C-terminal DNA binding domain of one of several transcription factors. They are reported to act as aberrant transcription factors with the N-terminal domain of EWSR1 as a strong transactivator. The other TET family members are homologous and are involved in strikingly similar translocation events giving rise to the production of structurally similar oncoproteins.
With the advent of widely used modern molecular techniques during the last decades, it has become obvious that EWSR1 is involved in development of diverse benign and malignant tumors with mesenchymal, neuroectodermal, and epithelial/myoepithelial features. As oncogenic transformation mediated by EWSR1-fusion proteins leads to such diverse tumor types, there must be a selection on a multipotent stem cell level.
In this review, we will focus on the wide variety of soft tissue and bone entities, including benign and malignant lesions, harboring EWSR1 rearrangement. Fusion gene analysis is the diagnostic gold standard in most of these tumors. We present clinicopathologic, immunohistochemical, and molecular features and discuss differential diagnoses.
## Ewing sarcoma
Arthur Purdy Stout and James Ewing were the first to describe this aggressive small, blue round-cell entity in 1918 and 1921, respectively. Later on, the chromosomal translocation was found by , the second breakthrough of translocation/fusion-gene associated sarcomas following alveolar rhabdomyosarcoma (ARMS). Subsequently, the fusion gene has been detected as mentioned in the introduction, being the genetic hallmark by an otherwise aspecific small blue, round-cell tumor.
Ewing sarcoma, the prototypic round-cell sarcoma, is relatively common in comparison to other small blue round-cell sarcomas. It arises in soft tissue and bone of children, adolescents, and young adults. Exceptionally, older patients are affected. The mean age is in the second to third decade. White males have the highest incidence and black females the lowest due to ethnic genetic preposition differences. Tumors can originate anywhere in the body, and around 80% of the neoplasms arise in the bone with preference sites in decreasing order of frequency: lower extremities, pelvis, upper extremities, ribs, spine, and craniofacial. Distribution in the soft tissue is extremities, chest wall, retroperitoneum, paravertebral, pelvis, and head and neck. Visceral organs, skin, and epidural spaces are rarely involved. The origin of the peripheral nerve as reported bycan clinically be confused with malignant peripheral nerve sheath tumor.
Macroscopically, these infiltrative lesions are (multi)nodular), fleshy, and often necrotic. A pseudocapsule can be present in soft tissue neoplasms. Post-therapy specimens show fibrosis, necrosis, and hemorrhage, often without visible tumor.
Histologically, Ewing sarcoma is composed of cellular sheets of relatively featureless small cells with round dark nuclei and inconspicuous cytoplasm. In some cases, cells are larger displaying more nuclear variability. The cytoplasm can appear clear due to retraction artefacts. Homer-Wright rosettes may be numerous in a subset of cases initially called peripheral primitive neuroectodermal tumors. Adamantinoma-like Ewing sarcoma shows more cohesive sheets and nests of cells with peripheral palisading, prominent desmoplastic stroma with production of hyaline membrane collagen, presence of keratin pearl formation, and comedo-like necrosis. These lesions are predestinated for misinterpretation as carcinoma, since keratins, including high molecular keratins, p40, and p63, are commonly positive.
Immunohistochemically, CD99 is specific in its distinct staining pattern of the cellmembrane. Nuclear FLI and ERG expression is commonly observed in the cases with corresponding fusion genes. Neuroendocrine markers may be expressed. Keratin-expression, often dot-like, was found in 1/3 of the cases; it can be confused with small-cell carcinoma, especially when combined with the expression of p40 and p63. This is of particular importance in the head and neck area. Expression of NKX2-2 in Ewing sarcoma seems to be highly sensitive, with imperfect specificity in comparison to other small, blue round-cell tumors. Expression of desmin is reported in a few cases, and can be confused with ARMS or desmoplastic small round-cell tumor (DSRCT). Ewing sarcoma is genetically characterized by binding of EWSR1 or other members of the TET/FET family to members of the ETS family. Approximately 85-90% of the Ewing's sarcomas display the translocation t(11;22)(q24;q12) resulting in the EWS/FLI1 fusion gene, and approximately 5-10% harbor a EWSR1-ERG fusion gene. The remaining cases show rare gene partners, such as ETV1, ETV4, and FEV, and EWSR1 can be substituted by FUS.
Although prognosis has improved markedly for patients with primary disease (5-year survival rate around 65%), presence of metastatic disease at time of diagnosis or early relapse leads to an adverse prognosis (5-year survival rate around 25-30%), with adequate surgical resection, aggressive multimodal chemotherapy, and adjuvant local radiotherapy being the optimal treatments. Differential diagnoses are listed in.
## Round-cell sarcomas with ewsr1-non-ets fusions
## Nfatc2-rearranged sarcoma
NFATc2-rearranged sarcoma was first described in 2009 by Szuhai et al.. It is another, apparently very rare, primitive round-cell tumor with a methylation profile distinct from Ewing sarcoma, probably due to the non-ETS fusion gene. Therefore, it belongs in the current WHO classification to the category of "round-cell sarcoma with EWSR1-non-ETS fusions".
NFATc2-rearranged sarcoma affects males and females with predominance of males at least in the first studies. There is a broad age range from childhood to the elderly with a median age in the late third decade of life.
As Ewing sarcomas, these neoplasms originate mainly in the bone with predilection for the long bones, particularly femur and humerus. A few cases were reported to be localized in soft tissue and intraabdominally.
Histology demonstrates sheets, lobules, cords, and trabeculae of small, blue, round cells or less commonly spindle cells with slightly irregular nuclei. Nuclear pleomorphism is described in some lesions. There is a variable stromal reaction, being sclerotic, hyaline, myxoid, myxohyaline, and chondromyxoid. Cartilaginous or osteoid-like areas are rarely described.
Immunohistochemistry shows reaction for CD99; for AGGRECAN; and inconsistently for panCK AE, S100, BCOR, WT1, ERG, and ETV-4. Desmin, NKX3-1, and SATB2 are negative.
Amplification of the EWSR1-NFATC fusion gene is typical and can support the diagnosis when present.
In a few cases, rearranged EWSR1 is substituted by FUS. However, such cases show a different transcriptomic profile. While EWSR1-NFATC tumors were strongly enriched in genes associated with inflammatory response, the FUS-NFATC2 tumors showed a signature of proliferation and drug resistance.
The outcome of patients is uncertain, since response to chemotherapy in unclear. Histological response to multimodal therapy seen in the resection-specimens was poor. Patient studies until now have been too small to draw definitive conclusions for prognosis. Favorable outcome is reported in few cases with long-term follow-up, mostly after complete resection. No data on radiotherapy effect are available.
Differential diagnoses are listed in.. Recently, more knowledge was gained based on advanced molecular technologies. EWSR1-PATZ-rearranged sarcoma seems to be a separate entity with a wide clinicopathological spectrum. VEZF1 is considered a paralogue of PATZ, and the few cases described with EWSR1-VEZF1 are similar to the PATZ-rearranged cases in terms of morphology and immunoprofile.
The age ranges from early childhood to the elderly, with an average age in the fourth decade. The anatomic sites vary, with lesions being located superficially and deep, mainly in the trunk (thorax, including lung; abdomen), and rarely in the head and neck and extremities. Intracranial localization is also reported.
Grossly, tumors are either well-circumscribed or infiltrating showing on the cut-surface a tan-yellow or grey-white color.
Histology is striking variable ranging from small blue round-cell morphologyto lesions showing a mixture of spindled, epithelioid, rhabdoid, ovoid, and round cells in varying fractions. The cells are commonly bland-looking with monomorphic nuclei and a low mitotic rate. They can be arranged in sheets, nests, and/or fascicles. Glandular structures as seen in synovial sarcoma, and pseudocystic spaces may be present. The stromal reaction is diverse and can be colleagenous, hyaline, or myxoid. There are often thick bands of collagen. Hyalinization of vessel walls are other possible features. Necrosis may occur but is mostly not prominent.
## 1.
Low-grade appearing tumors: the spindled, epithelioid, ovoid, and round cells are set in a hyaline stroma reminiscent of solitary fibrous tumors and myoepithelioma.
## 2.
Intermediate and high-grade appearing neoplasms have a round and/or ovoid morphology with few spindle cells and slight stromal component. The tumors of this second subgroup resemble other small, blue round-cell tumors, e.g., ARMS, BCOR-, and CIC-rearranged sarcoma.
Whether a transition may occur from low-grade to high-grade morphology as seen in myxoid/round-cell liposarcoma amongst others is yet unknown.
There is a polytypic immunophenotype with variable expression of neural, skeletal muscle, and epithelial markers. OLIG2 is also positive.
In contrast to NGS (or RT-PCR), FISH is not the method of choice for confirming the presence of the fusion gene due to the close proximity of the gene loci of PATZ and EWSR1 on chromosome 22q12.
Based on whole transcriptome sequencing, EWSR1-PATZ rearranged sarcomas are different from other EWSR1-related sarcomas.
Lesions can be aggressive or follow a more favorable course. It is unclear if the above-mentioned morphological grading is prognostic for outcome. In addition, deletions of CDKN2A/B and MDM2 gene amplification were associated with fatal outcome in one study and may therefore be a negative predictor of outcome.
## Small blue round-cell tumor with ewsr1-smarca5 rearrangement
A single case has been described in a 5-year-old female with a mass in the lumbosacral spinal canal with small blue round-cell histology and an immunoprofile like Ewing sarcoma. Cytogenetics showed a t(4;22) (q31;q22) as sole abnormality resulting in an EWSR1-SMARCA5 fusion. More cases are necessary for definitive categorization.
## Desmoplastic small round-cell tumor (dscrt)
The first description has been done by, and one year later, the consistent translocation t(11;22)(p13;q12) was found by Sawyer et al.. Ladanyi et al. detected the corresponding fusion gene EWSR1-WT1. WT1, a suppressor of transcription, is expressed in primitive, developing mesothelium. It is therefore not surprising that DSRCTs are classically located in the abdominal cavity with growth along the mesothelial membrane, often with multifocal spread at diagnosis. Origin in the small pelvis with ovary, or spermatic cord/paratesticular (tunica vaginalis), thoracic cavity/pleura, head and neck region, cranium/intracerebral, cauda equina, and extremities is rarely reported. Mostly adolescents and young adult males with a mean age in the second decade of life are affected, whereas females and older patients are rarely involved.
Grossly, this tumor is firm and shows a multinodular growth pattern with infiltration into adjacent structures and organs (e.g., liver). It has a solid and gray-white appearance with possible necrotic areas and hemorrhage.
Histologically, DSRCT is composed of irregular sheets, nests, trabecula, and cords of small cells with hyperchromatic nuclei and inconspicuous cytoplasm surrounded by a prominent desmoplastic stroma, which is a hallmark. When absent, other small round (or spindle) cell tumors could be superior in the differential diagnosis. Cells may vary in shape and size. The nuclei are round, ovoid, or spindly. In some cases, a rhabdoid/eosinophil or clear cytoplasm was noticed. The latter could be due to retraction during fixation process. Glandular/tubular or rosette-like structures have been identified in some cases. Mitotic figures may be numerous. Necrosis, possibly with calcification, can be present. The cells show immunohistochemically a polyphenotypic profile with expression of epithelial (EMA, broad-spectrum keratins (sometimes dot-like), myogenic (desmin, SMA in few cases), and neural markers (CD57, NSE). However, when this polyphenotype is incomplete, as sometimes seen, other small, blue round-cell tumors are conceivable, at least without the classical clinical context. WT1 is only of diagnostic value when an antibody toward the carboxy terminus is used showing a nuclear staining pattern. The presence of EWSR1-WT1 confirms the diagnosis in >95% of the cases. Multiple copies of the fusion gene have also been found. The differential diagnoses besides other small blue round-cell sarcomascan be carcinoids and small-cell carcinoma, including Merkel cell carcinoma because of the desmoplastic stroma reaction. Neuroblastoma, lymphoma, and blastemic Wilms tumor are probably less relevant in the differential-diagnostic spectrum(see. Treatment is challenging, and survival is low despite initial response to multimodal chemotherapy. Debulking is usually the surgical procedure, and tumor-free margins can often not be achieved when extended surgery is performed. For local control, several different additions have been tried with varying effects, such as total abdominal irradiation or surgery in combination with HIPEC. However, most patients relapse with locally disseminated disease not responsive to further treatment. Distant metastases may occur with involvement of lungs, liver, lymph nodes, bone, kidney, pancreas, spleen, adrenal gland, and small pelvis]. An indolent clinical course is reported in some mainly unusual cases.
## Myxoid liposarcoma (mls)
MLS is the only translocation-associated liposarcoma-subtype recapitulating more or less normal lipogenesis with maturation arrest. Limon et al. detected the most common translocation in 1986. The specific fusion genes FUS-DDIT3 (ca 90% of the cases) and, more rarely, EWSR1-DDIT3 (in up to 10% of cases) are the result of the t(12;16)(q13;p11) and t(12;22)(q13;q12), respectively. DDIT is an enhancer binding family of transcription factor involved in erythropoiesis and adipogenesis.
MLS is the most common liposarcoma arising in children, adolescents, and young adults. It comprises up to 35% of all liposarcomas and has, in 1/3 of cases, the tendency to metastasize to other soft tissue sites, including mediastinum and retroperitoneum, and also to bone, lung, and liver, with a consecutive fatal outcome. The classical primary localization (2/3 of the cases) is the deep soft tissues commonly of the thigh. Cases of the retroperitoneum are almost exclusively metastases, with some exceptions. At distal sites of the extremities, this tumor is exceedingly rare.
Macroscopically, the lesion is multinodular and gelatinous on cut surface. High grade areas are firm and grey-white in appearance.
Histology depicts a nodular growth pattern of relatively low cellularity with enhancement of cells at the periphery of the nodules. There is a proliferation of uniform bland, round-to-oval-shaped primitive cells intermingled with a variable amount of lipoblasts of different stages in an abundant myxoid stroma. Slightly pleomorphic nuclei are often associated with multivacuolated cytoplasm. In addition to the classical features, a nested pattern or islands of primitive cells, areas of extensive maturation, pseudoacini, or a cord-like growth pattern can be obvious. The very characteristic delicate plexiform ('chicken-wire') capillary vasculature is less obvious in cellular areas. When present, hemangiopericytoma-like vessels can be confused with solitary fibrous tumor, especially the lipomatous variant. Stromal hyalinization is rarely reported, which can be misleading if prominent. Hypercellular morphology with more large round cells with increased nuclear/cytoplasmic ratio, distinct nucleoli, and a small amount of intervening myxoid stroma is called round-cell liposarcoma and is associated with an inferior prognosis when more than 5% of the neoplasm is affected.
Immunohistochemistry shows variable expression of S100, which is of little value. The determination of the fusion gene is therefore an important confirmation that has consequences therapeutically. Recently, nuclear expression of DDIT3 as an appropriate immunohistochemical surrogate marker has been reported.
In the differential diagnosis is lipoblastoma, which rarely occurs in adults, at least in its primitive form, and myxoid pleomorphic liposarcoma when slight pleomorphism is present. Chondroid lipoma can be considered because of lipoblastic differentiation, and soft tissue angiofibroma shows some similarities in terms of the branching capillaries, which do not have such a delicate appearance in the latter. Small, blue round-cell tumors can be pondered when round-cell morphology without obvious lipoblasts is observed(seeNeo-adjuvant radiotherapy leading to a good response with maturation and hyalinization, followed by resection, is the optimal treatment. Recurrences are less frequent. When metastasized (up to 1/3 of the patients), the outcome is poor; however, a slow progression may be observed.
## Tumors with ewsr1/fus fused to the creb-family
ATF1, CREB1, and CREM are members of the CREB-family of transcription factors playing a pivotal role in diverse physiological processes. They act as fusion partners of EWSR1 or alternating FUS in several benign, intermediate, and fully malignant tumors, and the growing list includes mesenchymal, neuroectodermal, and epithelial neoplasms. Secondary genetic and/or epigenetic events seem to be mandatory for the specific oncogenesis. Whereas EWSR1-ATF1 and EWSR1-CREB are the two most characterized fusions, EWSR1-CREM is less well studied. Whether or not AFH (including its myxoid variant), PPMS, so-called mesothelioma, and EWSR1-CREM undifferentiated sarcoma are a spectrum of one tumor type will yield further studies in the future.
## Angiomatoid fibrous histiocytoma (afh)
This lesion was firstly recognized by Enzinger in 1979 with the main characteristics being reported in his seminal paper. In 2000, the first genetic report was published showing FUS-ATF1 as the result of t(12;16) (q13;p11). Later on, it became apparent that EWSR1/FUS-CREB/ATF1/CREM are the alternating candidates for the gene fusion, with EWSR1-CREB1 being the most common in soft tissue lesions. EWSR1-CREM positive cases are recently reported. Multiple copy numbers of the fusion gene seem to be associated with pleomorphism.
AFH affect mainly children, adolescents, and young adults. However, the age range is broad. When located in superficial soft tissue, clinical presentation is often a palpable, slowly growing indolent nodus imposing as hemangioma or lymph node.
Rarely accompanied systemic symptoms such as malaise, pyrexia, and anemia are documented. Classically, AFH arises subcutaneously in the extremities followed by the trunk and head and neck. Involvement of deep soft tissues and visceral sites is increasingly reported due to higher diagnostic standards (including molecular diagnostics). They include mediastinum, retroperitoneum, intraabdominal, lung, brain, bone, and ovary.
Grossly, the neoplasm is circumscribed, (multi)nodular or (multi)cystic, grayishyellow, and hemorrhagic, and can be as large as 10 cm.
Histologically, these circumscribed (multi)nodular and/or multicystic lesions possess a fibrous pseudocapsule that is surrounded by a prominent often lymph node-like mixed lymphocytic infiltrate with variable germinal center formation and presence of plasma cells. The inflammatory component may also be intermixed with tumor. The tumor cells are arranged in sheets, nodules/whorls, aggregates, short fascicles (storiform), and reticular formations (when myxoid). The cells are histiocytoid with a syncytial aspect showing an ovoid, epithelioid, or spindled appearance and bland-looking nuclei with fine chromatin and moderate amount of ill-defined eosinophilic cytoplasm. Centrally, a cannonball arrangement can be seen. Unusual characteristics are scattered large cells, nuclear grooving, and bizarre and irregularly folded nuclei. Rhabdoid, clear cells, or osteoclast-like giant cells and Ewing-like areas may be present. Nuclear palisading as seen in schwannomas has been rarely observed. Mitotic rate is usually low, but atypical mitoses are not a worrisome sign. Hemorrhage and blood-filled spaces are, when present, a hallmark of this lesion often associated with hemosiderin deposition. The stroma can be unremarkable, sclerotic, "desmoplastic", or myxoid. Edema can be prominent, and a slit-like pseudovascular pattern may be seen. Additionally, perivascular hyalinization is a possible sign.
Immunohistochemistry shows expression of ALK in almost all cases and desmin in approximately 50% of the cases. EMA, CD99, and CD68 are variably expressed. Other smooth muscle markers such as SMA and caldesmon are sometimes positive. CD21 may be positive in some cells. Myogenin and MYOD1 are consistently absent.
Intracranial myxoid mesenchymal tumors are suggested to be a variant of AFH. This meets the observations of peripherally located AFHs with myxoid changes. Differential diagnoses are listed in. AFH has a low to intermediate biologic potential, with most cases behaving benign. Recurrence is reported in up to 15% of the cases, especially when marginally excised. Metastases have been reported in up to 5%, most frequently in the regional lymph nodes and exceptionally in the lungs, liver, and brain. Pleomorphism and increased mitotic activity are not associated with worse outcome.
## Primary pulmonary myxoid sarcoma (ppms)
PPMS, an entity rendered by Thway et al. in 2011, is an EWSR1-CREB1 fusion gene associated neoplasm of the lung. It predominantly arises intrabronchially and involves the pulmonary parenchyma. It affects young to middle-aged adults, mainly women.
Morphologically, lesions are (multi)nodular with a pale and glistening cut surface on macroscopy. Microscopically, round-to-ovoid or spindle cells are situated in a myxoid matrix arranged in cords with possible reticular growth, as seen in extraskeletal myxoid chondrosarcoma (see there), nests, and sheets. Cells are relatively bland looking with at most slight nuclear pleomorphism. There is a low mitotic index and possibly focal necrosis. An inflammatory component is often obvious with presence of lymph follicles, either marginal or intralesional. Immunohistochemistry is of little value, with a faint reaction for EMA, but helps to exclude other lesions. Lesions may be benign or malignant (with described metastases in kidney and brain).
To what extent AFH and PPMS are related needs to be further explored. A close relationship is conceivable. Differential diagnoses are listed in. Non-specific; S100 (focal) NR4A3-EWSR1/TAF15/TCF12/TFG
## Ewsr1-crem undifferentiated sarcoma
Currently, these tumors are still relatively unexplored. One aggressive intraabdominal tumor in an adolescent was described consisting of swirls of uniform spindle cells with intercellular delicate collagen. Immunohistochemistry shows expression of vimentin, cytokeratin AE1/3, and CD56. EMA, CD34, ALK, synaptophysin, and DOG1 were focally positive. INI1 and H3K27me3 were retained. Negative were desmin, myogenin, S100, SOX10, NUT, CD31, and smooth muscle markers.
The other tumor was that of a 63-year-old woman with localization in the chest wall and a sclerosing epitheloid fibrosarcoma-like morphology with MUC4 and synaptophysin expression. The stroma was more fibrillary than sclerotic. After 17 months, there was no evidence of disease. A comparable case was included in the study by. Sequencing techniques should be applied in the right context to detect this rearrangement.
## Clear-cell sarcoma (ccs)
Enzinger was the first to describe CCSs systematically in 1965, also called melanoma of soft parts, because of overlapping morphological and immunohistochemical features with melanoma. In the years 1990, 1991, and 1992, the translocation (12;22)(q13;q12) was found, one year after the EWSR1-ATF1 gene fusion by Zucman et al. 1993. Later on, it became apparent that CREB1 and CREM are substituents of ATF1 involving a smaller subset of cases.
Young adults (30-40 years) with equal sex distribution are the main group of patients. However, there is a broad age range from children to the elderly. Additionally, there is a race distribution with the overwhelming majority of patients being Caucasian, whereas black people and Asians are uncommonly afflicted.
The deep soft tissue of the extremities is for the most part involved. Superficial soft tissue and skin localization does not exclude diagnosis. Extension into bone can be seen. Unusual tumor sites include the trunk, e.g., breast, anus, mediastinum, pleura, retroperitoneum, and the head and neck area. Symptoms depend on site and are nonspecific. Tumors can be large (up to 15 cm).
Grossly, tumors are firm and roughly spherical, with a smooth, nodular, or coarsely lobular surface. Most are well defined and surrounded by a fibrous pseudocapsule. Illdefined lesions are less frequently reported. The cut surface is usually gray to white. Brown areas and gelatinous foci are observed.
Microscopically, the classical pattern is that of a well delineated tumor with extension into adjacent structures consisting of nests or short fascicles of monomorphic round-to spindled cells with ample clear to eosinophilic cytoplasm, vesicular nuclei, and prominent nucleoli. Mitotic activity is variable but rarely brisk. Cells are separated by delicate fibrous septa. Typical are multinucleated giant cells with wreath-shaped nuclei. A diffuse growth pattern and pleomorphism are seen in some instances. Also reported are rhabdoid cells, an alveolar growth pattern, a seminoma-like appearance with a lymphocyterich fibrovascular stroma, and a dominant stromal reaction. Melanin pigment may be present, and necrosis can be found. Melanoma markers are expressed using immunohistochemistry. Melanoma is the most important differential diagnosis. It has to be discriminate from CCS because of the therapeutic consequences. Other differential diagnoses are listed in. Excision is the treatment of choice. In a large epidemiological study, radiotherapy and chemotherapy were applied in 34% and 20%, respectively. Whether this confers any survival advantage is unclear. At least some tumors are reported to be chemosensitive. Local recurrences and in-transit metastases are reported in around 20% of the cases. Sites of metastases are the lung (most commonly) and lymph nodes. The propensity to metastasize to lymph nodes is typical in comparison to most other sarcoma types. Overall, estimated 5-and 10-year survival is approximately 50% and 38%, respectively.
## Clear-cell sarcoma-like tumor of the gastrointestinal tract (osteoclastrich tumor of the gastrointestinal tract or malignant gastrointestinal neuroectodermal tumor)
In 1993, Ekfors et al. described the first case of a clear-cell sarcoma in the duodenum. However, a similar case was already reported in 1985 under the term malignant neuroendocrine tumor of the jejunum with osteoclast-like giant cells. In 1998, it became apparent that these tumors share the same genetic characteristics, with most tumors harboring an EWSR1-CREB1 fusion and less often an EWSR1-ATF1 fusion. Since then, there has been discussion whether these tumors are CCSs or a separate entity as they have morphological and genetic similarities.
Tumors arise in the gastrointestinal tract (small bowel, stomach, colon, and esophagus) of predominantly young adults and children. However, the age range is broad, including the elderly. An abdominal mass with pain and intestinal obstruction are the main clinical features. Lesions may be large, up to 15 cm. The appear macroscopically firm, solid, and tan-white. Microscopy shows primary involvement of submucosa and muscularis propria, occasionally with mucosal involvement. Cytomorphology resembles CSS. However, tumors show, besides the nested pattern, arrangement in sheets. Pseudoalveolar, pseudopapillary, microcystic, fascicular, and cord-like patterns, and rosette-like structures, are also reported. Mitotic activity is variable. Osteoclast-like giant cells are often numerous. Immunohistochemistry shows expression of S100 and SOX10. Other melanocytic markers (Melan A, HMB45) are in contrast to CSS commonly absent. Neural and neuroendocrine markers, including synaptophysin, NSE, and CD56, are inconsistently expressed. Rarely dot-like keratin-expression can be observed. Although there are some differences compared to CCS with regard to morphology and immunophenotype, lack of melanin pigment does not exclude CCS. Another important differential diagnosis regards melanoma. Although molecular alterations can often resolve this matter, not all melanomas harbor a BRAF mutation and not all clear-cell sarcoma-like tumors of the gastrointestinal tract have EWSR1 rearrangements.
These tumors show aggressive behavior with metastases to lymph nodes and the liver.
## Mesothelioma
In 2013 a (14;22)(q32;q12) translocation leading to a EWSR1-YY1 fusion was reported in two mesotheliomas, showing for the first time fusion genes in these neoplasms. In 2017, EWSR1/FUS-CREB fusions have been described in a subset of malignant mesotheliomas occurring mainly in young adults. However, the age range is broad, comprising patients from the childhood to the elderly. There is an equal sex distribution. The peritoneum seems to be mostly involved with pleural lesions less frequently reported. Extension into adjacent organs and structures and lymph node involvement are reported. Histologically, these lesions resemble AFH and in part CSS as described above under these headlines. Additionally, papillary projections, acinar, and tubular structures and psammomatous calcifications are reported as seen in classical mesotheliomas. The cells are epithelioid and histiocytoid with monomorphic round-to-oval nuclei and eosinophilic cytoplasm. The immunophenotype with keratin-and WT1 nuclear expression and absence of S100 differs from AFH and CSS. Overlapping positive immunohistochemical markers are EMA and desmin. Loss of BAP1 may occur in a minority of cases. Other differential diagnoses are listed in. EWSR1/FUS-ATF1/CREM are the described fusion genes showing the spectrum seen in other entities with EWSR1/FUS-CREB. Follow-up data show variable clinical presentation ranging from indolent to aggressive behavior.
## Myoepithelial tumors
Different from most other EWSR1-rearranged neoplasms, myoepithelial tumors have a normal counterpart, with myoepithelial cells being the outer layer of glands present in e.g., salivary glands, lung, skin adnexa, and mamma, but naturally not in soft tissue and bone.
The first myoepithelial tumor of soft tissue was published by Stout and Gorman in 1959 in a series of cutaneous lesions, and the first bone myoepithelioma was reported in 2001. Parachordoma is another term introduced 1977 in the English literature. Reports of cytogenetic analyses showed heterogeneous abnormalities.
Since EWSR1 rearrangement was mentioned in one myoepithelial carcinoma and one myoepithelioma of soft tissue 2007 and 2008; systematic analyses revealed that approximately 50% of myoepithelial tumors of skin, soft tissue, viscera, and bone harbor a EWSR1 fusion gene with a variety of gene partners, including PBX1, PBX3, ZNF 444, POU5F1, ATF1, and KLF17. EWSR1 seems to be rarely substituted by FUS. PLAG1 rearrangement and other genetic changes are alternatively observed.
The age range is broad from early childhood to the elderly. Extremities and limb girdles are most frequently involved, followed by the head and neck and trunk. Skin, subcutis, and deep soft tissue, including mediastinum and retroperitoneum, can be affected. Bone lesions most often arise in long tubular bones but also in small tubular bones and the axial skeleton, including iliac bone, sacrum, vertebra, ribs, skull, and jaw. Cortical destruction and extension into surrounding soft tissue may be present.
Macroscopically, tumors can be large with up to 20 cm. They are usually circumscribed and (multi)nodular. The cut surface often is white-grey in color with gelatinous areas. Calcification and ossification may be seen.
Microscopy is similar to salivary gland myoepithelial tumors showing a (multi)nodular appearance with well-circumscribed nodi/noduli variably infiltrating adjacent tissue. There is a broad spectrum in terms of architecture, cellularity, and cell composition. Growth patterns, often combined, are solid, nested, reticular, trabecular, cord-like, and glandular. Cells are epithelioid and/or spindled, having sometimes a clear cytoplasm, and/or plasmacytoid, and/or rhabdoid. Lesions called syncytial myoepitheliomas mainly occurring in skin show a sheet-like syncytial growth of ovoid to spindled or histiocytoid cells with pale eosinophilic cytoplasm. Criteria for malignancy were established in the largest series of soft tissue myoepithelial tumors, with tumors with benign cytomorphology or mild atypia (little variation in cell size, small relatively uniform nuclei, fine chromatin, inconspicuous nucleoli) diagnosed as myoepithelioma, whereas morderate to servere atypia (nuclear pleomorphism or vesicular or coarse chromatin, prominent nucleoli) represented features of myoepithelial carcinoma. Small round-cell morphology has been described in myoepithelial carcinomas. The matrix, variably present in myoepithelial tumors, can be (chondro)myxoid and/or collagenous/hyaline. Metaplastic cartilage or bone may occur. In myoepithelial carcinomas, malignant bone or cartilage can be observed. High mitotic rate and necrosis is reported in myoepithelial carcinomas. The immunohistochemical profile of myoepithelial tumors is variable showing per definition expression of broad-spectrum keratins and/or EMA and neuronal markers as S100, SOX10, and/or GFAP. P63 is positive in a subset of cases. Smooth muscle markers (SMA, calponin, and desmin) are possibly positive. INI1 is lost in a subset of myoepithelial carcinoma. MUC4 expression can be confusing when considering sclerosing epithelioid fibrosarcoma. Nuclear expression of brachyury, absent in myoepithelial tumors, distinguishes them from chordomas. Differential diagnoses are listed in.
## Serpine1/actb-fosb
Excision is the treatment of choice. Most of the lesions are superficially located with a benign morphology behaving indolent. Benign and malignant lesions have the potential for local recurrence. The metastatic rate of myoepithelial carcinoma is high with spread to lung, lymph nodes, bone, and soft tissue. Radiotherapy and chemotherapy are additional treatment options, but clinical effectiveness is variable.
## Low-grade fibromyxoid sarcoma (lgfms)/sclerosing epithelioid fibrosarcoma (sef)
These mostly deep situated sarcomas show overlapping features in terms of morphology, immunohistochemistry, genetic aberrations, and behavior. Therefore, it has been suggested that they form a spectrum of one entity.
LGFMS was firstly observed by Evans in 1987, and the first description of SEF was done by Meis-Kindblom eight years later. The chromosomal translocation, most typical for LGFMS, (t7;16)(q34;p11), has been described in 2003 by Reid et al. and the corresponding fusion gene FUS-CREB3L2 in the same year. Later on, it was shown that CREB3L1 is an alternative fusion partner of FUS and that FUS can be substituted by EWSR1. The genetic findings of LGFMS were also found in SEF and hybrid cases with predominance of EWSR1-CREBL3L1 in SEF. In one case, a PAX5-CREB3L1 was identified.
Both entities affect patients over a wide age range with a median in the 3th (LGFMS) and 4th (SEF) decade. Most often, these tumors occur in the deep soft tissue of the lower extremities, particularly thigh and limb girdles and the trunk. However, a wide variety of involved anatomic sites are reported, including intraabdominal, kidney, and bone.
Macroscopically, lesions are (multi)nodular with a grey-white whorled cut surface. Myxoid areas, if present, are visible. Infiltrative growth in adjacent structures can be seen.
Microscopically, LGFMS is characterized by alternating fibrous and myxoid areas with a whorled and bundled growth of uniform bland-looking slender fibroblastic spindle cells with elongated and tapered nuclei. A storiform, fascicular, and patternless architecture may be seen. Mitotic figures are sparse. There is scant cytoplasm. Typically, there are arcades of small blood vessels. In some cases, hyaline rosettes surrounded by round or oval cells are present. Such neoplasms were formerly called hyalinizing spindle cell tumor with giant rosettes. Cellular examples containing epithelioid cells show overlap with SEF and hybrid cases occur. A shift of the LGFMS pattern to SEF morphology is described in recurrences and metastases. Uncommonly noted are cell clusters, strands, palisades, and a retiform pattern. Thick collagen bundles are sometimes found in fibrotic areas. Nuclear pleomorphism and multinucleated giant cells are rarely observed and are mainly associated with recurrences and metastases. Cystic changes and osseous metaplasia may occur. SEF shows histomorphologically epithelioid/polygonal cells arranged in cords, nests, and sheets situated in a sclerotic stroma. Due to cellular shrinkage, a pseudovascular appearance can become apparent. The round-to-oval nuclei show at most slight pleomorphism and an open chromatin. There is a variable often low mitotic count. Chondro-osseous differentiation is exceptionally observed. The most reliable immunohistochemical marker is MUC4 with positivity in the majority of LGFMS, whereas SEFs are positive in around 70% of the cases. Other markers such as EMA, S100, CD34, SMA, and keratins (in SEF) are inconsistently expressed.
Differential diagnoses are listed in. Wide excision is the treatment of choice.
LGFMS typically shows a prolonged clinical course with recurrences and metastases. SEF seems to be more aggressive with much shorter survival; however, the outcome is variable. Wide spectrum of morphologies and histological grades; most often haphazardly arranged monomorphic spindle cells; variable stromal/perivascular hyalinization; and infiltrative growth into fat S100, CD34 (co-expression), and NTRK NTRK1-3 rearrangements (diverse fusion partners); RAF1 or BRAF fusions (rare)
## Extraskeletal myxoid chondrosarcoma (emc)
When initially described by Stout and Verner in 1953, it was thought that EMCs are true chondrosarcomas. The first large series delineating this tumor type more precisely was published by . In 1985, Hinrichs et al. reported for the first time the specific reciprocal translocation t(9;22)(q22;q11) leading to the most common fusion gene EWSR1-NR4A3, which was detected by. It seems that NR4A3 is necessarily involved. The described fusion partners besides EWSR1 are TAF15, TCF12, and TFG.
This in deep subcutis and soft tissue located sarcoma affects adults with a mean age of 50 years. Children are rarely involved. Males are slightly more often affected. The main sites are the proximal extremities and limb girdles followed by the distal extremities and trunk. Unusual sites are the head and neck area, including the intracranial cavity, the pelvic cavity/retroperitoneum, or intraabdominal and acral sites. Rarely bone lesions are also reported.
Macroscopically, tumors show a (multi)nodular configuration with relatively welldefined margins and variably a pseudocapsule. The cut-surface appears gelatinous with a tan color. Firm grey-white areas and hemorrhage may be seen.
Regarding microscopy, EMC is commonly a hypocellular lesion characterized by a multinodular growth pattern with presence of fibrous septa. The tumor nodules show peripheral accentuation of cellularity and are composed of bland-looking small round-tospindled cells with scanty eosinophilic cytoplasm set in a myxoid matrix resulting in a lace-like or reticular appearance. The nuclei are usually uniform, round-to-oval with an open chromatin or with hyperchromasia. There is low mitotic activity. Cellular areas lose their classical architecture owing to the limited myxoid matrix. They may be present in primary and recurrent lesions sometimes associated with pleomorphism of epithelioid, rhabdoid, and spindled cells. Immunohistochemistry is of little value depicting focal S100 reaction. GFAP, EMA, SMA, keratins, and p63 are expressed in a minority of cases with a focal staining pattern. Fusion gene analyses is especially helpful when classical features are less obvious.
The most important differential diagnosis is myoepithelial tumors of soft tissue. They show morphological, immunophenotypical (EMA/Keratins + and S100/SOX10/GFAP+), and genetic overlap with rearrangement of EWSR1 in a subset of cases. Fusion chimera involving NR4A3 are confirmatory for the diagnosis of EMC. For further differential diagnoses see.
Surgery is the treatment of choice. EMC shows a protracted clinical course with a high rate of recurrences and metastatic potential.
## Ewsr1-smad3-positive fibroblastic tumor (esft)
This recently defined tumor type was first described by . Few reports have followed since then.
Lesions usually present as a relatively small painless mass in the skin and superficial soft tissue of the extremities, mainly distal, especially the foot. Occurrence in bone is reported in one case localized in the tibia. There is a broad age range from the early childhood to the elderly and an obvious female preponderance.
Macroscopically, the neoplasms are nodular and firm, showing on cut surface a whitegrey solid appearance.
Histologically, tumors have a nodular configuration and consist of infiltrative growing intersecting long or short fascicles. The spindle cells possess uniform elongated nuclei with open chromatin. Mitotic activity is low. There is inconspicuous cytoplasm. Cellular areas merge with hyalinized areas showing sometimes calcifications. In some cases, a zonation pattern is seen often with central hyalinization. When located intradermal, an epidermal collarette may be present. Arrangement around blood vessels as seen in myopericytomas is sometimes observed. Myxoid and collagenous areas with the latter reminiscent of collagen rosettes are rarely reported. Immunohistochemically, the most reliable marker by now seems to be ERG demonstrating an homogeneous nuclear expression. Variable positive markers are SMA, keratins (both mostly week and focal), and SATB2. Reported negative stainings are EMA, desmin, S100, SOX10, CD34, CD31, MUC4, STAT6, TLE1, HMB45, and CD99. When a classical clinicopathologic constellation is present with expression of ERG, the diagnosis is straight forward. However fusion gene analysis may aid for the precise diagnosis, because benign and malignant lesions are in the differential diagnoses listed in. The clinical behavior appears to be benign, also when located in the bone, but lesions are prone to local recurrence even after 10 years.
## Epithelioid and spindle cell rhabdomyosarcoma with ewsr1/fus-tfcp2 fusion
These lesions were first described by . Hitherto-reported cases arose mainly in the bone and rarely in soft tissue, with predilection for the craniofacial bones. However, sites are heterogeneous, including also pelvis, femur, groin, and peritoneum. Intraosseous lesions show destruction of the cortex and expansion into soft tissue. The age range is broad, including pediatric patients and elderly patients. The average age is in the third decade. Males and females are affected (almost) equally with a slight female preponderance.
Macroscopically, a solid mass is reported.
Histologically, tumors consist of epithelioid and/or spindle cells. Whereas epithelioid cells are arranged in sheets, spindle cells show fascicular growth. The enlarged, relatively monomorphic round, oval, or elongated nuclei are vesicular with prominent nucleoli. The cytoplasm is scant or moderate, more or less intense eosinophilic, and can be rhabdoid in the epithelioid population. Round-cell morphology and uncommonly pleomorphism and hyperchromasia are reported. Real rhabdomyoblasts are not always present. There is a variable, sometimes prominent stromal reaction with sclerosing/hyalinized areas. Immunohistochemically, tumors were all positive with desmin and MYOD1 and to a much lesser degree with myogenin. ALK seems to be heterogeneously expressed in a large subset of cases, and broad-spectrum keratins are positive in almost all cases. S100 can be expressed without concomitant positivity for SOX10. Differential diagnoses are listed in. Fascicles of spindle cells with perivascular accentuation and alternating cellularity; staghorn vessels; necrosis; and heterologous differentiation S100, SOX10 (focal), and loss of H3K27me3
Inactivating mutations of NF1, CDKN2A/B, EED, and SUZ2
Most of the tumors behave extremely aggressively, with a reported median survival of 8 months. However, few patients with local disease and long-term follow-up showed no evidence of disease after treatment, with the mandible being a site of favorable prognosis.
## Retroperitoneal leiomyoma
The first cytogenetic analyzed retroperitoneal leiomyoma harbored a t(10;17)(q22;q21) translocation resulting in a KAT6B-KANSL1 fusion gene, and the second case was identified with a t(9;22)(q33;q12) leading to an EWSR1-PBX3 chimeric transcript. Both lesions occurred in woman 45 and 26 years old and showed a usual leiomyoma morphology, with leiomyocytes arranged in long fascicles and a classical immunoprofile with expression of smooth muscle markers. It is obvious that molecular heterogeneity may exist in these tumors.
## Simple (unicameral) bone cyst (sbc)
SBC was initially reported by Rudolf Virchow in 1876. It is a benign intramedullary cystic lesion involving the long bones in skeletally immature individuals. Boys are twice as more affected than girls. The reported peak is between the ages of 3 and 14 years. It commonly arises in the proximal humerus or proximal femur and less frequently in other long bones. Symptoms can be pain and swelling. A unilocular expansile cyst showing double-density fluid levels is radiologically characteristic.
Macroscopically, an often-fragmented thin fibrous membrane representing the cyst wall possibly with thickening and hemorrhage after trauma is seen (fracture). The fibrous pseudocystic structure is also microscopically obvious showing focal fibrin-like collagen with calcification and ossification. There is no true lining. Myofibroblastic cells and osteoclasts are not as prominent as typically seen in aneurysmal bone cyst (ABC), which is a differential diagnosis. Secondary changes as hemorrhage with resorption with chronic inflammation can be found after fracture. This can camouflage the classical microscopical features. In 2002, the translocation (16;20)(p11.2;q13) was identified as the sole cytogenetic abnormality in a SBC case. Recently, the corresponding FUS1-NFATC2 or alternatively EWSR1-NFATC2 fusion have been demonstrated in a subset of cases proving that SBC is a neoplasmBesides ABC, other differential diagnoses are cystic fibrous dysplasia, intraosseous ganglion, or lipoma. Conservative and surgical treatment are discussed in the literature. The recurrence rate is low but fractures are a known complication.
## Hemangioma of bone with an ewsr1-nfatc1 fusion
In 1982, Mulliken and Glowacki separated vascular anomalies into hemangiomas and vascular malformations based on their clinicopathological appearance and biological background. This was the base for the classification of the International Society for the Study of Vascular Anomalies (ISSVA). Vascular tumors arise by clonal cellular proliferation of vessels showing a disproportionate growth. In contrast, vascular malformations originate in utero as a result of mosaic mutations leading to erroneous development of vessels with proportionate growth.
Hemangiomas of the bone are relatively common and often incidental findings. They can be found at any age and arise often multifocal. The spine is the predilection site followed by the craniofacial bones. Radiologically, well-demarcated lucent lesions with coarse primary trabeculations are visible. When symptomatic, surgical intervention can be necessary and tissue will be obtained.
Macroscopically, the lesional tissue, often piecemeal, is soft and hemorrhagic with some bony fragments.
Microscopy is variable, showing thin and thick-walled vessels of different caliber. The endothelial lining is not remarkable. A stromal reaction can be prominent. A vascular malformation can be difficult to differentiate. Recently, a multifocal hemangioma has been described located in the occipital bone and clavicle showing a t(18;22)(q23;q12) with an EWSR1-NFATC1 fusion chimera.
Comparable lesions were found with EWSR1-NFATC1 or EWSR1-NFATC2 (own observation). Therefore, it seems to be a recurrent finding.
Prognosis is excellent.
# Conclusions
Several malignant and benign tumors harbor an EWSR1 rearrangement due to the central role of EWSR1 in different cell processes and vulnerability of the gene as consequence of frequent transcription. It was the first described fusion gene in sarcomas, and during the last decade(s) its promiscuous character has been shown by involvement in a variety of tumors. The recently described polyphenotypic sarcomas, which seem to be different entities with aggressive behavior, are interesting in this context. Whether they could be treated like Ewing sarcoma or require a more tailored approach is paramount to investigate.
Another intriguing point is that the same fusion partners are present in benign and malignant tumors, arguing that secondary genetic and epigenetic changes are mandatory to regulate and activate the required oncogenic pathways.
|
An Unusual Case of Parapharyngeal Desmoid Fibromatosis - A Rare Case Report
Desmoid fibromatosis (DF) is a rare locally aggressive, connective tissue malignancy developing in musculoaponeurotic tissues with an incidence of 2-4 per million population. We presented a case of a 3-year-old patient with a left parapharyngeal mass, histopathological examination suggesting DF, who underwent complete surgical excision without recurrence or requirement of cardiac resynchronization therapy.This is an open access journal, and articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as appropriate credit is given and the new creations are licensed under the identical terms.
heterochromatin. Studying the immunohistochemistry, DF stains are positive for nuclear B-catenin, vimentin, cyclooxygenase 2, tyrosine kinase PDGFRb, androgen receptor, and estrogen receptor but are negative for S-100, h-Caldesmon, CD34, and c-KIT. Multimodality imaging computed tomography (CT) and magnetic resonance are useful in the diagnosis, evaluation of treatment response, and surveillance of these tumors. The current initial treatment strategy for DT advocates for an active surveillance period. Active treatments in DT are multiple and include surgery, radiation therapy, systemic treatment, or a combination of these. [bib_ref] Desmoid-type fibromatosis, Garcia-Ortega [/bib_ref] cAse report year-old child presented with a swelling in the left side of the neck for 2 months. The swelling was initially small in size and slow-growing and painless. The patient had complaints of difficulty in swallowing and dyspnea in the supine position but no change of voice. On clinical examination, there was a 2 cm × 2 cm firm, nontender swelling in the left submandibular region externally with diffuse oropharyngeal swelling of approximately 3 cm × 2 cm in the soft palate on the left side. No lymph nodes were palpable in the neck [ [fig_ref] Figure 1: Clinical picture of the patient with left parapharyngeal mass with intraoperative and... [/fig_ref] ]. Cranial nerve function tests were normal, nasal endoscopy Case Report IntroductIon D esmoid fibromatosis (DF), also known as aggressive fibromatosis, deep fibromatosis, musculoaponeurotic fibromatosis, and desmoid tumor, is a locally aggressive, deep-seated connective tissue malignancy developing in the musculoaponeurotic tissues.DF is a rare tumor in western India, with a reported incidence of 2-4 per million population, and accounts for 0.03% of all neoplasms. Since DF lacks metastatic potential, it has now been classified as "intermediate, locally aggressive" tumor in the WHO Classification of Soft Tissue Tumors but has a high propensity for recurrence. Although a vast majority of these tumors are sporadic with a presentation in 15-60 years and being more common in females, with high proclivity to develop at sites of the surgical scar, DF may also be hereditary. Familial DF develops predominantly in patients with familial adenomatosis polyposis. β-catenin pathway is believed to play a key role in the pathogenesis of DF as intracellular levels of β-catenin are regulated by the adenomatous polyposis coli (APC) gene and the Wnt pathway with genetic alterations in the APC and CTNNB1 gene, resulting in the hereditary and sporadic DF. Histopathologically, DF is characterized by heterogeneous, poorly defined, and uniform proliferation of the spindle cells that mirror myofibroblasts wrapped within a stroma of abundant collagen and a vascular network lacking capsule.Consistently, there is no atypia, necrosis, or mitosis. The nuclei may contain euchromatin or showed no mass, but indirect laryngoscopy could not be performed. The origin of the mass could be suspected to be arising from the left parapharyngeal space clinically.
Fine needle aspiration cytology of the mass revealed proliferation of uniform spindle cells resembling myofibroblasts, in the background of abundant collagenous stroma. CT scan revealed a 36 mm × 50 × 60 mm lobulated hypodense mildly enhancing soft tissue mass in the left parapharyngeal space at the level of nasopharynx and oropharynx extending from pterygoid plates to hyoid bone with indentation of base of the tongue, with possible differentials being neurogenic or fibrous or mesenchymal tumor. Lesion could be seen abutting left side of the epiglottis and posterior most part of tongue, with loss of fat plane, left IJV, and ICA with focal loss of fat plane.
The patient underwent left parapharyngeal mass excision under general anesthesia. Elliptical incision was kept behind the posterior pillar of left tonsil to separate the tumor from parapharyngeal space. Dissection was carried out, and mass was dissected out from tonsillar bed. It was separated superiorly without surrounding tissue infiltration, but inferiorly, local infiltration was present in bed. Tumor was removed completely, and left tonsillectomy was done. The defect was closed primarily. Postoperative course was uneventful. The gross pathology report stated that the mass had lobulated irregular outer surface, with dimensions of 7.5 cm × 5 × 3 cm with surface membrane in certain areas and cut section of the mass showing yellowish areas of fat and hemorrhage. Microscopic examination revealed low-grade fibroblastic/myofibroblastic cells infiltrating skeletal and adipose tissue arranged in fascicles of the collagenous stroma with features, suggesting DF. The patient presented good esthetic and functional results, with no clinical and radiographic evidence of recurrence.
# Discussion
Aggressive fibromatoses are distinctive lesions that are defined as a group of nonmetastazing fibrous tumors which tend to invade locally developing from the tissue of the musculoaponeurotic system with an annual occurrence of 0.2-0.4 per 100,000, with origin in the head and neck accounting for 10%-25%. In the oral cavity, DF mostly presents as a painless mass or swelling in the involved region. Among the reported cases of pediatric aggressive fibromatosis of the head and neck, 74% have underwent large surgical resection as the primary treatment modality with 16% of these tumors reporting recurrence. Wang et al. (2014) after their extensive study proposed an age-based treatment of aggressive fibromatosis of the head and neck, in which they suggested that tumors which can be resected with negative margins without functional and cosmetic impairment should undergo surgical resection as a primary treatment modality. Conservative resection has to be done with chemotherapy in patients less than 18 years of age and radiotherapy in patients more than 18 years, if the surgical resection causes impairment. [bib_ref] Aggressive fibromatosis of the oral cavity in a 5 year old boy:..., Nair [/bib_ref] Our patient underwent complete surgical resection with negative margins without recurrence and is kept under active surveillance.
[fig] Figure 1: Clinical picture of the patient with left parapharyngeal mass with intraoperative and gross specimen of tumor [/fig]
[fig] Figure 2: Lobulated hypodense mildly enhancing soft tissue mass in the left parapharyngeal space [/fig]
|
Clinical implications of the quantitative detection of ID4 gene methylation in myelodysplastic syndrome
Background: Myelodysplastic syndrome (MDS) eventually transforms into acute leukemia (AL) in about 30% of patients. Hypermethylation of the inhibitor of DNA binding 4 (ID4) gene may play an important role in the initiation and development of MDS and AL. The aim of this study was to quantitatively assess ID4 gene methylation in MDS and to establish if it could be an effective method of evaluating MDS disease progression.Methods: We examined 142 bone marrow samples from MDS patients, healthy donors and MDS-AL patients using bisulfite sequencing PCR and quantitative real-time methylation-specific PCR. The ID4 methylation rates and levels were assessed.Results: ID4 methylation occurred in 27 patients (27/100). ID4 gene methylation was more frequent and at higher levels in patients with advanced disease stages and in high-risk subgroups according to WHO (P < 0.001, P < 0.001, respectively) and International Prognostic Scoring System (IPSS) (P = 0.002, P = 0.007, respectively) classifications. ID4 methylation levels changed during disease progression. Both methylation rates and methylation levels were significantly different between healthy donor, MDS patients and patients with MDS-AL (P < 0.001, P < 0.001, respectively). Multivariate analysis indicated that the level of ID4 methylation was an independent factor influencing overall survival. Patients with MDS showed decreased survival time with increased ID4 methylation levels (P = 0.011, hazard ratio (HR) = 2.371). Patients with ID4 methylation had shorter survival time than those without ID4 methylation (P = 0.008).Conclusions: Our findings suggest that ID4 gene methylation might be a new biomarker for MDS monitoring and the detection of minimal residual disease.
# Background
Myelodysplastic syndrome (MDS) is a heterogeneous group of clonal stem cell disorders characterized by abnormal differentiation and maturation of myeloid cells. The natural history of these syndromes may range from a chronic course, spanning many years, to a rapid course of leukemic progression, and 30% of cases of MDS eventually transform into acute leukemia (AL) [bib_ref] Cytogenetics in myelodysplastic syndromes, Ohyashiki [/bib_ref]. In myelodysplasia, malignant transformation at the level of a myeloid stem cell can result in chromosomal abnormalities that act as signatures of the disease and can be identified in precursor cells giving rise to granulocytes, monocytes, red cells, and platelets [bib_ref] Hypomethylating agents and other novel strategies in myelodysplastic syndromes, Garcia-Manero [/bib_ref]. For the development of the leukemic clone, further genetic events are required for the rapid expansion of leukemic blasts [bib_ref] Cytogenetic features in myelodysplastic syndromes, Haase [/bib_ref]. Many people regard MDS as the first stage of a hematological myeloid malignancy and consider MDS as the best intermediate stage research model for AL.
MDS pathogenesis and its frequent progression to AL are associated with aberrant cell clones with little ability for differentiation but with a high ability to proliferate [bib_ref] Apoptosis and antiapoptotic mechanisms in the progression of myelodysplastic syndrome, Kerbauy [/bib_ref] [bib_ref] Genetic abnormalities as targets for molecular therapies in myelodysplastic syndromes, Cilloni [/bib_ref]. This aberration of MDS is believed to be a multistep process of aberrant expression of the tumor suppressor gene (TSG) requiring both aberrant genetic and aberrant epigenetic factors such as abnormal hypermethylation of the gene promoter region [bib_ref] Cytogenetic features in myelodysplastic syndromes, Haase [/bib_ref] [bib_ref] Molecular pathways in myelodysplastic syndromes and acute myeloid leukemia: relationships and distinctions..., Bernasconi [/bib_ref] [bib_ref] Update on genetic and molecular markers associated with myelodysplastic syndromes, Valent [/bib_ref] [bib_ref] Current status of epigenetic treatment in myelodysplastic syndromes, Kuendgen [/bib_ref] [bib_ref] Epigenetic changes in the myelodysplastic syndrome, Issa [/bib_ref] [bib_ref] Myelodysplastic syndromes: an update on molecular pathology, Tormo [/bib_ref]. Previous studies have demonstrated that abnormal hypermethylation of many TSG promoter regions such as CpG islands are associated with MDS development and progression [bib_ref] Epigenetic inactivation of tumour suppressor gene KLF11 in myelodysplastic syndromes*, Potapova [/bib_ref] [bib_ref] DNA methylation predicts survival and response to therapy in patients with myelodysplastic..., Shen [/bib_ref] [bib_ref] Progressive chromatin repression and promoter methylation of CTNNA1 associated with advanced myeloid..., Ye [/bib_ref] [bib_ref] Aberrant DNA methylation is a dominant mechanism in MDS progression to AML, Jiang [/bib_ref] [bib_ref] Methylation status of fragile histidine triad (FHIT) gene and its clinical impact..., Lin [/bib_ref] [bib_ref] Promoter hypermethylation analysis in myelodysplastic syndromes: diagnostic & prognostic implication, Solomon [/bib_ref]. MDS patients treated with hypomethylating agents such as 5-azacytidine and decitabine achieved high overall response rates and hematological improvement rates [bib_ref] Hypomethylating agents and other novel strategies in myelodysplastic syndromes, Garcia-Manero [/bib_ref] [bib_ref] Decitabine in the treatment of myelodysplastic syndromes, Santos [/bib_ref] [bib_ref] Targeting DNA methylation for epigenetic therapy, Yang [/bib_ref] [bib_ref] DNA methylation changes following 5-azacitidine treatment in patients with myelodysplastic syndrome, Tran [/bib_ref]. In addition, it is well known that with the development of MDS, there are increases in abnormal blasts in bone marrow, which may be the result of abnormal expression of TSG in terms of cell cycle, cell differentiation, and cell proliferation.
To date, the role of aberrant promoter methylation in MDS has not been investigated exhaustively, and previous studies have mainly focused on cycling-dependent kinase inhibitors even if other genes have been found to be targeted by aberrant hypermethylation in MDS. For example, hypermethylation of the inhibitor of DNA binding 4 (ID4) gene may play an important role in hematological [bib_ref] Global assessment of promoter methylation in a mouse model of cancer identifies..., Yu [/bib_ref] and solid cancers [bib_ref] Inhibitor of differentiation 4 (Id4) is a potential tumor suppressor in prostate..., Carey [/bib_ref]. ID4 contributes to cell differentiation and proliferation [bib_ref] Inhibitor of differentiation 4 (Id4) is a potential tumor suppressor in prostate..., Carey [/bib_ref] [bib_ref] ID4: a new player in the cancer arena, Dell'orso [/bib_ref] [bib_ref] Silencing of the inhibitor of DNA binding protein 4 (ID4) contributes to..., Chen [/bib_ref] [bib_ref] The execution of the transcriptional axis mutant p53, E2F1 and ID4 promotes..., Fontemaggi [/bib_ref] [bib_ref] Inhibitor of differentiation 4 drives brain tumor-initiating cell genesis through cyclin E..., Jeon [/bib_ref] , and abnormal methylation of the ID4 gene in hematological cancers plays an important role in diagnosis and prognosis [bib_ref] Silencing of the inhibitor of DNA binding protein 4 (ID4) contributes to..., Chen [/bib_ref] [bib_ref] Differential methylation pattern of ID4, SFRP1, and SHP1 between acute myeloid leukemia..., Uhm [/bib_ref]. Aberrant methylation of the ID4 promoter region is associated with invasiveness, growth, and recurrence of many solid tumors [bib_ref] Silencing of the inhibitor of DNA binding protein 4 (ID4) contributes to..., Chen [/bib_ref] [bib_ref] Differential methylation pattern of ID4, SFRP1, and SHP1 between acute myeloid leukemia..., Uhm [/bib_ref] [bib_ref] Multiplexed methylation profiles of tumor suppressor genes and clinical outcome in lung..., Castro [/bib_ref] [bib_ref] Promoter methylation-associated loss of ID4 expression is a marker of tumour recurrence..., Noetzel [/bib_ref] [bib_ref] Aberrant hypermethylation of ID4 gene promoter region increases risk of lymph node..., Umetani [/bib_ref] [bib_ref] Frequent DNA methylation but not mutation of the ID4 gene in malignant..., Hagiwara [/bib_ref] and with initiation and progression in hematological cancers including leukemia and lymphoma [bib_ref] Global assessment of promoter methylation in a mouse model of cancer identifies..., Yu [/bib_ref] [bib_ref] Silencing of the inhibitor of DNA binding protein 4 (ID4) contributes to..., Chen [/bib_ref] [bib_ref] Differential methylation pattern of ID4, SFRP1, and SHP1 between acute myeloid leukemia..., Uhm [/bib_ref] [bib_ref] ID4 methylation predicts high risk of leukemic transformation in patients with myelodysplastic..., Wang [/bib_ref]. However, the methylation status of the ID4 gene in MDS patients and its clinical significance are poorly described.
Due to the high frequency of DNA methylation in MDS and its involvement in leukemia transformation, we hypothesized that the analysis of DNA methylation levels is a potential predictor for prognosis. This was achieved using a rapid and quantitative methylation detection methodology based on real-time PCR. The aim of the present study was to investigate the clinical impact of ID4 gene methylation on MDS prognosis in terms of methylation rate and methylation levels of the ID4 gene.
# Methods
## Subjects for study
We selected 142 bone marrow samples harvested from the ilium by osteostixis from individuals who visited the Chinese PLA General Hospital between May 2007 and June 2010. Samples for bisulfite sequencing study corresponded to bone marrow specimens from a healthy donor (age: 45 years, gender: male), bone marrow from a patient with MDS who was diagnosed with refractory anemia (MDS-RA) (age: 40 years, gender: male) and bone marrow from a patient with acute myeloid leukemia (MDS-AML) (M6) (age: 42 years, gender: male). Both patients with MDS and AML were diagnosed according to the World Health Organization (WHO) criteria. The prognostic score for each patient was calculated using the International Prognostic Scoring System (IPSS). For methylight analysis, we selected bone marrow samples from 100 patients with MDS, 20 healthy donors and 22 patients with MDS-AML from our institution. Ten samples in each group were randomly selected to assess the correlation between bisulfite sequencing PCR (BSP) and methylight; methylight was performed only in patients with high BSP results because of the detection threshold of methylight. Patients' characteristics are shown in [fig_ref] Table 1: Clinical characteristics of the study groups [/fig_ref]. Normal bone marrow samples (n = 20) were obtained for negative controls. In addition, for further positive and negative controls, we included malignant hematological cell lines including NB4 (acute leukemia cell line) and 293 (renal cell line) obtained from the American type culture collection (ATCC).
The study was approved by the local internal review boards and ethics committees. Written informed consent was obtained from all subjects.
## Dna extraction
DNA was extracted using a Genomic DNA extraction kit (Promega, Madison, WI, USA). The ID4 gene DNA sequence was obtained from the National Center for Biotechnology Information (NCBI).
## Bisulfite sequencing pcr
Sodium bisulfite treatment was provided to 1 μg of each DNA sample using the Epi bisulfite kit (Qiagen, Venlo, Limburg, Netherlands) according to the manufacturer's instructions. Bisulfite-treated genomic DNA was altered so that all unmethylated CpG sites were converted to TG, and methylated CpG sites remained CG. The methylated or unmethylated status of the promoter region of the ID4 gene was determined by PCR amplification of the bisulfite-treated genomic DNA using specific PCR primers and [fig_ref] Figure 1: Methylation frequency of CpG sites and association between the ID4 methylation percentage [/fig_ref].
The PCR conditions were 95°C for 15 min, followed by 40 cycles of 50 s at 94°C, 45 s at 53°C, and 60 s at 72°C
, and a final extension of 72°C for 10 min. The PCR reaction mixture contained 2 μl of bisulfite-modified DNA, 1 μl of 25 mM dNTPs, 2.5 μl of 10× Herman's buffer, 1.5 μl of 10 pmol/μl of each primer, and 0.5 μl of Hotstar Taq DNA Polymerase (Invitrogen, Carlsbad, CA, USA), for a final volume of 25 μl.
PCR products were analyzed on a 1.5% agarose gel, and then were purified using a QIAquick Gel Extraction Kit (Qiagen) and transformed into Escherichia coli cells using the pGEM-T easy vector (Promega). Ten individual clones were selected for vector extraction by QIAprep Spin Mini Prep Kit (Qiagen). Vector DNA samples were sequenced by Invitrogen.
## Methylight
After bisulfite treatment and purification, the promoter of the ID4 gene was evaluated by methylight. The primers and the probe for methylight were located in the BSP-amplified sequences. The primers and the ID4 gene probe had 10 CpG sites. The primers and the internal control gene MYOD1 probe had no CpG sites [bib_ref] MethyLight: a high-throughput assay to measure DNA methylation, Eads [/bib_ref]. Both the ID4 gene and MYOD1 gene probes were labeled with 6-fluorescein amidite (FAM) at the 5′ end and the quencher tetramethylrhodamine (TAMRA) at the 3′ end. All primers and probes were synthesized by Invitrogen. The primers and probes used for PCR are shown in .
The 20-μl reaction mixture included 2.5 μmol of each primer, 1.25 μmol of probe, 2 μl of bisulfite-converted DNA, and 10 μl of mix buffer (Qiagen). Conditions were as follows: 95°C denaturation for 10 min, followed by 45 cycles at 95°C denaturation for 30 s, and 57°C annealing and extension for 1 min. Methylight PCR reactions were performed in an MX3000p device (Stratagene, La Jolla, CA, USA). To control the accuracy of the methylight reactions, we used DNA from NB4 cell lines previously known to be methylated and DNA from 293 cell lines previously known to be unmethylated for the genes being analyzed.
The methylation levels of the ID4 gene were calculated individually using the relative quantification method. The standard sample was made of bisulfite-treated DNA from the NB4 cell line. A standard curve was produced for ID4 and MYOD1 (as the internal control gene) [bib_ref] MethyLight: a high-throughput assay to measure DNA methylation, Eads [/bib_ref] by a 10-fold dilution series of four different plasmid concentrations. Every run had a standard curve. The positive samples had amplification curves, but negative samples had no amplification curves.
# Statistical analysis
Statistical analysis was performed using SPSS 18.0 for Windows (IBM, Armonk, NY, USA). We used the chisquare test or the Fisher's exact test to evaluate categorical data. The Mann-Whitney non-parametric test was used for comparisons of continuous data.
The Spearman's rank correlation was used for correlation of methylation levels and bone marrow blast levels.
## Table 2 sequences of the primers and probes
Forward Reverse
[formula] BSP ID4 primer 5′-GTTTGATTGGTTGGTTATTTTAGAT-3′ 5′-ACCGAAAAAAAAATAACCCAC-3′ 5′-CACCAAAAAAAAAATAACCCAC-3′ Methylight PCR MYOD1 primer 5′-CCAACTCCAAATCCCCTCTCTAT-3′ 5′-TGATTAATTTAGATTGGGTTTAGAGAAGGA-3′ MYOD1 probe 5′-TCCCTTCCTATTCCTAAATCCAACCTAAATACCTCC-3′ ID4 primer 5′-TCGGAGTTTTCGTTTTCGTT-3′ 5′-CGATACTACTCACAACCGCG-3′ ID4 probe 5′-AGCGGGTTTCGTTCGGTTCG-3′ [/formula]
ID4, inhibitor of DNA binding 4; BSP, bisulfite sequencing PCR.
Overall survival (OS) was calculated from diagnosis day until the day of death or the last follow-up date (censored on 31 December 2012). OS was analyzed according to the Kaplan-Meier method and the log-rank test. The Cox proportional hazard regression model was used for the adjustment of independent prognostic factors in multivariate survival analysis. Two-tailed P values ≤ 0.05 were considered statistically significant.
# Results
## Detection of cpg methylation frequency by bsp
We detected the methylation status of bone marrow samples from one healthy donor, one MDS patient, and one AML patient using BSP. Ten clones from every bone marrow sample were chosen. [fig_ref] Table 3: ID4 methylation status of NBM, MDS, and AML [/fig_ref]. There was an association between the ID4 methylation percentage using BSP and methylight. Using linear regression, we obtained a goodness of fit of r 2 = 0.7168. Using an exponential curve, the goodness of fit was r 2 = 0.8485 [fig_ref] Figure 1: Methylation frequency of CpG sites and association between the ID4 methylation percentage [/fig_ref].
## Detection of id4 gene methylation status by methylight and correlation with clinical characteristics
As summarized in [fig_ref] Table 4: Correlation between clinical characteristics and ID4 gene methylation [/fig_ref] , 27 patients showed ID4 gene hypermethylation. Patients with ID4 gene methylation had higher levels of white blood cell (WBC) counts and bone marrow blast counts than those without (P = 0.036, P = 0.001). ID4 methylation levels were correlated with bone marrow blast counts in patients with MDS (r = 0.388, P < 0.001, [fig_ref] Figure 2: Methylation level of the ID4 gene according to the percentage of bone... [/fig_ref]. There were no significant differences in other clinical features, such as age, sex, initial hemoglobin levels, and platelet counts, between the MDS patients with and without ID4 gene hypermethylation (P > 0.05).
## Association between id4 gene methylation status (by methylight) and mds subtypes
As summarized in [fig_ref] Table 5: Association between ID4 gene methylation status and MDS subtypes [/fig_ref] , those who harbored ID4 gene hypermethylation according to WHO subtypes included four (9.76%) of the 41 patients with RA, four (18.18%) of the 22 patients with refractory cytopenia with multilineage dysplasia (RCMD), one (14.29%) of the 7 patients with refractory anemia with ringed sideroblasts (RARS), eleven (52.38%) of the 21 patients with refractory anemia with excess blasts (RAEB)-1, and seven (77.78%) of the 9 patients with RAEB-2. There was a significant difference in the rates of ID4 gene methylation between the different WHO subtypes (P < 0.001). MDS patients with highrisk subtypes (RAEB-1 and RAEB-2) had a significantly higher incidence of ID4 gene methylation (60%) than those with low-risk subtype (RA, RCMD, and RARS, 12.86%, P < 0.001). Another risk stratification, the IPSS (which weights the effects of blast quantity, karyotype, and cytopenia), also showed a different methylation positivity rate between the low-risk group (8.33%, 2/24 patients with low and intermediate risk (INT) 1) and the high-risk group (63.64%, 7/11 patients with INT 2 and high) (P = 0.002). As summarized in [fig_ref] Table 5: Association between ID4 gene methylation status and MDS subtypes [/fig_ref] , ID4 gene methylation levels was used to evaluate the WHO subtypes (P < 0.001), WHO risk stratification (P < 0.001), and IPSS risk stratification (P = 0.007). However, both the ID4 gene methylation positivity rates and methylation levels of the MDS patients with different karyotypes showed no significant differences (P = 1.000, P = 1.000).
# Prognostic factors analysis
A multivariate Cox survival analysis of age, sex, hemoglobin levels, platelet (PLT) counts, WBC counts, bone marrow blast levels, subtypes by IPSS and ID4 methylation level on OS was performed for the MDS patients [fig_ref] Table 6: Multivariate analysis of prognostic factors for overall survival [/fig_ref]. Results showed that sex, age, IPSS risk group, and ID4 methylation levels were significant independent risk factors of poor OS in patients with MDS. IPSS risk group was the strongest factor on OS (HR = 7.374).
The estimated median survival time of all patients was 39.0 months. Using the IPSS risk groups, patients with a higher risk had a shorter median survival (51.4 months for the low risk group, 42.1 months for the INT 1 risk group, 35.2 months for the INT 2 risk group, and 22.0 months for the high risk group).
ID4 methylation level was an independent factor affecting OS. Patients with MDS showed poor survival with increased ID4 methylation levels (P = 0.011, HR = 2.371). According to the multivariate analysis, age and sex were significant predictors of OS (P = 0.007, HR = 6.277 and P < 0.001, HR = 4.119, respectively), whereas marrow blast levels did not show statistical significance (P = 0.245). Patients with ID4 methylation had shorter survival time than those who had unmethylated ID4 (P = 0.008, [fig_ref] Figure 3: Overall survival of MDS patient groups for ID4 gene with and without... [/fig_ref]. There was a significant difference in the median survival between the methylated ID4 group and the unmethylated group (30.2 and 44.3 months, respectively).
## Sequential analysis of id4 gene methylation status during mds evolution
For one patient, bone marrow samples were available over five stages. With a follow-up time of 20 months, this patient underwent a disease progression from MDS-RA and MDS-RAEB to AL and transplantation. ID4 methylation levels of MDS-RA, MDS-RAEB, AL, and days 37 to 200 after transplantation were 0.45, 0.86, 1.47, and 0, respectively. ID4 gene methylation levels changed with the number of bone marrow blasts. Four additional patients with MDS with sequential samples were in the same situation: ID4 gene hypermethylation reflected disease stage and became negative after transplantation or demethylation treatment, and this was associated with bone marrow blasts [fig_ref] Figure 4: Relationship between ID4 methylation levels and BM blasts during disease status in... [/fig_ref].
# Discussion
The aim of this study was to investigate the biological significance and the prognostic relevance of the ID4 gene methylation status and methylation levels in MDS patients. Our results, which used for the first time quantitative real-time methylation-specific PCR (methylight) for ID4 gene methylation detection, indicated that aberrant methylation of the ID4 gene may play an important role in the prognosis of MDS. Firstly, we detected the methylation of the ID4 gene promoter region CpG sites using BSP. There was a sharp increase in methylation positivity rates from NBM, MDS to AL, which corresponded with the methylight result. As the intermediate stages of AL, MDS had moderate methylation levels of the ID4 gene, intermediate to those for NBM and AL. This is a clear indication that ID4 methylation detection could be used to monitor MDS progression.
In this study, 27 patients showed aberrant ID4 gene methylation, which was present in each WHO subtype of MDS, with increasing amounts in ascending order of RA, RCMD, RARS, RAEB-1, and RAEB-2. Methylation occurred more frequently in both high-risk WHO and IPSS subtypes of MDS. Wang et al. reported that ID4 methylation was present in 35.4% of adult patients newly diagnosed with MDS and occurred more frequently in advanced-stage MDS by methylation-specific PCR (MSP) [bib_ref] ID4 methylation predicts high risk of leukemic transformation in patients with myelodysplastic..., Wang [/bib_ref]. These discrepancies may be the result of the different methods used. Besides the methylation positivity rate, methylation levels were also used to evaluate the relationship with the WHO subtypes and IPSS. ID4 gene hypermethylation occurred in more advanced stage of MDS samples, as well as with higher methylation levels. As an Many studies have reported that more bone marrow blasts in patients with MDS provide more opportunities for leukemia transplantation [bib_ref] Prognostic factors and life expectancy in myelodysplastic syndromes classified according to WHO..., Malcovati [/bib_ref]. Previously, it was shown that both bone marrow blasts and ID4 methylation were independent factors for leukemia-free survival (LFS) estimation in MDS patients [bib_ref] ID4 methylation predicts high risk of leukemic transformation in patients with myelodysplastic..., Wang [/bib_ref]. In our study, not only ID4 methylation levels had high correlation with bone marrow blasts, but there was also a significant difference between the ID4 methylated and unmethylated groups, in terms of bone marrow blasts. This hinted that ID4 methylation levels may be a substitute for the degree of dysplastic cells for interpreting pathogenetic conditions. After treatment by either transplantation or demethylation, ID4 gene methylation in all sequential samples was negative, corresponding to decreased bone marrow blasts. To some extent, ID4 gene methylation may be a potential new marker for minimal residual disease (MRD) detection. In addition, the detection of ID4 gene methylation could offer some guidance to MDS treatment stratification. In our multivariate analysis, IPSS risk groups was an independent factor that affected the OS of patients with MDS. This finding was consistent with their clinical state. Our investigations indicated that the ID4 methylation levels were higher in advanced-stage MDS, indicating that ID4 methylation might be a molecular event playing an important role in the progression of MDS. In our study, ID4 methylation levels were considered to affect the OS of MDS patients. Meanwhile, patients with ID4 methylation had a shorter survival time. These results hinted that ID4 methylation levels may be an independent prognostic factor for overall survival.
Our results revealed that while there was no significant correlation between ID4 gene methylation and genetic abnormalities, taken together they may offer greater potential as biomarkers for MDS. As previously reported, the occurrence and progression of cancers includes both genetic and epigenetic alterations [bib_ref] Myelodysplastic syndromes, Orazi [/bib_ref] [bib_ref] Myelodysplastic syndromes, Scott [/bib_ref] , and for epigenetic alterations, ID4 gene methylation could be used for auxiliary diagnosis and prognosis of MDS patients without significant genetic alterations. Combining these two mechanisms could cover more scope for disease monitoring, treatment effect evaluation and MRD detection.
The present study investigated the status and levels of ID4 gene methylation in seven patients with MDS at different stages of MDS to directly interpret the close correlation of disease evolution with ID4 gene methylation for the first time. In terms of disease progression, ID4 gene methylation status of one sequential MDS patient sharply increased from MDS-RA and MDS-RAEB to AML. After transplantation, ID4 methylation was negative, supporting the role of ID4 gene methylation in MRD progression.
It is noteworthy that drugs with demethylating activity are increasingly used and studied in hematological malignancies [bib_ref] Phase 1 study of low-dose prolonged exposure schedules of the hypomethylating agent..., Issa [/bib_ref] [bib_ref] Characterization of DNA demethylation effects induced by 5-Aza-2′-deoxycytidine in patients with myelodysplastic..., Mund [/bib_ref] [bib_ref] Sensitization by 5-aza-2′-deoxycytidine of leukaemia cells with MLL abnormalities to induction of..., Niitsu [/bib_ref]. One could speculate that the presence of increased DNA methylation levels in MDS in clinical remission might provide a rationale for the use of these drugs in affected patients. Further clinical studies are necessary to clarify this possibility and to define the accuracy of methylation MRD estimation.
# Conclusions
Our findings provide evidence that aberrant DNA methylation is present in MDS patients and can potentially be used for prognosis. These findings may offer the possibility of prognostic estimation for many patients who currently lack adequate markers.
[fig] Figure 1: Methylation frequency of CpG sites and association between the ID4 methylation percentage. (A) Methylation frequency of CpG sites in the ID4 gene in NBM, MDS and MDS-AL. Filled circle and empty circle represent one methylated CpG site (CG) and unmethylated CpG site (TG), respectively. P < 0.001. (B) Association between the ID4 methylation percentage using bisulfite sequencing PCR and methylight. Using linear regression, we obtained a goodness of fit of r 2 = 0.7168. Using an exponential curve, the goodness of fit was r 2 = 0.8485. AL, acute leukemia; MDS, myelodysplastic syndrome; NBM, normal bone marrow; TSS, transcriptional start site. [/fig]
[fig] Figure 2: Methylation level of the ID4 gene according to the percentage of bone marrow blasts. ID4 gene methylation levels increase with the number of bone marrow blasts (r = 0.388, P < 0.001). ID4, inhibitor of DNA binding 4. epigenetic abnormality, high ID4 gene hypermethylation may increase the risk of disease progression in patients with MDS patients. [/fig]
[fig] Figure 4: Relationship between ID4 methylation levels and BM blasts during disease status in five MDS patient samples at different disease stages. ID4 gene methylation levels increased or decreased with the number of bone marrow blasts. (A) ID4 methylation level of one patient during different disease stages MDS-RA, MDS-RAEB, AML, day 37 to day 200 after transplantation. (B) ID4 methylation level of four patients in different disease stages. (C) BM blasts of four patients in different disease stages. Stage 1: new diagnosis; stage 2: the first period of demethylation treatment for patient 2 and patient 3 and 1 month after transplantation for patient 4 and patient 5; stage 3: the second period of demethylation treatment for patient 2 and patient 3 and 2 months after transplantation for patient 4 and patient 5. AML, acute myeloid leukemia; ID4, inhibitor of DNA binding 4; MDS-RA, MDS with refractory anemia; MDS-RAEB, MDS with refractory anemia with excess blasts. [/fig]
[fig] Figure 3: Overall survival of MDS patient groups for ID4 gene with and without methylation. Patients with ID4 methylation had a shorter survival than those without ID4 methylation (P = 0.008). ID4, inhibitor of DNA binding 4. [/fig]
[table] Table 1: Clinical characteristics of the study groups [/table]
[table] Table 3: ID4 methylation status of NBM, MDS, and AML [/table]
[table] Table 4: Correlation between clinical characteristics and ID4 gene methylation [/table]
[table] Table 6: Multivariate analysis of prognostic factors for overall survival [/table]
[table] Table 5: Association between ID4 gene methylation status and MDS subtypes [/table]
|
Recycling of a selectable marker with a self-excisable plasmid in Pichia pastoris
## Primer sequences (5'-3') or a short description of the plasmids endonucleases
Cre-F GCTGAATTCATGTCCAATTTACTGACCG EcoRI For secreted expression of Arl using 5 ARL expression cassettes after zeocin-resistance marker excision GS115/pZACH-(phy) 6 GS115/P-6c For secreted expression of Phy using 6 PHY expression cassettes after zeocin-resistance marker excision pZACH-phy/pZACH-(phy) 6 C-Phy/P-6c For secreted expression of Phy using 7 PHY expression cassettes after zeocin-resistance marker excision pZACH-phy/pZACH-(phy) 6 /pZACH-(phy) 6 C-Phy/P-6c/P-6c For secreted expression of Phy using 13 PHY expression cassettes after zeocin-resistance marker excision pZACH-(phy) 6 /pGACH P-6c/GH Control pZACH-(phy) 6 /pGACH-SLY1 P-6c/SLY1 For secreted expression of phytase using 6 PHY expression cassettes and overexpression of Sly1p after zeocin-resistance marker excision pZACH-(phy) 6 /pGACH-SEC1 P-6c/SEC1 For secreted expression of phytase using 6 PHY expression cassettes and overexpression of Sec1p after zeocin-resistance marker excision pZACH-(phy) 6 /pGACH-SLY1/pGACH P-6c/SLY1/GH Control pZACH-(phy) 6 /pGACH-SLY1/pGACH-SEC1 P-6c/SLY1/SEC1 For secreted expression of phytase using 6 PHY expression cassettes and overexpression of Sly1p and Sec1p after zeocin-resistance marker excision GS115/C-Arl 0.093±0.014 . Schematic of strategy for zeocin-resistance marker excision.
[formula] Cre-G357C-R CTTTTCGGATGCGCCGCATAACCAG Cre-G357C-F CTGGTTATGCGGCGCATCCGAAAAG Cre-R CATGGGCCCCTAATCGCCATCTTCCAGC ApaI AOX1-lox71-F CAATACCGTTCGTATAGCATACATTATACGAAGTTATTAACATCCAAAGACGAAAGGTTGAATG AOXTT-A TGGGGGTTCCGCACAAACGAAGGTC AOXTT-F TTTGTGCGGAACCCCCACACACCATAG Zeo_lox66_A GATCTTACCGTTCGTATAATGTATGCTATACGAAGTTATGATCTCATGACCAAAATCCC AOX1-G-F ATACGAACGGTAAGATCTAACATCCAAAGACG GAP-G-F ATACGAACGGTAAGATCTTTTTTGTAGAAATG AOXTT-G-A CTATACGAACGGTATTGAAGCTATGGTGTGTGGG HIS4-F ATGCTATACGAACGGTATTAAATAAGTCCCAGTTTCT HIS4-R TTCGTTTGTGCGGATCCATGACATTTCCCTTGCTACC 3AOX-F TACCGTTCGTATAGCATACA 3AOX-R GGATCCGCACAAACGAAGGT SLY1-F CGTGGTACCATGCTTCATTTGAATG PmlI SLY1-R GATCCGCGGTTACTTATCGTCATCATCCTTGTAATCTTTTGCTTCGGCA SacII SEC1-F CGTGGTACCATGTACCCATACGATGTTCCAGATTACGCTGCTTCTGATCTGATTAA PmlI SEC1-R GATCCGCGGCTATTTCCAAAATTTCTTCA SacII RT-G1 GTCGGGACACGCCTGAAACT RT-G2 CCACCTTTTGGACCCTATTGAC RT-Phy1 TGGTTGGGGTAGAATCAC RT-Phy2 TGCTTCTGAGGAGGATGA Prtx1 AGGTGCTGATAAGATTGT Prtx2 TTGCTTGTTAGCCTTT PrtA1 GAGAAATACCGCTCCTAC PrtA2 GCAACATACTTCCACCAA P1 TCAGATCTCCTGATGACTGA P2 TTTCTTTTTGACCAACTGGC P3 GTACCAACCACACTTTGCCA P4 CGATCAGCTCCTCAAATTGG Phy-S GAGGTTCCAGACGATATGAAATTG Phy-A TCTATTAACGTCGGCCA RT-AOX1-1 GAGAGTTCTTCTGGTGTTGG RT-AOX1-2 GAATCAAGCCTCAGTACGAG RT-DAK2-1 GGCTGGAACTTCTTTAGTGC RT-DAK2-2 TGTTTGCTCTGGCTGGAATG RT-DAS1-1 TTGGTTCGGTAGGCTTGTCT RT-DAS1-2 CCAGCTAAAGGTGACGAGTT RT-FBA1-2-1 GCACCTTCTTATGGCATACC RT-FBA1-2-2 TTCATCGTCTGTTTCCTCCG [/formula]
The Zeo R transformants were shifted from YPDSZ plates to YPM. After methanol induction, YPM cultures were streaked onto YPD plates. The single colony from last step was spotted on both YPD and YPDZ plates. Colonies that can grow on the YPD plate but not YPDZ plate, which indicated that the Cre-Zeo R cassette might have been excised. The result will be verified by PCR using primers pairs P3 and P4. Transcription levels of PHY and four methanol utilization related genes in the different P. pastoris strains. All four strains were harvested after 96 h of methanol induction. Subsequently, mRNA levels were analyzed by real-time PCR. PHY and four methanol utilization related genes (AOX1, DAS1, DAK2 and FBA1-2) levels were normalized relative to GS115/C-Phy. Statistical significance was examined using a two tailed by unpaired T-test analysis. *P < 0.05, **P < 0.01, ***P < 0.001, ns: no significant difference.
[fig] 8, Figure 9, Figure 10, Figure 11, Figure: 3-6 Table S2 .......................................................................................................................7 Table S3 .......................................................................................................................S1 …...................................................................................................................S2 .....................................................................................................................S3 .....................................................................................................................S4 .....................................................................................................................12 Figure S5 .....................................................................................................................13 [/fig]
[fig] Figure S2: PCR Assay of excision of the zeocin-resistance marker using primer pair P3 and P4.Lane M: DNA marker; lane 1: control (wild type P. pastoris GS115); lane 2: the transformant without methanol induction; lanes 3-12: colonies can grow on YPD but not YPDZ inFigure 2. [/fig]
[fig] Figure S4: Excessive copy numbers of PHY down-regulated the transcription level of methanol utilization genes. [/fig]
[table] Table S2: Marker recycling frequencies. [/table]
|
Global Metabolic Profiling of Infection by an Oncogenic Virus: KSHV Induces and Requires Lipogenesis for Survival of Latent Infection
Like cancer cells, virally infected cells have dramatically altered metabolic requirements. We analyzed global metabolic changes induced by latent infection with an oncogenic virus, Kaposi's Sarcoma-associated herpesvirus (KSHV). KSHV is the etiologic agent of Kaposi's Sarcoma (KS), the most common tumor of AIDS patients. Approximately one-third of the nearly 200 measured metabolites were altered following latent infection of endothelial cells by KSHV, including many metabolites of anabolic pathways common to most cancer cells. KSHV induced pathways that are commonly altered in cancer cells including glycolysis, the pentose phosphate pathway, amino acid production and fatty acid synthesis. Interestingly, over half of the detectable long chain fatty acids detected in our screen were significantly increased by latent KSHV infection. KSHV infection leads to the elevation of metabolites involved in the synthesis of fatty acids, not degradation from phospholipids, and leads to increased lipid droplet organelle formation in the infected cells. Fatty acid synthesis is required for the survival of latently infected endothelial cells, as inhibition of key enzymes in this pathway led to apoptosis of infected cells. Addition of palmitic acid to latently infected cells treated with a fatty acid synthesis inhibitor protected the cells from death indicating that the products of this pathway are essential. Our metabolomic analysis of KSHV-infected cells provides insight as to how oncogenic viruses can induce metabolic alterations common to cancer cells. Furthermore, this analysis raises the possibility that metabolic pathways may provide novel therapeutic targets for the inhibition of latent KSHV infection and ultimately KS tumors.
# Introduction
Many metabolic pathways are dramatically altered in cancer cells. These alterations are thought to provide cancer cells with the necessary energy and substrates for rapid cell division. Otto Warburg first demonstrated that most cancer cells have increased levels of glycolysis, even in the presence of oxygen, indicating that cancer cells dramatically alter their metabolism [bib_ref] On respiratory impairment in cancer cells, Warburg [/bib_ref]. The increased aerobic glycolysis seen in most cancer cells, now termed the Warburg effect, is often accompanied by decreased oxygen usage, indicating a dramatic shift in the source of energy for tumor cells. Cancer cells become dependent on increased glycolysis and thus require increased glucose uptake for survival [bib_ref] 2-deoxy-D-glucose causes cytotoxicity, oxidative stress, and radiosensitization in pancreatic cancer, Coleman [/bib_ref] [bib_ref] Evaluation of 2-deoxy-D-glucose as a chemotherapeutic agent: mechanism of cell death, Aft [/bib_ref] [bib_ref] Hypersensitization of tumor cells to glycolytic inhibitors, Liu [/bib_ref] [bib_ref] Enhancement of radionuclide induced cytotoxicity by 2-deoxy-D-glucose in human tumor cell lines, Shrivastava [/bib_ref]. In addition to the Warburg effect, many other metabolic changes occur in most tumor cells, including increases in lipogenesis, amino acid metabolism, and the pentose phosphate pathway among others.
Recently, global changes in cellular metabolism have been studied using metabolomic approaches [bib_ref] Dynamics of the cellular metabolome during human cytomegalovirus infection, Munger [/bib_ref] [bib_ref] Systems-level metabolic flux profiling identifies fatty acid synthesis as a target for..., Munger [/bib_ref]. Metabolomics generally involves the use of gas chromatography-mass spectrometry (GC-MS) and/or Liquid chromatography-mass spectrometry (LC-MS) to simultaneously detect changes in a wide variety of metabolites [bib_ref] Dynamics of the cellular metabolome during human cytomegalovirus infection, Munger [/bib_ref] [bib_ref] Hypoxia-induced metabolic shifts in cancer cells: Moving beyond the Warburg effect, Weljie [/bib_ref] [bib_ref] GC/MS-based metabolomics reveals fatty acid biosynthesis and cholesterol metabolism in cell lines..., Lin [/bib_ref] [bib_ref] Temporal proteome and lipidome profiles reveal hepatitis C virus-associated reprogramming of hepatocellular..., Diamond [/bib_ref]. Metabolomic approaches have allowed for the determination of global alterations of metabolism in tumor cells as well as in virally infected cells.
As non-living entities, viruses do not inherently have their own metabolism. However, upon infection, viruses dramatically alter the metabolism of the host cell. Viral alteration of host cell metabolism can provide the substrates necessary for viral replication. For example, alteration of host cell metabolism can provide the increased nucleotides necessary for genome replication or increased free amino acids needed for rapid viral protein synthesis. Virally-induced alterations of host metabolic pathways are likely to also be important for viral pathogenesis. Viral metabolomic studies were first used to identify changes in host cellular metabolism induced by human cytomegalovirus (HCMV) lytic infection [bib_ref] Dynamics of the cellular metabolome during human cytomegalovirus infection, Munger [/bib_ref] [bib_ref] Systems-level metabolic flux profiling identifies fatty acid synthesis as a target for..., Munger [/bib_ref]. These studies found that HCMV infection leads to the alteration of many key metabolic pathways including changes that are essential for lytic replication. Subsequently, changes in the cellular metabolic profile were determined for cells infected by other viruses, including hepatitis C virus (HCV), human immunodeficiency virus (HIV) and herpes simplex virus (HSV) [bib_ref] Temporal proteome and lipidome profiles reveal hepatitis C virus-associated reprogramming of hepatocellular..., Diamond [/bib_ref] [bib_ref] Metabolite profiles of human immunodeficiency virus infected CD4+ T cells and macrophages..., Hollenbaugh [/bib_ref] [bib_ref] Divergent effects of human cytomegalovirus and herpes simplex virus-1 on cellular metabolism, Vastag [/bib_ref]. These studies identified numerous metabolic alterations in infected cells. However, metabolomic studies have yet to be done on viruses that directly induce oncogenesis.
Kaposi's Sarcoma-associated herpesvirus (KSHV) is the etiologic agent of Kaposi's Sarcoma (KS). KS is the most common tumor in parts of central Africa as well as in AIDS patients worldwide. KS lesions are predominantly populated by spindle cells of endothelial origin [bib_ref] Biology of Kaposi's sarcoma, Ensoli [/bib_ref]. At late stages, all spindle cells in the tumor are infected with KSHV [bib_ref] Kaposi's sarcoma-associated herpesvirus gene expression in endothelial (spindle) tumor cells, Staskus [/bib_ref]. Like all herpesviruses, KSHV has both lytic and latent viral phases. During lytic replication, there is broad viral gene expression and the virus is rapidly replicated, ultimately leading to cell death. During latency, viral gene expression is limited, supporting expression of genes necessary for the maintenance of the latent episome and for infected cell survival. In KS spindle cells, KSHV is predominantly in the latent state, however, a few percent of the cells express viral markers of lytic replication [bib_ref] Concurrent expression of latent and a limited number of lytic genes with..., Krishnan [/bib_ref]. KSHV infection of cultured endothelial cells yields similar levels of latent and lytic infection [bib_ref] De novo infection and serial transmission of Kaposi's sarcoma-associated herpesvirus in cultured..., Lagunoff [/bib_ref].
Our previous studies found that KSHV infection of cultured endothelial cells directly induces the Warburg effect [bib_ref] Induction of the Warburg effect by Kaposi's sarcoma herpesvirus is required for..., Delgado [/bib_ref]. KSHV infection of endothelial cells leads to increased glucose uptake as well as increased lactic acid production. Expression of the glucose transporter, GLUT3, and the first rate-limiting enzyme in glycolysis, hexokinase-2, were also increased following latent infection. KSHV infection of endothelial cells also leads to significant decreases in oxygen consumption. Interestingly, both inhibition of lactate dehydrogenase, an essential enzyme for the conversion of pyruvate to lactate, or inhibition of glucose processing by a non-hydrolysable glucose analog, led to increased death of cells latently infected with KSHV [bib_ref] Induction of the Warburg effect by Kaposi's sarcoma herpesvirus is required for..., Delgado [/bib_ref]. These results indicate that glycolysis is required for the survival of cells latently infected with KSHV. This was the first evidence that KSHV directly induces a metabolic alteration common to most cancer cells. Furthermore, as observed in many cancer cells, this alteration is required for the survival of infected cells.
To determine if KSHV induces other metabolic changes similar to cancer cells, we performed a global screen to detect metabolites in mock-and KSHV-infected cells at 48 and 96 hours post infection (hpi). At 48 and 96 hpi, greater than 90% of the cells expressed a latent marker of infection, while only 1-5% of the cells expressed a marker of lytic replication. By 48 hpi, KSHV altered greater than one-quarter of the approximately 200 detectable metabolites and nearly one-third were altered by 96 hpi. Metabolites of several pathways commonly dysregulated in tumor cells were altered by KSHV, including glycolysis, amino acid metabolism, the pentose phosphate pathway and lipogenesis. KSHV infection leads to elevated levels of over half of the detectable metabolite products of de novo fatty acid synthesis (FAS or lipogenesis), and was concurrent with increased lipid droplets in latently infected endothelial cells. Lipid droplets are lipid-rich cytoplasmic organelles (diameter ,1-100 mm) that store cellular energy in the form of triacylglycerols and also store phospholipids and sterols, the building blocks of membranes [bib_ref] Lipid droplets finally get a little, Farese [/bib_ref] [bib_ref] Lipid droplets in inflammation and cancer, Bozza [/bib_ref]. Therefore, lipid droplets provide both readily accessible energy as well as cellular membrane material.
Fatty acids are essential substrates for energy metabolism and are the major components of all biological lipid membranes [bib_ref] Fatty acid synthase and the lipogenic phenotype in cancer pathogenesis, Menendez [/bib_ref]. Animal fatty acids can be derived exogenously from diet or endogenously by de novo lipogenesis. Normal cells predominantly acquire fatty acids from dietary sources rather than de novo lipogenesis. However, the majority of fatty acids from cancer cells are derived by de novo lipogenesis even when exogenous fatty acids are available [bib_ref] Metabolism of neoplastic tissue. IV. A study of lipid synthesis in neoplastic..., Medes [/bib_ref]. Induced lipogenesis is thought to aid in rapid cell division by generating the raw material needed for membrane production and the required metabolites for cell proliferation. Previous studies have shown that increased lipogenesis is required for the survival of many cancer cells [bib_ref] De novo fatty-acid synthesis and related pathways as molecular targets for cancer..., Mashima [/bib_ref]. We show here that the induction of FAS is also required for the survival of endothelial cells latently infected with KSHV. Inhibitors of FAS led to greatly increased apoptotic death of latently infected cells, as compared to their mock-infected counterparts. This effect was reversed by the addition of palmitc acid, the first fatty acid metabolite that is produced downstream of the drugs we used to block FAS. Therefore, the products of FAS are necessary for the survival of endothelial cells latently infected with KSHV. Inhibition of fatty acid synthesis provides a potential therapeutic target for KSHV and ultimately KS tumors.
# Results
## Kshv infection induces host cell metabolism
To determine global changes in host cell metabolism induced by latent KSHV infection, we utilized a metabolomic approach. Tertimmortalized human dermal microvascular endothelial cells (TIME cells) [bib_ref] De novo infection and serial transmission of Kaposi's sarcoma-associated herpesvirus in cultured..., Lagunoff [/bib_ref] were mock-or KSHV-infected and harvested at 48 and 96 hpi for metabolic profiling. As determined by immunofluorescence in all experiments, greater than 90% of the infected cells expressed the latency-associated nuclear antigen (LANA), a viral marker of latent infection, while less than 5% of the cells had detectable ORF 59 staining, a viral marker of lytic infection [bib_ref] De novo infection and serial transmission of Kaposi's sarcoma-associated herpesvirus in cultured..., Lagunoff [/bib_ref]. We harvested and analyzed metabolites from the lysates of duplicate infections at both 48 and 96 hpi in three independent experiments. Therefore, a total of six biological replicates at both time points were examined. Both GC-MS and LC-MS analysis was used to analyze a broad spectrum of metabolites and was performed by Metabolon (North Carolina). Detection of 194 distinct characterized biochemicals could be unambiguously identified in our samples. A large fraction of these metabolites were significantly altered by latent KSHV infection, nearing one-third of compounds by 96 hpi (TableS1). Specifically, at 48 hpi, 58 metabolites were determined to be altered with a pvalue of ,0.05, 45 metabolites were induced and 13 metabolites were decreased. By 96 hpi, 63 metabolites were determined to be altered with a p value of ,0.05, 51 metabolites were induced and 12 metabolites were decreased.
## Author summary
In recent years there has been a resurgence in the study of metabolic changes in tumor cells. To determine if an oncogenic virus alters similar metabolic pathways as cancer cells, we measured the levels of a large number of metabolites in endothelial cells infected with Kaposi's Sarcoma-associated herpesvirus (KSHV). KSHV is the etiologic agent of Kaposi's Sarcoma (KS), the most common tumor of AIDS patients world wide. Latent KSHV infection of endothelial cells altered a significant proportion of the host cell metabolites. Many metabolic pathways that are altered in most tumor cells were also altered by KSHV. In particular, KSHV upregulated fatty acid synthesis, a pathway that provides membrane material and metabolites critical for cell proliferation. Inhibitors of fatty acid synthesis kill many types of tumor cells and we found that these inhibitors led to death of cells latently infected with KSHV. In summary, we found that a directly oncogenic virus alters the same host metabolic pathways that are dysregulated in many cancer cells and that inhibition of these pathways can be used to kill off infected cells, thereby providing novel therapeutic targets for KSHV and ultimately KS tumors.
A variety of curation procedures were carried out by Metabolon to ensure the quality of the data set presented. System artifacts, mis-assignments, and background noise were limited by Metabolon data analysts through the use of proprietary visualization and interpretation software to confirm the consistency of peak identification among the various samples. Bradford assays were used to normalize each sample by protein concentration and each biochemical in the mock samples was re-scaled to yield a median equal to 1. KSHV-infected cells are shown as fold change compared to mock-infected cells. 194 named biochemical were unambiguously identified in our samples. Following log transformation and imputation with minimum observed values for each compound, Welch's two-sample t-tests were used to determine which biochemicals are significantly altered by KSHV infection at 48 and 96 hpi across the 6 biological replicates. An estimate of the false discovery rate (q-value) was calculated to take into account the multiple comparisons that normally occur in metabolomicbased studies. Low q-values were calculated for significantly different biochemical (p#0.05) for both comparisons, indicating a low estimated false discovery rate. The q-values for the significantly altered compounds are 0.06 and 0.09 at 48 hours and 96 hpi respectively.
Several major metabolic pathways that are often altered in cancer cells were significantly altered by KSHV infection . In the KSHV latently infected cells there is a general increase in amino acid metabolism including a large number of amino acids and a number of metabolites involved in the anabolic pathways of these amino acids. Induction of amino acid metabolism is thought to allow for the increased biosynthetic demand of rapidly dividing cells. Polyamine metabolism intermediates putrescine, spermidine, and spermine were also found at elevated levels in KSHV infected samples at 48 and 96 hpi. Polyamines are important for eukaryotic cell growth and differentiation and increased polyamine synthesis has been associated with cancer cells. Supporting our previous findings that increased glucose uptake is required for the survival of KSHV infected cells [bib_ref] Induction of the Warburg effect by Kaposi's sarcoma herpesvirus is required for..., Delgado [/bib_ref] , the metabolomic data show a significant increase in glucose in the infected samples at 96 hpi. The data also reveal increased levels of the major glycolytic intermediates 3-phosphoglycerate and 2-phosphoglycerate, and a significant increase in phosphoenolpyruvate (PEP). In addition to glycolysis, the metabolomic data indicate that glucose may also be utilized in the pentose phosphate pathway, which synthesizes both nucleotides and NADPH. The pentose phosphate pathway intermediates ribose 5-phosphate and the ribulose 5-phosphate and/or xylulose 5-phosphate (isobars that cannot be differentiated) were found at significantly elevated levels in KSHV infected samples at 48 hpi. Additionally, by 96 hpi, 6-phosphogluconate was measured at significantly elevated levels in KSHV infected samples. The elevation of these intermediates suggests that increased levels of nucleotides, nucleic acids and NADPH are required during infection. Importantly, it is known that NADPH is a necessary co-factor for fatty acid biosynthesis. In the KSHVinfected cells there were significant increases in many metabolites involved in de novo fatty acid synthesis (FAS) as described more extensively below. Of note, many of the metabolites that are increased by infection, are increased to higher levels at 96 hours as compared to 48 hours post infection. For example, increased glucose is only increased 1.38 fold at 48 hours with a p-value of greater than 0.05, while it is increased more than 10 fold at 96 hours with a p-value below 0.05 . At 96 hpi, many of the metabolites involved in FAS are also increased to a greater extent than at 48 hpi, suggesting that most of the changes are unlikely due to binding and entry but rather are dependent upon KSHV gene expression.
## Kshv infection induces host cell lipogenesis
Lipogenesis is induced in many cancer cells and involves the production of fatty acids and lipid material. Induction of lipogenesis allows increased cell proliferation and cancer cell survival. Our metabolomics study identified significant increases in many long chain fatty acids (LCFA) [fig_ref] Figure 1: KSHV infection of endothelial cells induces fatty acid production through de novo... [/fig_ref] , the building blocks of lipid material. In KSHV-infected cells, over 50% of the LCFA that could be detected were significantly increased, while no LCFA were decreased [fig_ref] Figure 1: KSHV infection of endothelial cells induces fatty acid production through de novo... [/fig_ref]. Observed elevation in LCFA may be due to increased lipid synthesis or degradation of phospholipids [bib_ref] Malignant transformation alters membrane choline phospholipid metabolism of human mammary epithelial cells, Aboagye [/bib_ref]. Analysis of the metabolomics data indicated an increase in choline and phosphocholine, fatty acid precursor metabolites generally associated with increased FAS, and decreases in glycerophosphorylcholine and glycerol-3-phosphate, metabolites associated with fatty acid production due to degradation of phospholipids [fig_ref] Figure 1: KSHV infection of endothelial cells induces fatty acid production through de novo... [/fig_ref]. This supports the conclusion that the increase in LCFAs in KSHV-infected cells is due to synthesis and not degradation of fatty acid containing material.
## Kshv infection induces lipid droplet formation
Lipid droplets are small cytoplasmic organelles that contain triacyglycerols for cellular energy storage, as well as, phospholipids and sterols for membrane production. Excess cellular LCFA can lead to the formation of lipid droplets [bib_ref] Longchain fatty acids induce lipid droplet formation in a cultured human hepatocyte..., Fujimoto [/bib_ref]. To determine if increased lipogenesis in KSHV-infected cells leads to increased lipid droplet formation, mock-and KSHV-infected TIME cells were stained with Oil-Red-O, a specific stain for lipid droplets, and hematoxylin, a stain for cell nuclei. In every experiment greater than 85% of the cells expressed LANA, a marker of latent infection while less than 3% of the cells expressed ORF 59, a marker of lytic replication. [fig_ref] Figure 2: KSHV-infected cells induce the formation of lipid droplet organelles [/fig_ref] shows representative images of high level staining for each condition. Many of the KSHVinfected cells had large numbers of lipid droplets while we only rarely saw significant lipid droplet staining in the mock-infected cells. To confirm and quantify this data we used flow cytometry to analyze mock-and KSHV-infected cells using another lipid droplet specific stain, Nile Red. There was much stronger Nile Red staining in KSHV-infected cells at 48 hpi as compared to mock-infected cells, indicating that, as was seen in the immunohistochemical staining, KSHV-infected cells have higher levels of lipid droplets [fig_ref] Figure 2: KSHV-infected cells induce the formation of lipid droplet organelles [/fig_ref]. To ensure that the increase in lipid droplets was due to viral gene expression we also stained cells infected with UV-inactivated KSHV. UV-inactivated virus still binds and enters cells, but viral gene expression is completely ablated. At 48 hpi, there was a slight increase in staining of the UV-inactivated virus as compared to mock infected cells but it was much lower than the levels seen in KSHV-infected cells (not shown). By 96 hpi, the staining of UV-inactivated virus was identical to mock-infected cells, while KSHV-infected cells still demonstrated a significant increase [fig_ref] Figure 2: KSHV-infected cells induce the formation of lipid droplet organelles [/fig_ref]. To determine if viral gene expression is continuously needed for lipid droplet formation or if lipid droplet formation is simply switched on by the virus, we passaged latently infected cells for over two weeks. As has been described before, long term passage of TIME cells leads to a loss of the viral episome [bib_ref] De novo infection and serial transmission of Kaposi's sarcoma-associated herpesvirus in cultured..., Lagunoff [/bib_ref]. At two weeks post infection,only around 40% of the cells maintain the viral episome with lower staining levels. Concurrent with loss of the viral episome, lipid droplet formation decreased to mock levels, indicating that continuous KSHV gene expression is required for lipid droplet formation (data not shown). Taken together, these data indicate that KSHV latent infection induces host cell lipogenesis and the infected cells store excess lipids in the form of lipid droplets.
## Fatty acid synthesis inhibitors induce apoptosis in kshvinfected cells
Inhibitors of FAS have been shown to selectively kill cancer cells and inhibit tumor cell growth [bib_ref] De novo fatty-acid synthesis and related pathways as molecular targets for cancer..., Mashima [/bib_ref]. To determine if KSHVinduced lipogenesis is necessary for the maintenance of latently infected endothelial cells, we treated mock-and KSHV-infected TIME cells or 1uhDMVECs with the FAS inhibitors 5-(Tetradecyloxy)-2-Furoic Acid (TOFA) or C75 in three biological replicates. TOFA inhibits the enzyme, acetyl-CoA carboxylase (ACC1), thereby preventing the conversion of acetyl-CoA into malonyl-CoA during lipogenesis [fig_ref] Figure 3: Fatty acid synthesis inhibitors selectively induce cell death in KSHV latently infected... [/fig_ref].Tumor cells are very sensitive to TOFA treatment, usually leading to cancer cell death via apoptosis [bib_ref] Acetyl-CoA carboxylasealpha inhibitor TOFA induces human cancer cell apoptosis, Wang [/bib_ref]. Forty-eight hpi, mock-and KSHV-infected TIME cells were treated with DMSO (control) or with TOFA for an additional 48 hours. After treatment, the percent of cell death was determined by a trypan blue exclusion assay. While KSHVinfected TIME cells have slightly higher basal cell death levels than their mock-infected counterparts, treatment with TOFA for 48 hours results in a very large dose-dependent increase in cell death in KSHV-infected cells while there is only a small increase [fig_ref] Figure 3: Fatty acid synthesis inhibitors selectively induce cell death in KSHV latently infected... [/fig_ref]. After normalizing for the basal cell death rate in the absence of drug, around 60% of KSHV-infected cells die due to treatment with 20 mg/mL TOFA while only 9% of mock-infected cells die due to treatment with the same concentrations of TOFA.
Fatty acid inhibitor C75 specifically inhibits fatty acid synthase (FASN), thereby preventing the catalysis of acetyl-CoA and malonyl-CoA into palmitic acid [fig_ref] Figure 3: Fatty acid synthesis inhibitors selectively induce cell death in KSHV latently infected... [/fig_ref]. Cancer cells are generally highly sensitive to C75 treatment compared to non-malignant cells. At 48 hpi, mock-and KSHV-infected cells were treated with C75 for 24 hours and cell death rates were determined by a trypan blue exclusion assay. Treatment with C75 for 24 hours induces a significantly larger dose-dependent increase in cell death in KSHV versus mock-infected cells [fig_ref] Figure 3: Fatty acid synthesis inhibitors selectively induce cell death in KSHV latently infected... [/fig_ref]. After normalizing for the basal death rate in the absence of drug, 70% of KSHV-infected cells die due to 4 mg/mL C75 treatment, while only 13% of mock-infected cells die due to identical C75 treatment. To determine the mechanism of cell death induced by TOFA and C75 in latently infected endothelial cells, classic markers of apoptosis were examined. Caspase 3 cleavage is an early event in many forms of programmed cell death and leads to the induction of PARP cleavage [bib_ref] Identification and inhibition of the ICE/CED-3 protease necessary for mammalian apoptosis, Nicholson [/bib_ref] [bib_ref] Human ICE/CED-3 protease nomenclature, Alnemri [/bib_ref]. Antibodies that recognize the cleaved forms of these proteins were used to determine if the treated cells are undergoing apoptosis. Endothelial cells were infected with KSHV for 48 hours and then treated with TOFA for an additional 36 hours or C75 for an additional 24 hours. Treatment of KSHV-infected cells with TOFA or C75 led to high levels of Caspase 3 and PARP cleavage [fig_ref] Figure 4: Fatty acid synthesis inhibitors selectively induce apoptosis in KSHV latently infected cells [/fig_ref]. Little or no induction of Caspase 3 or PARP cleavage was identified when mock-infected cells were treated with the same doses of C75 or TOFA. These results demonstrate that inhibition of fatty acid synthesis induces cell death by apoptosis specifically in KSHVinfected endothelial cells but not in mock-infected cells. Therefore, endothelial cells latently infected with KSHV require induced lipogenesis for survival.
## Metabolites downstream of palmitic acid are necessary for the survival of latently infected cells
Inhibition of ACC1 by TOFA could lead to apoptosis through inhibition of the synthesis of necessary downstream metabolites or through increased accumulation of precursors that could potentially be harmful to the cell. To determine if downstream pathways are necessary for latent infected cell survival we added palmitic acid, a metabolite that is downstream of the enzyme ACC1. Palmitic acid is the fundamental fatty acid precursor that iscreated by the conversion of acetyl-CoA and malonyl-CoA by the enzyme FASN. Addition of palmitic acid to tumor cell culture media has been shown to overcome blocks in fatty acid synthesis in tumor cells [bib_ref] Fatty acid synthesis is a therapeutic target in human liposarcoma, Olsen [/bib_ref]. Therefore, we tested if the addition of palmitic acid could overcome TOFA induced cell death in KSHV infected endothelial cells. As seen above, 20 mg/ml of TOFA induces death in approximately 60% of the cells latently infected with KSHV but did not increase cell death in mock-infected cells [fig_ref] Figure 5: Palmitic Acid rescues TOFA induced cell death in KSHV latently infected endothelial... [/fig_ref]. Addition of palmitic acid slightly but not significantly increased cell death in untreated mock-and KSHV-infected cells and in mock-infected cells treated with TOFA [fig_ref] Figure 5: Palmitic Acid rescues TOFA induced cell death in KSHV latently infected endothelial... [/fig_ref]. However, the addition of palmitic acid significantly decreased the percentage of dead cells to around 35% in cultures of KSHV-infected cells treated with TOFA [fig_ref] Figure 5: Palmitic Acid rescues TOFA induced cell death in KSHV latently infected endothelial... [/fig_ref]. Thus, the addition of palmitic acid partially rescues KSHV infected cell death when a key step of fatty acid synthesis is blocked with TOFA. These results were confirmed with a fluorescent assay in TIME cells with a Dead Green viability stain [fig_ref] Figure 5: Palmitic Acid rescues TOFA induced cell death in KSHV latently infected endothelial... [/fig_ref]. These results indicate that KSHV-infectedcell death is not due to the build up oftoxic fatty acid synthesis precursors, butrather that the synthesis of downstream fatty acid metabolites is necessary for the survival of latent infection.
# Discussion
It has long been known that viruses can alter specific metabolic processes. However, global changes in metabolism have only recently been examined for a few lytic viruses. The first landmark study of viral metabolism was performed on HCMV infected cells [bib_ref] Dynamics of the cellular metabolome during human cytomegalovirus infection, Munger [/bib_ref]. This study found changes in many metabolic pathways including glycolysis and fatty acid synthesis. Subsequently, other viruses have been examined including HCV, HIV, Dengue virus and HSV [bib_ref] Temporal proteome and lipidome profiles reveal hepatitis C virus-associated reprogramming of hepatocellular..., Diamond [/bib_ref] [bib_ref] Metabolite profiles of human immunodeficiency virus infected CD4+ T cells and macrophages..., Hollenbaugh [/bib_ref] [bib_ref] Divergent effects of human cytomegalovirus and herpes simplex virus-1 on cellular metabolism, Vastag [/bib_ref] [bib_ref] Metabolomics approach for investigation of effects of dengue virus infection using the..., Birungi [/bib_ref]. In all cases, there were dramatic changes in host cell metabolic pathways. Interestingly, viruses from different families induce some common metabolic pathways. For example, lytic infection with either HCMV or HCV induced glycolytic pathways [bib_ref] Systems-level metabolic flux profiling identifies fatty acid synthesis as a target for..., Munger [/bib_ref] [bib_ref] Temporal proteome and lipidome profiles reveal hepatitis C virus-associated reprogramming of hepatocellular..., Diamond [/bib_ref]. We recently demonstrated that KSHV also induces glycolysis during the latent phase of infection and this was confirmed by our metabolomics data [bib_ref] Induction of the Warburg effect by Kaposi's sarcoma herpesvirus is required for..., Delgado [/bib_ref]. We expected that the metabolic needs for latent and lytic phases of viral infection would be very different. However, the upregulation of glycolysis occurs both during latent KSHV infection and upon lytic infection with a G) were treated with 0 mg/mL or 20 mg/mL TOFA for 36 hours (C & E) or 0 mg/mL or 4 mg/mL C75 for 24 hours (D & F). Cells were then treated with Image-It Dead Green viability stain and visualized on an inverted fluorescent microscope. doi:10.1371/journal.ppat.1002866.g003 slow growing herpesvirus, HCMV, but not with lytic infection by HSV, a rapidly replicating herpesvirus [bib_ref] Systems-level metabolic flux profiling identifies fatty acid synthesis as a target for..., Munger [/bib_ref] [bib_ref] Divergent effects of human cytomegalovirus and herpes simplex virus-1 on cellular metabolism, Vastag [/bib_ref] [bib_ref] Induction of the Warburg effect by Kaposi's sarcoma herpesvirus is required for..., Delgado [/bib_ref].
A number of enveloped lytic viruses have been shown to be inhibited by a block in FAS [bib_ref] Systems-level metabolic flux profiling identifies fatty acid synthesis as a target for..., Munger [/bib_ref] [bib_ref] Fatty acid synthase expression is induced by the Epstein-Barr virus immediate-early protein..., Li [/bib_ref] [bib_ref] Evidence for attachment of fatty acid to varicella-zoster virus glycoproteins and effect..., Namazue [/bib_ref] [bib_ref] Phospholipid biosynthesis and poliovirus genome replication, two coupled phenomena, Guinea [/bib_ref] [bib_ref] Fatty acid synthase is up-regulated during hepatitis C virus infection and regulates..., Yang [/bib_ref] [bib_ref] Expression profile of lipid metabolism-associated genes in hepatitis C virus-infected human liver, Fujino [/bib_ref] [bib_ref] Dengue virus nonstructural protein 3 redistributes fatty acid synthase to sites of..., Heaton [/bib_ref]. It was proposed that lipogenesis provided membrane material for the envelopment of membrane containing viruses.Inhibition of FAS leads to decreased infectious virus for HCMV [bib_ref] Systems-level metabolic flux profiling identifies fatty acid synthesis as a target for..., Munger [/bib_ref] , influenza virus [bib_ref] Systems-level metabolic flux profiling identifies fatty acid synthesis as a target for..., Munger [/bib_ref] , Epstein-Barr virus (EBV) [bib_ref] Fatty acid synthase expression is induced by the Epstein-Barr virus immediate-early protein..., Li [/bib_ref] , varicella-zoster virus (VZV) [bib_ref] Evidence for attachment of fatty acid to varicella-zoster virus glycoproteins and effect..., Namazue [/bib_ref] , poliovirus [bib_ref] Phospholipid biosynthesis and poliovirus genome replication, two coupled phenomena, Guinea [/bib_ref] , HCV [bib_ref] Fatty acid synthase is up-regulated during hepatitis C virus infection and regulates..., Yang [/bib_ref] [bib_ref] Expression profile of lipid metabolism-associated genes in hepatitis C virus-infected human liver, Fujino [/bib_ref] , yellow fever virus, West Nile virus and Dengue virus [bib_ref] Dengue virus nonstructural protein 3 redistributes fatty acid synthase to sites of..., Heaton [/bib_ref]. Interestingly, treatment of EBV infected cells with fatty acid synthase (FASN) inhibitors C75 and cerulenin lead to reduced immediate-early and early gene expression [bib_ref] Fatty acid synthase expression is induced by the Epstein-Barr virus immediate-early protein..., Li [/bib_ref]. While the precise mechanism has yet to be investigated, it is not consistent with a block in envelopment. Dengue virus, which replicates on lipid droplets, has been shown to require induced fatty acid synthesis for replication at early stages as well [bib_ref] Dengue virus nonstructural protein 3 redistributes fatty acid synthase to sites of..., Heaton [/bib_ref]. Further work is necessary to determine if inhibition of FAS blocks KSHV lytic replication.
Interestingly, KSHV induces and requires lipogenesis during latent infection of endothelial cells. The increase in lipogenesis and lipid droplet accumulation was seen in the bulk of the population where greater than 85% of the cells expressed latent markers and less than 5% of the cells express lytic markers. This indicates that increased lipid droplets occurred in most of the latently infected cells. However, this does not rule out the possibility that lipid droplet formation requires a paracrine factor produced by lytic cells. Increased lipogenesis may predispose infected cells to oncogenesis as lipogenesis is upregulated in most cancer cells and is required for their survival. Inhibition of FAS led to high levels of apoptosis in KSHV latently infected cells, while the same concentration of inhibitors had little effect on the mock-infected cells. These effects were overcome by the addition of exogenous palmitic acid, a key FAS metabolite downstream of the early FAS inhibitor blockage. This indicates that downstream fatty acid metabolic processes are necessary for KSHV latency and that FAS inhibitors are not causing cell death simply due to build up of toxic pre-cursors. Importantly, apoptosis is also induced in many cancer cells following inhibition of FAS [bib_ref] De novo fatty-acid synthesis and related pathways as molecular targets for cancer..., Mashima [/bib_ref] [bib_ref] Acetyl-CoA carboxylasealpha inhibitor TOFA induces human cancer cell apoptosis, Wang [/bib_ref] [bib_ref] Fatty acid synthase inhibitors cerulenin and C75 retard growth and induce caspasedependent..., Ho [/bib_ref]. The question remains as to why KSHV latently infected cells require induced FAS. With regards to cancer cells, it has been hypothesized that induced FAS is necessary to meet the increased need for membranes due to high proliferation rates, as well as, for other important metabolic intermediates required for rapid cell division. However, in our hands, KSHV does not induce significant increases in proliferation of endothelial cells [bib_ref] Latent KSHV infection of endothelial cells induces integrin beta3 to activate angiogenic..., Dimaio [/bib_ref]. Therefore, KSHV likely requires specific FAS intermediates for survival or as a source of energy. Lipid droplet organelles can be used as energy storage for cells and could be critical for survival of the latently infected cells. Further experiments are necessary to identify the exact metabolites in the lipogenesis pathway that are necessary for the survival of cells latently infected with KSHV. The mechanism of induction of FAS by KSHV is also not clear. We did not see significant changes in the RNA levels or protein levels ofthree major FAS enzymes, ACC1, FASN or ACLY. We also found that induction of lipid droplets does not occur in all cell types as we did not see increased lipid droplets following high level infection of human foreskin fibroblast cells (data not shown). Further work on the cellular mechanisms are also ongoing.
All current drugs for herpesviruses specifically target only lytic replication steps. FAS inhibitors may provide novel targets for the inhibition of KSHV as they inhibit KSHV latency. Therefore, it is possible that FAS inhibitors could be used to treat KS tumors as they have the capacity to kill latently infected cells in the tumor. Overall, our metabolomics studies have identified a number of metabolic pathways altered by KSHV that are also commonly altered in cancer cells including glycolysis, amino acid synthesis, the pentose phosphate pathway, polyamine metabolism and fatty acid synthesis. These alterations in metabolism provide novel targets for inhibition of KSHV and possibly KS tumors. Experiments to identify the role of other metabolic pathways identified in our screen on KSHV pathogenesis are ongoing.
# Materials and methods
## Cell lines
Tert-immortalized microvascular endothelial (TIME) cells or primary human dermal microvascular endothelial cells (1uhDM-VECs) were maintained as monolayer cultures in EBM-2 media (Lonza or PromoCell) supplemented with a bullet kit containing 5%fetal bovine serum (FBS), vascular endothelial growth factor, basicfibroblast growth factor, insulin-like growth factor 3, epidermalgrowth factor, and hydrocortisone.
## Reagents and antibodies
Staurosporin (Sigma) was diluted in DMSO and were used at a final concentration of 1 mM. 5-(tetradecyloxy)-2-furancarboxylic acid (TOFA) (Sigma) and C75 (Calbiochem) were diluted in DMSO and used at the indicated final concentrations. Palmitic Acid (Sigma) was diluted in DMSO to 100 mM and used at a final concentration of 24 mM. Image-It Dead Green Viability Stain (Invitrogen) was diluted in DMSO and was used at a final concentration of 100 nM. Antibodies recognizing cleaved Caspase-3 (Cell Signaling), Poly-A Ribose Polymerase (Cell Signaling) ORF 59 (Advanced Biotechnology inc.) were used at 1:1000 dilutions as specified by the manufacturer and antibodies to b-Actin (Sigma) was used at 1:10,000. Antibodies that recognize KSHV protein, LANA (kind gift from Polson and Ganem) was used at 1:1000 dilutions.
## Viruses, infection, and immunofluorescence
KSHV inoculum from induced BCBL-1 cells was concentrated, titered and used to infect TIME and primary hDMVECs as previously described [bib_ref] Persistent activation of STAT3 by latent Kaposi's sarcoma-associated herpesvirus infection of endothelial..., Punjabi [/bib_ref]. All cells were infected in serum-free EBM-2 media for 3-4 hours, after which the medium was replaced withcomplete EGM-2. Infectionrates were assessed for each experiment by immunofluorescence and only experiments where greater than 85% of the cells expressed LANA and less than 5% of the cells expressed ORF59 were used. Prior to harvesting cells for immunoblot or enzymatic assays, an aliquotof mock-or KSHV-infected 1uhDMVECs or TIME cells were seeded on LabTekPermanoxfour-well chamber slides (Intermountain Sci) and fixed with4% (vol/vol) paraformaldehyde in phosphatebuffered saline. Infection rates were monitored using antibodies against thelatent KSHV protein LANA, as described previously [bib_ref] De novo infection and serial transmission of Kaposi's sarcoma-associated herpesvirus in cultured..., Lagunoff [/bib_ref]. For lytic replication, fixed cells were incubated with antibody to ORF 59 followed by incubation with fluor-conjugated secondary antibodies(goat anti-rabbit Alexa Fluor 488 and goat anti-mouse Alexa Fluor594 Molecular Probes/Invitrogen. Cells were mounted in medium containing DAPI (49, 69-diamidino-2phenylindole)before being viewed under a Nikon Eclipse E400 fluorescencemicroscope (Nikon, Inc.).
# Immunoblot analysis
TIME cells were mock-or KSHV-infected, harvested using a cell scraper and pelleted. An aliquot of thecells was seeded onto chamber slides for immunofluorescenceanalysis as described above. Cell pellets were washed once in cold phosphate-buffered saline and then resuspended in RIPA lysis buffer (50 mMTris-HCl, pH 7.6, 150 mMNaCl, 1 mM EDTA, 1% NP-40, 0.5%deoxycholate, 0.1% sodium dodecyl sulfate, 1 mM sodium orthovanadate,1 mM sodium fluoride, 40 mM b-glycerophosphate, and CompleteMini protease inhibitor tablet [Roche]). Proteinconcentrations were determined by the bicinchoninic acid assay(Pierce). 15-50 mg of protein was fractionated on a sodiumdodecyl sulfatepolyacrylamide gel, transferred to a polyvinyldifluoride membrane, blotted with the indicated primary antibody diluted as described above, and subsequently with horseradish peroxidaseconjugated donkeyanti-goat (1:20,000), goat anti-rabbit (1:10,000), goat anti-mouse (1:10,000).Immunoreactive proteins were visualized by chemiluminescence using the Amersham ECL Plus Western blotting detection reagents(GE Healthcare).
Trypan blue exclusion assay and Image-It Dead Green Viability stain 48 hpi, mock-or KSHV-infected TIME cells or 1uhDMVECs were seeded sub-confluently (60,000 cells per well) into 12 well plates and treated with TOFA (0 mg/mL, 10 mg/mL or 20 mg/ mL) for 48 hours or treated with C75 (0 mg/mL, 4 mg/mL or 8 mg/mL) for 24 hours. Cells were trypsinized and pelleted with cellular supernatant for 5 minutes and cell pellet was resuspended in ,50 uL media. Cell death rates were determined by counting cells using a hemocytometer after addition of trypan blue, which is actively exported from the cytoplasm of live cells but remains in dead cells. Cell death rate (%) = number of dead cells/number of total cells (*100%). For the Dead Green viability assay, TIME cells or 1uhDMVECs were mock-or KSHV-infected for 48 hours, seeded sub-confluently in 12 well plates and treated 0 mg/mL or 4 mg/mL C75 for 24 hours, or 0 mg/mL or 20 mg/mL TOFA for 24 hours. Cells were then treated with Image-It Dead Green Viability stain for 15 min and pictures were taken with a fluorescent microscope. For Palmitic Acid rescue experiments, 48 hpi, mock-and KSHV infected TIME cells were seeded subconfluently (60,000 cells/well) in 12 well plates and treated with TOFA at 20 mg/ml in the presence or absence of 24 mM Palmitic Acid for an additional 48 hours. Both trypan blue and Dead Green viability assays were completed as described above.
Lipid droplet staining 48 hpi, mock-and KSHV-infected TIME cells were seeded in a chamber slide, fixed with 10% Formalin in PBS for 40-60 minutes and washed with 60% Isopropanol for 5 minutes. The cells were then exposed to fresh Oil-Red-O working solution for 10 minutes to stain the lipid droplets. Cells were then washed with dH2O, counterstained with hematoxylin for 1 minute to stain the nuclei and then washed and mounted in Aqua Poly/Mount (Polysciences Inc.). Pictures were taken under bright field microscopy. Images represent frames with high levels of lipid droplets for each condition.
## Flow cytometry
Mock-and KSHV-infected cells were trypsinized to remove cells from plates, counted and processed for flow cytometry. For detection of lipid droplets, the Lipid Droplets Florescence Assay Kit (Cayman Chemical) was used as recommended by the manufacturer. Briefly, 48 hpi TIME cells were washed in 1 mL Assay Buffer, fixed for 10 minutes at room temperature, washed with 1 mL Assay buffer, resuspended in Nile Red Staining Solution for 15 minutes at room temperature, washed and resuspended in Assay Buffer. Samples were analyzed by FACScan caliber flow cytometer (Becton-Dickinson, Franklin Lakes, NJ) or by FACS Canto flow cytometer. All Samples were analyzed by FloJo flow cytometry analysis software (Tree Star, Inc.; Ashland, OR).
## Supporting information
Table S1 Relative levels of allthe metabolites identified in KSHV infected cells compared to mock infected cells at 48 and 96 hours post infection. Values shown are normalized levels from 6 distinct KSHV and mock infections of TIME cells. The major metabolic pathways are indicated in the right columns and the biochemical name and the mass spectrometry platforms used are indicated in the middle columns. Metabolites that are significantly increased (p,0.05) are shaded in red and those significantly decreased (p,0.05) are shaded in green. Numbers in blue are altered in KSHV infected cells but with lower statistical significance (0.05,p,0.01). (DOCX)
[fig] Figure 1: KSHV infection of endothelial cells induces fatty acid production through de novo synthesis. Relative levels of metabolites from mock-(control) and KSHV-infected (infected) TIME cells are shown at 48 hpi (light blue and pink respectively) and 96 hpi (dark blue and red respectively). A) Box and whisker plots showing relative levels of fatty acid metabolites significantly altered by KSHV infection. B) Box and whisker plots of metabolites that differentiate production of long chain fatty acids (LCFAs) by synthesis or degradation of phospholipids indicating that increased fatty acid metabolites in KSHV infected cells come from increased synthesis. doi:10.1371/journal.ppat.1002866.g001 in cell death in the mock-infected cultures (fig. [/fig]
[fig] Figure 2: KSHV-infected cells induce the formation of lipid droplet organelles. A) 48 hpi, TIME cells were fixed and stained with Oil-Red-O, a specific stain for lipid droplets, and hematoxylin, to stain cell nuclei. Cells were visualized by bright field microscopy. B) Mock-and KSHV-infected TIME cells were harvested at 48 hpi, fixed and stained with Nile Red, a specific fluorescent stain for lipid droplets. Staining was analyzed by flow cytometry. C) Mock-, KSHV-and UV-irradiated-KSHV-infected TIME cells were harvested at 96 hpi and stained as in B for lipid droplets. doi:10.1371/journal.ppat.1002866.g002 [/fig]
[fig] Figure 3: Fatty acid synthesis inhibitors selectively induce cell death in KSHV latently infected endothelial cells. A) Schematic of lipogenesis pathway and where specific inhibitors block the pathway. At 48 hpi, mock-and KSHV-infected TIME cells were treated with B) 0, 10, or 20 mg/mL of TOFA and incubated for an additional 48 hours or C) 0, 4 or 8 mg/mL C75 and incubated for an additional 24 hours. Cell death rates were then determined by a trypan blue exclusion assay. Cell death rate (%) is # of dead cells/# of total cells. 48 hpi, TIME cells (D & E) or 1u hDMVECs (F &The FAS inhibitor results were confirmed in TIME cells (fig. 3D and E) and 1uhDMVECs (fig. 3Fand G) by a microscopy-based fluorescence assay for dead cells (Image-It Dead Green Viability Stain). The fluorescent substrate is taken up by dead cells but is excluded from live cells. Treatment with either TOFA or C75 induced high levels of cellular fluorescence in KSHV-infected cells but not in mock-infected cells. [/fig]
[fig] Figure 4: Fatty acid synthesis inhibitors selectively induce apoptosis in KSHV latently infected cells. 48 hpi, mock-(M) and KSHV-(K) infected cells were treated with A) 20 mg/mL TOFA for 36 hours or B) 4 mg/mL C75 for 24 hours and then harvested forWestern blot analysis with antibodies that recognize the cleaved PARP or cleaved caspase-3. doi:10.1371/journal.ppat.1002866.g004 [/fig]
[fig] Figure 5: Palmitic Acid rescues TOFA induced cell death in KSHV latently infected endothelial cells. A) At 48 hpi, mock-and KSHVinfected TIME cells were treated with DMSO (control) or 20 mg/mL of TOFA, in the presence or absence of 24 mM Palmitic Acid, and incubated for an additional 48 hours. Cell death rates were then determined by a trypan blue exclusion assay. Cell death rate (%) is # of dead cells/# of total cells. B) At 48 hpi, TIME cells were treated as in panel A. Cells were then treated with Image-It Dead Green viability stain and visualized on an inverted fluorescent microscope. doi:10.1371/journal.ppat.1002866.g005 [/fig]
|
Identification of a genetic locus for autosomal dominant infantile cataract on chromosome 20p12.1-p11.23 in a Chinese family
Purpose: To map a gene responsible for infantile cataract in a large four-generation, non-consanguineous Chinese family. Methods: Twenty-two family members including 17 cataract patients in the Chinese family were analyzed clinically. All family members were genotyped with 382 microsatellite markers that provide genome-wide coverage every 10 cM. Linkage analysis was performed to identify the chromosomal location of the infantile cataract gene in the family. Candidate genes were studied by direct DNA sequence analysis. Results: Genome-wide linkage analysis provided evidence for a genetic locus for infantile cataract on chromosome 20p12.2-20p11.23. The maximum LOD score was 5.15 for marker D20S471 at a recombination fraction of 0. Fine mapping defined the cataract gene within a 7.4 Mb interval between markers D20S915 and D20S912. No mutation was detected in potential candidate genes, BFSP1 and CHMP4B. Conclusions: Our results suggest that there is a new gene for infantile cataract on chromosome 20p12.2-p11.23. Our results suggest that new genes for infantile cataract could be found through further study of candidate genes at the 20q locus, which may provide insights into the pathogenic mechanisms of cataracts.
Congenital cataract is a significant cause of hereditary visual impairment in childhood. Its incidence is estimated to be 2.2-2.49 per 10,000 live births and may account for onethird of total blindness in infants [bib_ref] Measuring and interpreting the incidence of congenital ocular anomalies: lessons from a..., Rahi [/bib_ref] [bib_ref] Infantile cataract in the collaborative perinatal project: prevalence and risk factors, Sangiovanni [/bib_ref] [bib_ref] Aetiology of congenital and paediatric cataract in an Australian population, Wirth [/bib_ref]. Congenital cataract is genetically and clinically highly heterogeneous, and various inheritance patterns have been reported including autosomal dominant, autosomal recessive, and X-linked forms [bib_ref] Genetic and segregation analysis of congenital cataract in the Indian population, Vanita [/bib_ref] [bib_ref] A locus for isolated cataract on human Xp, Francis [/bib_ref]. To date, 21 genetic loci have been identified, and 16 causative genes have been identified for autosomal dominant congenital cataract (ADCC), the most common familial form [bib_ref] Genetic origins of cataract, Shiels [/bib_ref]. The phenotype of ADCC also varies markedly in morphology, affecting the nuclear, cortical, polar, and other sections of the lens [bib_ref] Clinical and genetic heterogeneity in autosomal dominant cataract, Ionides [/bib_ref]. Total cataract is characterized by opacity of all lens fibers. For autosomal dominant total cataract, one mutation in the heat shock transcription factor (HSF4) gene was reported by us [bib_ref] Novel HSF4 mutation causes congenital total white cataract in a Chinese family, Ke [/bib_ref] , and another mutation was identified in a gene encoding the major intrinsic protein (MIP, MIP26) of the lens [bib_ref] A novel mutation in major intrinsic protein of the lens gene (MIP)..., Gu [/bib_ref]. For autosomal recessive total cataract, a mutation was reported in the gene encoding the α-A component of αcrystallin (CRYAA) [bib_ref] Recessive congenital total cataract with microcornea and heterozygote carrier signs caused by..., Khan [/bib_ref]. There are three ADCC families linked to chromosome 20 including a Japanese autosomal dominant posterior cataract family linked to chromosome 20p12-20q12 [bib_ref] An autosomal dominant posterior polar cataract locus maps to human chromosome 20p12-q12, Yamada [/bib_ref] and a Chinese family with autosomal dominant progressive congenital zonular nuclear cataract that is linked to chromosome 20p12.2-20p11.23 [bib_ref] An autosomal dominant progressive congenital zonular nuclear cataract linked to chromosome 20p12.2-p11.23, Li [/bib_ref]. Very recently, a white (U.S.A.) family with progressive childhood posterior subcapsular cataracts has been reported to be linked to 20q, and CHMP4B was identified as the pathogenic gene for ADCC at this locus [bib_ref] CHMP4B, a novel gene for autosomal dominant cataracts linked to chromosome 20q, Shiels [/bib_ref].
In this study, we analyzed a four-generation, nonconsanguineous Chinese family diagnosed as having infantile total cataract. It is different from other congenital cataract because there is no amblyopia presented posteriorly with intraocular lens transplantation. Because the cataract was present at the age of 10-12 years and showed progressive development of lens opacities within one to two years and decreased visual acuity. Our genome-wide linkage screen mapped the new infantile cataract pathogenic gene on chromosome 20p12.2-20p11.23. The maximum LOD score reached 5.15 for marker D20S471 at a recombination fraction of 0. The infantile cataract gene was further defined within a 12.5 cM (7.5 Mb) interval between markers D20S915 and D20S912. In this region, the potential candidate genes include BFSP1 and the newly-reported pathogenic gene, CHMP4B. However, direct DNA sequence analysis did not identify any mutation in the two candidate genes. Our results suggest that a gene for infantile total cataract is located on chromosome 20p11.23.
# Methods
## Study participants:
We recruited a four-generation, nonconsanguineous Chinese family from the Henan province with multiple family members diagnosed as having total cataract. The family displayed an autosomal dominant inheritance pattern [fig_ref] Figure 1: Pedigree structure of a Chinese family with autosomal dominant total cataract [/fig_ref]. There are 17 patients. Blood samples were obtained from 14 living patients.
Complete medical history analyses and ophthalmic examinations were performed on all living family members. The examinations include assessment of visual acuity and a detailed examination of the ocular lens under a slit lamp to determine the disease status. The results are shown in [fig_ref] TABLE 1: CLINICAL CHARACTERISTICS OF FAMILY MEMBERS AT RISK FOR CONGENITAL TOTAL CATARACT IN... [/fig_ref]. In this family, the phenotype of lens opacity showed total cataract in morphology [fig_ref] Figure 2: Slit-lamp image of the lens opacity from an individual with congenital total... [/fig_ref]. The initial clinical manifestation for all affected members was opacity of the lens and decreased visual acuity at the age of 10-12 years. Interestingly, the cataract phenotype starts to manifest as early as just after birth [fig_ref] Figure 2: Slit-lamp image of the lens opacity from an individual with congenital total... [/fig_ref].
Informed consent was obtained from all participants, and the study was in accordance with the tenets of the Declaration of Helsinki on human subject research. Genotyping: Genomic DNA was extracted from the peripheral blood of 22 family members with a DNA isolation kit for mammalian blood (Wizard Genomic DNA Purification kit; Promega, Madison, WI). The DNA samples were quantified by a spectrophotometer and diluted to 25 ng/μl for polymerase chain reaction (PCR) amplification. The initial genome-wide screen was performed with 382 highly polymorphic fluorescent markers (PRISM Linkage Mapping Set MD-10, Applied Biosystems, Foster City, CA) that have an average spacing interval of 10 cM over the entire human genome. Other fluorescently labeled markers were selected and designed according to the Marshfield Clinic Medical Genetics database for fine mapping. Each PCR genotyping reaction was performed in a 5 μl volume containing 50 ng of genomic DNA, 10 μM dye-labeled primer pairs, 0.5 μl 10X PCR buffer (GeneAmp Buffer II, Applied Biosystems), 0.5 μl 10 mM dNTP mix, and 0.2 U of Taq DNA polymerase (AmpliTaq Gold Enzyme, Applied Biosystems). Amplification was performed in a GeneAmp 9700 PCR system (Applied Biosystems). PCR products (1 μl) from each DNA sample was pooled and mixed with 0.2 μl of Liz Size Standard-500 and 9 μl of Hi-Di formamide (both from Applied Biosystems, Foster City, CA). The mixture was then fractionated by electrophoresis and visualized on a 3100 Genetic Analyzer (Applied Biosystems). The GenScan 3.1 and GeneMapper 2.5 software (Applied Biosystems) were used to analyze the alleles. Linkage analysis: Two-point linkage analysis was performed using the MLINK program from the LINKAGE program package (version 5.2). The linkage was performed assuming an autosomal dominant inheritance pattern, 99% penetrance, and a disease gene frequency of 0.0001.
Multipoint linkage analysis was computed using GENEHUNTER version 2.1 running on the Linux operating system, and the marker order and positions were based on the Marshfield Clinic Medical Genetics map [fig_ref] TABLE 2: TWO-POINT LOD SCORES FROM LINKAGE ANALYSIS OF TOTAL CATARACT FOR MARKERS ON... [/fig_ref]. Haplotype analysis was performed using the Cyrillic 2.1 (Cyrillic Software, Setauket, NY) program and manual prediction. [fig_ref] Figure 1: Pedigree structure of a Chinese family with autosomal dominant total cataract [/fig_ref] illustrates the typical cataract identified in this pedigree. The cataract was present around 10-12 years of age and showed progressive development of lens opacities within one to two years and decreased visual acuity, but it is different from others because no amblyopia is presented posteriorly with IOL transplantation. Slit lamp examinations were performed to characterize the lens phenotype. The phenotype is lamellar cataract. Perinuclear-shaped total opacities were restricted to the lamellae and nucleus. The total cataract morphology was identified for all affected members in the family (data not shown).
Mutational analysis: All exons and exon-intron boundaries of candidate genes, BFSP1 and CHMP4B, were amplified by PCR, purified, and sequenced using the BigDye Terminator Cycle Sequencing v3.1 kit (Applied Biosystems).
# Results
We clinically characterized a large Chinese family with a diagnosis of total cataract, which manifests as early as at birth (infantile cataract, [fig_ref] Figure 2: Slit-lamp image of the lens opacity from an individual with congenital total... [/fig_ref]. A genome-wide genotyping scan was performed for 22 members of the family including 14 living patients, five normal individuals, and three married spouses, using 382 microsatellite markers covering the entire human autosome every 10 cM. We obtained a positive two-point LOD score of 2.73 at θ=0 with marker D20S195 [fig_ref] TABLE 2: TWO-POINT LOD SCORES FROM LINKAGE ANALYSIS OF TOTAL CATARACT FOR MARKERS ON... [/fig_ref]. The two markers flanking D20S195 were uninformative. Fine mapping was then carried out with markers near D20S195, and multiple markers showed LOD scores greater than 3.0 [fig_ref] TABLE 2: TWO-POINT LOD SCORES FROM LINKAGE ANALYSIS OF TOTAL CATARACT FOR MARKERS ON... [/fig_ref]. Marker D20S471 showed the highest LOD score of 5.15. Multipoint linkage analysis showed a peak LOD score of nearly 5.0 for the chromosomal region from D20S910 to D20S471 . These results revealed that the gene for infantile total cataract in the Chinese family is linked to markers D20S910 to D20S471 on chromosome 20 with significant LOD scores.
Haplotype analysis was constructed for eight markers on chromosome 20p12.2-p11.23. Obligate recombination events were identified in patient IV-4 between D20S604 and D20S915 and in individual IV-7 between D20S915 and D20S910. Thus, D20S915 was defined as the left flanking marker for the locus. One non-obligate recombination event was detected in a normal individual, IV-8, between D20S471 and D20S912, which defines D20S912 as the right flanking marker for the locus. Mutational analysis of two candidate genes on chromosome 20, BFSP1 and CHMP4B, did not reveal any disease-associated mutation.
# Discussion
Using genome-wide genetic linkage analysis, we have shown that a gene for an autosomal dominant infantile total cataract is located on the short arm of chromosome 20, within a 7.4 Mb interval between markers D20S915 and D20S912. Recently, Shiels at el. [bib_ref] CHMP4B, a novel gene for autosomal dominant cataracts linked to chromosome 20q, Shiels [/bib_ref] reported linkage of autosomal dominant progressive childhood posterior sub-capsular cataracts to chromosome 20q in a Caucasian family and identified the pathogenic gene as CHMP4B in a refined disease interval of 3.5 cM. We used direct DNA sequence analysis for mutational analysis of CHMP4B, but no mutation was identified. Furthermore, CHMP4B is located outside of the refined disease locus. Thus, the infantile cataract gene on chromosome 20p12.2-p11.23 in the Chinese family is not CHMP4B.
One Japanese autosomal dominant posterior cataract family has been linked to chromosome 20p12-20q12 [bib_ref] An autosomal dominant posterior polar cataract locus maps to human chromosome 20p12-q12, Yamada [/bib_ref]. A Chinese autosomal dominant progressive congenital zonular nuclear cataract family has been linked to chromosome 20p12.2-p11.23 [bib_ref] An autosomal dominant progressive congenital zonular nuclear cataract linked to chromosome 20p12.2-p11.23, Li [/bib_ref]. Interestingly, the Chinese family with infantile total cataract under this study is also linked to the same region. It is possible that the same gene is responsible for three different types of cataracts in the Japanese family, the Chinese family, and the family under this study. Our . Multipoint linkage analysis spanning 19 cM between total cataract phenotype and markers from D20S604 to D20S871 using the GENEHUNTER 2.1 program. Genetic distance between the markers is as indicated in [fig_ref] TABLE 1: CLINICAL CHARACTERISTICS OF FAMILY MEMBERS AT RISK FOR CONGENITAL TOTAL CATARACT IN... [/fig_ref] infantile total cataract locus is 2 cM smaller than the progressive congenital zonular nuclear cataract locus identified in the other Chinese family and much smaller than the posterior cataract locus identified in the Japanese family. On the other hand, due to the highly clinical and genetic heterogeneity of congenital cataracts, it may be possible that the distinct genes contributed to the different cataract phenotype at this specific chromosomal region. Future identification of these specific genes should be able to distinguish the two hypotheses.
There are no other obvious candidate genes for autosomal dominant total cataract on chromosome 20p12.2-p11.23 except for BSFP1. BSFP1 encodes the beaded filament structural protein 1, a lens-specific intermediate filament-like protein, which functions as a major cytoskeletal element of the eye lens and is essential to the optical properties of eye lens. Mutations in BFSP2 have been reported to be associated with autosomal dominant congenital cataract [bib_ref] A juvenile-onset, progressive cataract locus on chromosome 3q21-q22 is associated with a..., Conley [/bib_ref] [bib_ref] Autosomal-dominant congenital cataract associated with a deletion mutation in the human beaded..., Jakobs [/bib_ref]. In a consanguineous family of Indian origin with autosomal recessive juvenile onset cortical cataract, linkage was detected with markers between D20S852 and D20S912 (peak LOD=5.4 with D20S860), and one homozygous deletion in BFSP1 was identified [bib_ref] Autosomal recessive juvenile onset cataract associated with mutation in BFSP1, Ramachandran [/bib_ref]. Heterozygous carriers did not develop cataracts. Direct DNA sequence analysis of the entire coding region and exon-intron boundaries of BFSP1 has been previously conducted in the Japanese and Chinese cataract families linked to chromosome 20, but no mutation was identified [bib_ref] An autosomal dominant posterior polar cataract locus maps to human chromosome 20p12-q12, Yamada [/bib_ref] [bib_ref] An autosomal dominant progressive congenital zonular nuclear cataract linked to chromosome 20p12.2-p11.23, Li [/bib_ref]. We also performed direct DNA sequence analysis of BFSP1 for the proband from the family under this study but did not detect any mutation. Although we cannot exclude the possibility that a BFSP1 mutation in the promoter or an intron may be associated with cataract in the family, this is unlikely because our results are consistent with the findings by Yamada et al. [bib_ref] An autosomal dominant posterior polar cataract locus maps to human chromosome 20p12-q12, Yamada [/bib_ref] , Li et al. [bib_ref] An autosomal dominant progressive congenital zonular nuclear cataract linked to chromosome 20p12.2-p11.23, Li [/bib_ref] , and Ramachandran et al. [bib_ref] Autosomal recessive juvenile onset cataract associated with mutation in BFSP1, Ramachandran [/bib_ref] that heterozygous carriers for a BFSP1 deletion were phenotypically normal.
In summary, these results indicate that there is new gene on chromosome 20p12.2-p11.23 that is responsible for infantile total cataract. The disease gene interval has been defined between markers D20S915 and D20S912, a 7.4 Mb region. Future studies of the candidate genes within the locus should identify the specific gene, which will provide further important insights into the genetic basis of infantile cataracts.
# Acknowledgments
Dr. Q. K. Wang and Dr. L. Wang are equally responsible for the conduct of the research reported in this article and can be considered to be co-corresponding authors. We are grateful to the patients and their family members for their cooperation in this study. This study was supported by the China National Program for High Technology Research and Development ("863"Programs; No.2002BA711A07). The study was approved by the Ethics Committee of the Beijing University
[fig] Figure 1: Pedigree structure of a Chinese family with autosomal dominant total cataract. The genotyping results are shown for markers D20S604, D20S915, D20S910, D20S98, D20S470, D20S471, D20S912, and D20S871. The haplotype inherited from the affected parents is shown on the left of each pair. The vertical dark-box bars represent the mutant haplotype, and the open bars are normal haplotypes. [/fig]
[fig] Figure 2: Slit-lamp image of the lens opacity from an individual with congenital total cataract from the Chinese family. A pre-operative photo of the left eye from patient III:1 (see [/fig]
[table] TABLE 1: CLINICAL CHARACTERISTICS OF FAMILY MEMBERS AT RISK FOR CONGENITAL TOTAL CATARACT IN THE CHINESE FAMILY.Y: affected and with IOL transplantation; N: unaffected and without IOL transplantation. OD = right eye; OS = left eye. [/table]
[table] TABLE 2: TWO-POINT LOD SCORES FROM LINKAGE ANALYSIS OF TOTAL CATARACT FOR MARKERS ON CHROMOSOME 20P12-20Q12 IN THE CHINESE [/table]
|
COVID‐19 infection and its consequences among surgical oncology patients: A systematic analysis, meta‐analysis and meta‐regression
We conducted this meta-analysis to address the outcomes in cancer patients after oncologic surgery during COVID-19 pandemic. The primary endpoint was the COVID-19-related mortality rate. Higher body mass index was significantly and negatively associated with higher all-cause mortality and in-hospital COVID-19 infection rates. Male sex, preoperative respiratory disease, and smoking history were positively and significantly associated with increased all-cause mortality rates. Furthermore, male sex was positively and significantly associated with the COVID-19 infection rate. K E Y W O R D S cancer, COVID-19, meta-analysis, mortality, surgery 1 | INTRODUCTION During the global COVID-19 pandemic, elective surgical procedures have frequently been rescheduled, mainly to preserve medical resources as well as minimize the exposure of patients and health care providers to COVID-19. 1 As a result, maintaining the surgical oncology workflow has been a clinical challenge. 2 Thus, researchers and clinicians have shared and reported institutional experiences, global recommendations, suggestions for adjusted workflows, and international guidelines to optimize cancer care without compromising oncologic surgery outcomes during the COVID-19 pandemic. 3,4 However, one size does not fit all, as travel restrictions, national infection and vaccination rates, availability of medical infrastructure, accessibility to medical resources, and numbers of available providers mandate personalized surgical oncology workflows that follow international guidelines for minimizing the transmission of COVID-19 infection and maintaining patient and workforce safety. Consequently, collaborative work and timely feedback on the feasibility and efficiency of the adjusted surgical workflows are critically needed. Surgical oncology strategies take into account the need for timely and optimized treatment modalities for downstaging and systemic control, preoperative (nononcologic) interventions to minimize negative perioperative outcomes such as wound complications and length of hospital stay, 5-7 and the presence of comorbidities that J Surg Oncol. 2022;125:813-823.wileyonlinelibrary.com/journal/jso
cannot be modified to determine the patient's readiness for oncologic surgery. [bib_ref] Cancer surgery scheduling during and after the COVID-19 first wave: The MD..., Tzeng [/bib_ref] Additional factors must be considered during the COVID-19 pandemic: uncertain COVID-19 infection trajectory, actual infection risk among health care providers, and expected vaccine protection rates. [bib_ref] Surgical decision-making and prioritization for cancer patients at the onset of the..., Tzeng [/bib_ref] [bib_ref] Personal protective equipment and COVID-19: a review for surgeons, Stewart [/bib_ref] Moreover, some surgical procedures may require interventional radiologic or endoscopic procedures and/or admission to an intensive care unit (ICU) postoperatively or emergency room (ER) perioperatively. Furthermore, virtual perioperative appointments, limited numbers of caregivers and visitors, and additional barriers designed to protect patients and health care providers from infection may further complicate interventions and patient advocacy. [bib_ref] Management of cancer surgery cases during the COVID-19 pandemic: considerations, Bartlett [/bib_ref] [bib_ref] A combined approach to priorities of surgical oncology during the COVID-19 epidemic, Mazzaferro [/bib_ref] [bib_ref] Cancer patients in SARS-CoV-2 infection: a nationwide analysis in China, Liang [/bib_ref] The available published data on, international recommendations for, and global experience with infections in cancer patients may help overcome the challenges described above. [bib_ref] Management of cancer surgery cases during the COVID-19 pandemic: considerations, Bartlett [/bib_ref] [bib_ref] Safety recommendations for evaluation and surgery of the head and neck during..., Givi [/bib_ref] [bib_ref] Gynaecologic cancer care during COVID-19 pandemic in India: a social media survey, Kumari [/bib_ref] [bib_ref] Surgical prehabilitation in patients with cancer: state-of-the-science and recommendations for future research..., Carli [/bib_ref] Nevertheless, our understanding of the COVID-19 pandemic is evolving, and the oncologic surgery workflow must be revisited, updated, and optimized as a result to account for the range of institutional, local, and national conditions during the pandemic. Therefore in this systematic review, meta-analysis, and meta-regression, we analyzed the measurable reported outcomes and identified potential predictors of poor outcome in cancer patients who had oncologic surgery during the COVID-19 pandemic.
# | methods
## | literature search
A systematic review and meta-analysis were conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. [bib_ref] The PRISMA statement for reporting systematic reviews and meta-analyses of studies that..., Liberati [/bib_ref] The Ovid MEDLINE, Ovid Embase, Clarivate Analytics Web of Science, PubMed, and Wiley-Blackwell Cochrane Library databases were searched for articles published in English from December 1, 2019, to September 21, 2020. The databases were searched for the following terms: COVID-19, severe acute respiratory syndrome coronavirus 2, SARS-CoV-2, coronavirus infections, novel coronavirus, cancer, neoplasms, tumor, leukemia, lymphoma, melanoma, carcinoma, sarcoma, oncology, surgery, surgical, neurosurgery, resection, perioperative care, perioperative period, and postoperative complications. The search terms were combined with "or" if they represented similar concepts and with "and" if they represented different concepts. The complete search strategies are detailed in Tables S1-S4.
## | study selection
Comparative studies eligible for this systematic review and metaanalysis consisted of those with reported outcomes of patients who were scheduled for oncologic surgery during the COVID-19 pandemic. Studies with fewer than five patients were excluded. For duplicate published studies from the same institution, only the most complete reports were included. In addition, the bibliographies of all studies and meta-analyses were searched to identify additional articles (i.e., backward snowballing). Abstracts, conference presentations, reviews, and expert opinions were excluded.
## | data extraction and endpoints
Excel software (Microsoft Corporation) was used for extraction of data from the studies we identified. Continuous variables were reported as mean (±SD) values, and categorical variables were expressed as frequencies. Data on study period, study center, country, type of cancer, type of study, and sample size were retrieved. The following patient characteristics were abstracted: age, male sex, body mass index (BMI), diabetes, renal insufficiency, coronary artery disease, acute kidney injury, peripheral artery disease, smoking history, pre-existing pulmonary disease, and dyslipidemia. Of note, smoking history data were not consistent across studies. Some authors reported this variable as "history of smoking," whereas other authors added more details, such as "former" versus "current" smoker.
Comorbidities (e.g., pre-existing pulmonary disease, dyslipidemia), smoking history, BMI, and sex were examined as separate variables, although the possible collinearity could not be assessed or ignored.
The primary endpoint was the COVID-19-related mortality rate, which was calculated for the entire patient population. The secondary endpoints were length of hospital stay and the rates of in-hospital
# | statistical analysis
For the short-term categorical and continuous outcomes, pooled event rates (PERs) and pooled means with their 95% confidence intervals (CIs) were calculated using the DerSimonian-Laird (inverse variance) method. Subgroup analysis was conducted to evaluate the primary endpoint according to the type of cancer.
Univariable meta-regression was performed to explore the relationship of the endpoints with preoperative characteristics. Each study was weighted according to the inverse of the variance of the estimate of that study, and between-study variance was estimated using a DerSimonian-Laird estimator. The meta-regression results were reported using regression coefficients (i.e., β), SEs, and p values.
Hypothesis testing for equivalence was two-tailed with a 0.05 significance level except for the subgroup analysis. A significance level of 0.06 was adopted. Heterogeneity assessment was performed based on the Cochran Q test with I 2 values. Individual study inference analysis was performed via leave-one-out sensitivity analysis for the primary endpoint. Funnel plots by graphical inspection and Egger regression testing were used for assessment of publication bias regarding the primary endpoint.
All statistical analyses were performed using the R computing language (version 3.6.2) and RStudio software with the meta and metafor packages.
# | results
# | overall results
We identified a total of 1936 studies in our database search. After exclusion of duplicates and irrelevant articles, we screened 1026 potentially relevant articles. We then assessed 72 full-text articles for eligibility. Twenty-eight studies with a total of 3508 patients met our inclusion criteria. The sample sizes ranged from 5 to 621. An outline of the systematic review process performed according to Preferred
Reporting Items for Systematic Reviews and Meta-Analyses guidelines is shown in [fig_ref] Figure 1: shows a forest plot of the primary endpoint of COVID-19-related mortality [/fig_ref]. The characteristics and demographics of the patients in the included studies are summarized in . We used the Newcastle-Ottawa Quality Assessment Scale for cohort studies to critically appraise the quality of the included studies .3.2 | Meta-analysis of the outcomes The PER for in-hospital COVID-19 infection was 3.00% (95% CI:
1.88%-4.73%) . The PER for all-cause mortality was 2.68% (95% CI: 1.23%-5.72%) , and that for surgery postponement because of COVID-19 infection was 2.80% (95% CI:
1.49%-5.18%) . The PER for overall COVID infection was 3.49% (95% CI: 2.34%-5.17%) . The PER for length of hospital stay was 7.26 days (95% CI: 5.03-10.48 days) .
The PER for hospital readmission was 2.74% (95% CI: 1.93%-3.88%) . The PER for postoperative complications was 11.44% (95% CI: 7.30%-17.48%) . The PER for ER visits was 2.18% (95% CI: 0.38%-11.51%) . The PER for surgical recovery was 92.03% (95% CI: 73.86%-97.92%) . The PER for COVID-19 infection recovery was 72.85% (95% CI: 61.84%-81.62%) . The PER for ICU admission was 3.82% (95% CI:
1.28%-10.87%) [fig_ref] Figure 1: shows a forest plot of the primary endpoint of COVID-19-related mortality [/fig_ref]. The PER for need for a ventilator was 9.85% (95% CI: 1.98%-37.20%) [fig_ref] Figure 1: shows a forest plot of the primary endpoint of COVID-19-related mortality [/fig_ref]. The PER for pulmonary complications was 5.96% (95% CI: 3.24%-10.71%) [fig_ref] Figure 1: shows a forest plot of the primary endpoint of COVID-19-related mortality [/fig_ref]. The patient outcomes are summarized in .
Subgroup analysis of the primary endpoint showed a nonsignificant trend of higher COVID-19-related mortality rates in patients with thoracic or lung cancer than in those with other cancers (subgroup difference, p = 0.30) [fig_ref] Figure 1: shows a forest plot of the primary endpoint of COVID-19-related mortality [/fig_ref].
## | meta-regression
BMI was significantly and negatively associated with all-cause mortality (p < 0.001) and in-hospital COVID-19 infection (p = 0.0064) rates. Also, male sex was positively and significantly associated with the COVID-19 infection rate (p = 0.0479) and an increased all-cause mortality rate (p < 0.001). Finally, preoperative respiratory disease (p = 0.0079) and smoking history (p = 0.0151) were positively and significantly associated with an increased all-cause mortality rate.
Meta-regression outcomes are summarized in . In these cohorts of cancer patients scheduled for oncologic surgery during the COVID-19 era, the calculated early mortality rate after COVID-19 infection was 27%. This aligns with published data demonstrating that the mortality rate may reach 20% in general surgical patients [bib_ref] Clinical characteristics and outcomes of patients undergoing surgeries during the incubation period..., Lei [/bib_ref] and even higher (25%) in cancer patients who undergo surgical intervention. [bib_ref] Mortality in patients with cancer and coronavirus disease 2019: A systematic review..., Saini [/bib_ref] The investigators in the latter study reported that the high mortality rate was attributed to COVID-19 infection in the patients. [bib_ref] Mortality in patients with cancer and coronavirus disease 2019: A systematic review..., Saini [/bib_ref] Over time, the mortality rate during the COVID-19 pandemic has improved, likely due to improved understanding of the infection trajectory and implementation of more optimal care algorithms. [bib_ref] Mortality in patients with cancer and coronavirus disease 2019: A systematic review..., Saini [/bib_ref] Furthermore, we report herein the PER for all-cause mortality which was 2.68%. These rates are similar to the COVID-19 mortality rate of 3.6% that was reported by Rajasekaran et al. [bib_ref] Patient safety associated with the surgical treatment of bone and soft tissue..., Rajasekaran [/bib_ref] Therefore, authors have suggested following international guidelines and implementing recommended adjustments to the surgical workflow for the purpose of overcoming the pandemic-related restrictions on the workflow. [bib_ref] Collateral damage: the impact on outcomes from cancer surgery of the COVID-19..., Sud [/bib_ref] We agree that safety should come first, so surgical scheduling may have to be adjusted according to the national situation and medical resources. Still, maintenance of safety should not compromise patient outcomes. We must also remember that cancer patients require special considerations and extra measures. [bib_ref] A practical approach to the management of cancer patients during the novel..., Ho [/bib_ref] In our meta-analysis, the PER for postoperative complications was 11%. Authors reported a postoperative complication rate of 12%
# | discussion
in patients with gynecologic cancer, [bib_ref] Performing gynecologic cancer surgery during the COVID-19 pandemic in Turkey: a multicenter..., Dursun [/bib_ref] and Lisa et al. [bib_ref] Breast reconstruction in a coronavirus disease, Lisa [/bib_ref] reported a lack of a significant difference in the complication rate after breast reconstruction before and after the start of the COVID-19 pandemic. Our analysis also showed that BMI is negatively associated with all-cause mortality and COVID-19 infection rates in cancer patients.
This aligns with published data demonstrating that higher BMI or obesity contributes to higher morbidity and mortality rates after COVID-19 infection. [bib_ref] High prevalence of obesity in severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) requiring..., Simonnet [/bib_ref] [bib_ref] More severe obesity leads to more severe COVID-19 in study, Kuehn [/bib_ref] [bib_ref] Obesity and COVID-19: what makes obese host so vulnerable?, Mohammad [/bib_ref] [bib_ref] T lymphopaenia in relation to body mass index and TNF-alpha in human..., Tanaka [/bib_ref] [bib_ref] Changes in nutritional status impact immune cell metabolism and function, Alwarawrah [/bib_ref] The detrimental effect of obesity after COVID-19 infection may be attributed to the observed negative effects of obesity of impaired immune function and decreased lung capacity and reserve. [bib_ref] High prevalence of obesity in severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) requiring..., Simonnet [/bib_ref] [bib_ref] T lymphopaenia in relation to body mass index and TNF-alpha in human..., Tanaka [/bib_ref] [bib_ref] Changes in nutritional status impact immune cell metabolism and function, Alwarawrah [/bib_ref] [bib_ref] Global pandemics interconnected-obesity, impaired metabolic health and COVID-19, Stefan [/bib_ref] Smoking is another major risk factor for poor oncologic sur- Investigators have observed that delaying or canceling surgery for cancer may impact mortality, as well. Also, as we acknowledged earlier, cancer patient characteristics overlap (sex, BMI, comorbidities, smoking history, and even age). We collected and analyzed each of these variables separately, but the overlap of these factors is impossible to ignore and assess. Furthermore, we sought to present as much valuable data as possible, but we were challenged by the heterogeneous definitions of some outcomes in the studies (i.e., recovery rates). Also, a point to consider is the possible impact of missing data on the patients who underwent treatment outside a hospital or were lost to follow-up, especially when it comes to the impact of missing data on recovery rates.
During the first wave of the COVID-19 pandemic, the rate of oncologic surgical procedures decreased by 80%. [bib_ref] Cancer surgery scheduling during and after the COVID-19 first wave: The MD..., Tzeng [/bib_ref] Furthermore, authors reported that 44% of cancer patients needed ICU admission. [bib_ref] Clinical characteristics and outcomes of patients undergoing surgeries during the incubation period..., Lei [/bib_ref] These rates have improved greatly since that time. Balancing the competing requirements of safety, timely oncologic care, and optimal use of operation units, ERs, and ICUs is always essential before setting workflows for oncologic surgical care. [bib_ref] Management of cancer surgery cases during the COVID-19 pandemic: considerations, Bartlett [/bib_ref] [bib_ref] Approaching surgical triage during the COVID-19 pandemic, Brindle [/bib_ref] [bib_ref] Personal protective equipment and COVID-19: a review for surgeons, Stewart [/bib_ref] [bib_ref] Cancer patients in SARS-CoV-2 infection: a nationwide analysis in China, Liang [/bib_ref] [bib_ref] Safety recommendations for evaluation and surgery of the head and neck during..., Givi [/bib_ref] [bib_ref] Surgical prehabilitation in patients with cancer: state-of-the-science and recommendations for future research..., Carli [/bib_ref] We must face the reality that the COVID-19 pandemic may be prolonged, as we are still learning about the course of the disease, and not everyone has access to and/or received the vaccine. Moreover, new strains are emerging, and the vaccine does not provide 100% protection. Data demonstrate that cancer does not appear to be associated with an elevated risk of COVID-19 infection as long as preventive measures are taken for selected patients. [bib_ref] Breast cancer surgery during the COVID-19 pandemic peak in the UK: operative..., Macinnes [/bib_ref] However, COVID-19 infection is always a risk for cancer patients during preparation for surgery.
Therefore, we extensively reviewed published data to guide effective clinical care for cancer patients who need surgical interventions. As the rates of COVID-19 infection decline and access to vaccines expands, our hope is that this progress enables full oncologic surgical capacity while maintaining the safety of patients and providers and prioritization of the oncologic surgical workflow based on institutional resources and the national situation. [bib_ref] Plastic surgery and the COVID-19 pandemic: a review of clinical guidelines, Ozturk [/bib_ref] [bib_ref] Epidemiology and transmission of COVID-19 in 391 cases and 1286 of their..., Bi [/bib_ref] Our data reported herein constitute a reliable, solid resource to help oncologic surgeons better understand the trajectory and outcomes of COVID-19 infection and identify cancer patients at increased risk for poor surgical outcomes.
[fig] Figure 1: shows a forest plot of the primary endpoint of COVID-19-related mortality. The PER for this endpoint was 27.15% (95% CI:18.38%-38.16%). A funnel plot of the results of publication bias assessment and leave-one-out analysis is shown inFigure 2. [/fig]
|
Visual Function and Brief Cognitive Assessment for Multiple Sclerosis in Optic Neuritis Clinically Isolated Syndrome Patients
Background: In this study, we hypothesized that clinically isolated syndrome-optic neuritis patients may have disturbances in neuropsychological functions related to visual processes. Methods: Forty-two patients with optic neuritis within 3 months from onset and 13 healthy controls were assessed at baseline and 6 months with MRI (brain volumes, lesion load, and optic radiation lesion volume) and optical coherence tomography (OCT) (peripapillary retinal nerve fiber layer [RNFL], ganglion cell and inner plexiform layers [GCIPLs], and inner nuclear layer). Patients underwent the brief cognitive assessment for multiple sclerosis, high-contrast and low-contrast letter acuity, and color vision. Results: At baseline, patients had impaired visual function, had GCIPL thinning in both eyes, and performed below the normative average in the visual-related tests: Symbol Digit Modalities Test and Brief Visuospatial Memory Test-Revised (BVMT-R). Over time, improvement in visual function in the affected eye was predicted by baseline GCIPL (P = 0.015), RNFL decreased, and the BVMT-R improved (P = 0.001). Improvement in BVMT-R was associated with improvement in the high-contrast letter acuity of the affected eye (P = 0.03), independently of OCT and MRI metrics. Conclusion: Cognitive testing, assessed binocularly, of visuospatial processing is affected after unilateral optic neuritis and improves over time with visual recovery. This is not related to structural markers of the visual or central nervous system.
O ptic neuritis (ON) is a frequent presentation of demyelinating clinically isolated syndrome (CIS) [bib_ref] Optic neuritis, Jenkins [/bib_ref]. Research studies often include cognitive outcomes when investigating disability accrual and risk for multiple sclerosis (MS) conversion in CIS cohorts. Although most studies have found deficits in information processing speed and visuospatial memory [bib_ref] Cognitive impairment at the onset of multiple sclerosis: relationship to lesion location, Reuter [/bib_ref] [bib_ref] Longitudinal MRI and neuropsychological assessment of patients with clinically isolated syndrome, Uher [/bib_ref] , the prevalence of cognitive impairment in CIS patients varies from one study to another [bib_ref] Cognitive impairment in multiple sclerosis with regards to disease duration and clinical..., Brochet [/bib_ref]. One explanation is that researchers do not always assess vision and damage in the visual pathways when testing cognition in CIS patients with ON. In this study, we hypothesize these factors can influence visiondependent cognitive tests in ON-CIS patients.
MS researchers have shown relationships between visual function and cognitive performance, for tests depending on visual [bib_ref] The influence of corrected visual acuity on visual attention and incidental learning..., Davis [/bib_ref] [bib_ref] Visual acuity is associated with performance on visual and non-visual neuropsychological tests..., Feaster [/bib_ref] [bib_ref] Low contrast visual acuity testing is associated with cognitive performance in multiple..., Wieder [/bib_ref] [bib_ref] Visual pathway measures are associated with neuropsychological function in multiple sclerosis, Nguyen [/bib_ref] and nonvisual inputs [bib_ref] Low contrast visual acuity testing is associated with cognitive performance in multiple..., Wieder [/bib_ref] [bib_ref] Visual pathway measures are associated with neuropsychological function in multiple sclerosis, Nguyen [/bib_ref]. However, these studies included heterogeneous cohorts for past ON history [bib_ref] The influence of corrected visual acuity on visual attention and incidental learning..., Davis [/bib_ref] [bib_ref] Visual acuity is associated with performance on visual and non-visual neuropsychological tests..., Feaster [/bib_ref] [bib_ref] Low contrast visual acuity testing is associated with cognitive performance in multiple..., Wieder [/bib_ref] [bib_ref] Visual pathway measures are associated with neuropsychological function in multiple sclerosis, Nguyen [/bib_ref]. Furthermore, they assessed patients with long disease duration [bib_ref] Visual pathway measures are associated with neuropsychological function in multiple sclerosis, Nguyen [/bib_ref] and other MS phenotypes (i.e., relapsing-remitting and progressive). Therefore, they may have observed the effects of the central nervous system (CNS) damage accrual on both vision and cognitive functions.
Markers of retinal damage derived from optical coherence tomography (OCT), namely, peripapillary retinal nerve fiber layer (pRNFL) and combined ganglion cell and inner plexiform layers (GCIPLs), have inconsistent associations with cognitive outcomes [bib_ref] Visual pathway measures are associated with neuropsychological function in multiple sclerosis, Nguyen [/bib_ref] [bib_ref] Retinal nerve fiber layer atrophy is associated with physical and cognitive disability..., Toledo [/bib_ref] [bib_ref] Cognitive impairment in patients with multiple sclerosis is associated with atrophy of..., Coric [/bib_ref] [bib_ref] Peripapillary retinal nerve fibre layer as measured by optical coherence tomography is..., Bsteh [/bib_ref] [bib_ref] Cognitive impairment and optic nerve axonal loss in patients with clinically isolated..., Anhoque [/bib_ref]. Some studies included patients with a previous ON [bib_ref] Visual pathway measures are associated with neuropsychological function in multiple sclerosis, Nguyen [/bib_ref] [bib_ref] Cognitive impairment and optic nerve axonal loss in patients with clinically isolated..., Anhoque [/bib_ref]. Hence, their findings may have been affected by pre-existing retinal layer changes. Other studies excluded eyes with previous ON [bib_ref] Retinal nerve fiber layer atrophy is associated with physical and cognitive disability..., Toledo [/bib_ref] [bib_ref] Peripapillary retinal nerve fibre layer as measured by optical coherence tomography is..., Bsteh [/bib_ref] , but, by including patients with long disease duration, and other factors, such as brain atrophy or white matter (WM) lesion accrual, may have influenced their results. A recent study [bib_ref] Cognitive impairment in patients with multiple sclerosis is associated with atrophy of..., Coric [/bib_ref] grouped the patients according to the ON history and reported an association between GCIPL thinning and cognitive impairment evident only in the group without a history of ON.
In this study, we studied only patients with unilateral ON-CIS and analyzed affected (AEs) and nonaffected eyes (NAEs) separately. We assessed patients within 3 months after the onset of ON and at 6 months, after visual recovery [bib_ref] The course of visual recovery after optic neuritis, Beck [/bib_ref] , and when ON-related retinal thinning is evident [bib_ref] Early retinal atrophy predicts long-term visual impairment after acute optic neuritis, Sanchez-Dalmau [/bib_ref] [bib_ref] Timing of retinal neuronal and axonal loss in MS: a longitudinal OCT..., Balk [/bib_ref]. We used MRI to evaluate the possible contribution of damage to the posterior visual pathways and the CNS. We chose the brief cognitive assessment for MS (BICAMS) [bib_ref] Recommendations for a brief international cognitive assessment for multiple sclerosis (BICAMS), Langdon [/bib_ref] because it includes visually and nonvisually dependent tests.
# Methods
We recruited 42 patients with ON within 3 months from onset, from the National Hospital of Neurology and Neurosurgery and the Moorfields Eye Hospital, London, United Kingdom. Exclusion criteria were a history of past neurological episodes or previous ON episodes, antibodies against aquaporin-4 and myelin oligodendrocyte glycoprotein (routinely assessed), other medical conditions potentially affecting the CNS, including cardiovascular risk factors, and the eye, and high refractive errors (.26.0 or +6.0 D). We also recruited 13 healthy controls (HCs).
Clinical assessments (patients alone), OCT, and MRI were performed at baseline and after 6 months.
All subjects gave written informed consent (Study Ref: 13/LO/1762; 13/0231-CIS2013).
We scored physical disability using the Expanded Disability Status Scale. For visual function, we assessed each eye separately. We scored: high-contrast letter acuity (HCLA) using logMAR high contrast (100%) Early Treatment Diabetic Retinopathy Study charts at 4 m, low-contrast letter acuity (LCLA) (2.5% and 1.25%) as the total number of letters read correctly using Sloan letter charts at 2 m, and the color vision as the total error score (TES) at the Farnsworth-Munsell test [bib_ref] The Farnsworth-Munsell 100 hue test in the first episode of demyelinating optic..., Ménage [/bib_ref].
We used BICAMS that includes the Symbol Digit Modalities Test (SDMT) for information processing speed, the California Verbal Learning Test-II (CVLT-II) for verbal memory, and the Brief Visuospatial Memory Test-Revised (BVMT-R) for visuospatial memory.
We recorded years of education for each patient and calculated z-scores using the data set provided by the BICAMS initiative (https://www.bicams.net) [bib_ref] The utility of regression-based norms in interpreting the minimal assessment of cognitive..., Parmenter [/bib_ref]. In all tests, participants used their habitual distance corrective lenses. For HCLA, subjects also used pin-hole correction with their habitual lenses, and we used the best-corrected scores.
We used spectral-domain OCT (Spectralis v.1.7.1.0, Heidelberg Engineering) with eye tracking for measurement accuracy (21) without pharmacological pupil dilatation. For each eye, we acquired 1) three 3.4 mm peripapillary circular scans (automatic real time [ART] 100) manually centered around the optic nerve with the highest quality chosen for analysis and 2) a macular volume scan-we manually centered the scan around the fovea (vertical alignment, ART 10-25). We used the baseline peripapillary and macular scans as references for the follow-up scans.
We obtained individual macular layers with an automated segmentation software (HRA/Spectralis Viewing Module version 5.6.4.0). We performed quality control (QC) according to OSCAR-IB criteria [bib_ref] The OSCAR-IB consensus criteria for retinal OCT quality assessment, Tewarie [/bib_ref] and excluded the scans failing the QC. In the case of minor failures of the automated segmentation, whenever possible, we performed manual correction.
For pRNFL thickness, we used the global average of the thickness. We used a 1-3-6 mm grid on the thickness map for the macular scan, selecting the 1-3-mm ring values for our analysis. We combined the ganglion cell layer and the inner plexiform layer in the GCIPL [bib_ref] On Behalf of the IMSVISUAL Consortium. The APOSTEL recommendations for reporting quantitative..., Cruz-Herranz [/bib_ref] , and we recorded the values of the inner nuclear layer.
For each subject, we used the OCT measurements from each eye.
We used a 3T Philips Achieva system (Philips Medical Systems, Best, Netherlands).
The MRI protocol included 2D short-tau inversionrecovery coronal sequences of the orbits; 3D T1-weighted turbo field echo, 3D fluid-attenuated inversion recovery, and 2D proton density (PD)/T2-weighted turbo spin-echo of the brain; sagittal PD/T2-weighted images of the spinal cord; and pregadolinium and postgadolinium T1-weighted imaging of the orbits, brain, and spinal cord (patients at baseline alone).
We determined lesion dissemination in space and time according to the revised 2017 McDonald criteria (24) for MS diagnosis.
We outlined WM lesions on the 2D PD/T2-weighted images using JIM v6.0 (Xinapse systems) and computed lesion volumes.
The lesion masks were coregistered to the 3D-T1 images (25) for lesion filling [bib_ref] A multi-time-point modality-agnostic patch-based method for lesion filling in multiple sclerosis, Prados [/bib_ref]. Subsequently, we used Geodesic Information Flows (27) for brain extraction and tissue segmentation.
At baseline, after registering the 3D-T1 images in the Montreal Neurological Institute (MNI152) space, we used the Juelich Histological Atlas (https://fsl.fmrib.ox.ac.uk/fsl/ fslwiki/Atlases/Juelich) to create binary optic radiation masks. We then obtained the volume of the lesions in the optic radiations.
We assessed the percentage of brain volume change (PBVC) with SIENA [bib_ref] Accurate, robust, and automated longitudinal and cross-sectional brain change analysis, Smith [/bib_ref].
We assessed group differences between CIS-ON patients and HCs for demographic characteristics using the 2-sample t test for continuous variables and the chi-square test for categorical variables.
We conducted 3 statistical analyses.
1. At baseline, group differences in brain volumes and brain parenchymal fraction (BPF) were assessed with linear regression adjusting for age and sex.
We used mixed effect models to assess the differences in OCT metrics between patients' NAEs and HCs' eyes adjusting for age, sex, months from onset, ethnicity (categorized as Caucasian and non-Caucasian). We also tested the effect of lesion volumes (WM and optic radiation) and BPF on the model.
In patients, we used the same model for the differences in OCT metrics and visual outcomes between patients' AEs and NAEs adjusting for steroid treatment at the onset.
## For the longitudinal analysis, group differences in pbvc
were assessed with linear regression adjusting for age and sex.
We used mixed effect models to assess differences in changes over time in OCT metrics between patients' NAEs and HCs' eyes adjusting for age, sex, and months from onset. However, we also added time and group · time interaction as explanatory variables. If significantly different, PBVC was added to the model.
In patients, we used a similar model to assess the difference between eyes in changes in visual outcomes and OCT metrics.
In patients, a similar model, not stratified by eyes, was used for BICAMS scores, also adjusting for education. We used raw scores as they attain the same significance level of z-scores [bib_ref] The utility of regression-based norms in interpreting the minimal assessment of cognitive..., Parmenter [/bib_ref] , whereas we used Z-scores for descriptive purposes.
If a variable x showed a significant change over time, we calculated the difference between 6 months and baseline (labeled as Dx). We used the Dvariables in the subsequent analysis.
1. We applied multiple linear regression models, adjusted for age and sex, to assess the effects of OCT metrics, MRI parameters, and visual scores on cognition.
At baseline, each cognitive score was entered separately as a response variable. Predictors were OCT metrics, if significantly different between patients and controls or between AEs and NAEs, and visual outcomes, if significantly different between patients' eyes. We also explored the effect of BPF and lesion volume on the model. For the longitudinal model, we entered only significant D cognitive scores as response variables in the model. The significant DOCT metrics and Dvisual outcomes were entered separately as predictors. We explored the effect of PVBC and lesion volume change on the model.
Statistical analyses were performed with Stata v.14.1 (Stata Corporation, College Station, TX).
Only results associated with P , 0.05 were considered statistically significant and subsequently reported in this article. Because of the exploratory nature of the study, correction for multiple comparisons was not performed.
# Results
Thirty patients and 13 HCs completed the study at 6 months. Patients and HCs did not differ significantly for age, sex, brain volumes, and PBVC [fig_ref] TABLE 1: Demographic characteristics and brain volumes of patients and healthy controls at baseline... [/fig_ref]. None of the patients had comorbidities or were taking treatments altering their mental states.
Thirty-six (86%) of 42 baseline CIS patients presented with brain lesions and 30 of them had lesions involving the optic radiations [fig_ref] TABLE 2: Clinical and MRI characteristics of patients at baseline and 6 months [/fig_ref].
After quality control of OCT acquisitions, we discarded the following patients' OCT data: 9/84 (11%) baseline and 4/60 (7%) 6-month pRNFL; 8/84 (10%) baseline and 3/60 (6%) 6-month macular scans. HCs' OCT scans passed the QC.
The AE GCIPL was significantly lower than NAE GCIPL (coeff. = 211.5 mm, P , 0.0001), which was lower than in HCs (coeff. = 27.7 mm, P = 0.002) [fig_ref] FIG. 1: Visual outcomes [/fig_ref]. The Raw score 60.6 (9) 67 (6)
## Gcipl in the naes was not associated with the lesion volumes or bpf.
AEs had significant deficits for HCLA (coeff. = 0.12 logMAR, P , 0.0001), 2.5% and 1.25% LCLA (coeff. = 215.3 and coeff. = 210.2 letters, respectively; P , 0.0001), and color vision TES (coeff. = 100, P , 0.0001) compared with NAEs [fig_ref] FIG. 2: Optical coherence tomography metrics [/fig_ref].
At baseline, although all patients performed in the normative average for CVLTII, 10 (24%) patients and 4 (10%) showed deficits (,1.5 z-score) for SDMT and BVMT-R, respectively. Two (5%) patients had ,1.5 z-score in both tests matching the definition for cognitive impairment [fig_ref] FIG. 2: Optical coherence tomography metrics [/fig_ref].
The pRNFL decreased significantly in the AEs (coeff. = 213.05 logMAR, P = 0.02), whereas the GCIPL did not change significantly [fig_ref] Table 3 ,: Fig [/fig_ref] , [fig_ref] FIG. 2: Optical coherence tomography metrics [/fig_ref].
All the visual scores in the AEs improved ( The improvement in the BVMT-R score was significantly associated with the improvement in the AE HCLA (R 2 = 0.17; B = 219.9; 95 CI = 238.3 to 21.5, P = 0.03) [fig_ref] FIG. 4: Scatter plots of D bvmt-r and Dvisual scores in the affected eyes [/fig_ref]. The PBVC did not affect the significance of the correlation (P = 0.045), whereas the lesion volume change did (P = 0.07). PVBC and lesion volume change did not affect the other correlations between DBVMT-R and DCVLTII and Dvisual and DOCT parameters. Original Contribution CONCLUSION Patients with clinically isolated, unilateral ON performed worse in BICAMS visually dependent tests (BVMT-R and SDMT) than in nondependent (CVLTII). Over time, both the CVLTII and BVMT-R had an improvement possibly influenced by the general relapse recovery.
We did not find correlations between these changes and most of our visual outcomes. However, the BVMT-R improvement was associated with the recovery of visual function (HCLA). We believe this result is of clinical interest, particularly as it was independent from brain atrophy.
The correlation between BVMT-R improvement and AE HCLA recovery is interesting as the NAE remains visually noncompromised. It is possible, however, that higher visual processing, tested in this study by BVMT-R, depends on binocular afferent stability and/or symmetry. When this is disrupted, there is impairment of visually dependent cognitive performance. As the AE improves, binocular visual symmetry is restored and visual cognitive performance normalizes. Lesion accumulation may have an impact on this process. The mechanisms for this are speculative but may include higher visual neuroplastic changes. Furthermore, bidirectional relationships between vision and cognition may occur [bib_ref] Low contrast visual acuity testing is associated with cognitive performance in multiple..., Wieder [/bib_ref].
LCLA, although recovering significantly, remained impaired in patients' AEs, but we could not find correlations with SDMT or BVMT-R, in contrast to previous studies [bib_ref] Visual pathway measures are associated with neuropsychological function in multiple sclerosis, Nguyen [/bib_ref]. These studies, however, showed a correlation between LCLA and cognition, independently of the ON history. As LCLA has been related to CNS damage [bib_ref] Validity of low-contrast letter acuity as a visual performance outcome measure for..., Balcer [/bib_ref] , particularly global and regional brain atrophy [bib_ref] Relationship of optic nerve and brain conventional and non-conventional MRI measures and..., Frohman [/bib_ref] , in MS patients with long disease duration, LCLA may be related to cognitive performance as a surrogate marker of neurodegeneration and not just of visual impairment. This would also explain similar results in neurodegenerative diseases [bib_ref] Vision and cognition in Alzheimer's disease, Rizzo [/bib_ref] [bib_ref] Visual-spatial ability in Parkinson's disease, Crucian [/bib_ref]. In our cohort, instead, in the absence of brain atrophy, LCLA could reflect visual pathway damage more than CNS alterations, as shown by the correlation with GCIPL. We found no associations between BICAMS scores and OCT metrics. The presence of demyelination, inflammation, and neuroaxonal damage at baseline may have different and possibly opposing effects on OCT masking potential correlations with BICAMS. Longitudinal studies are required to assess if, with the disease progression, relationships arise.
Finally, SDMT did not improve over time. Deficits in information processing speed are known to characterize CIS cognitive impairment (5). SDMT did not correlate with visual function or damage to visual pathways. This suggests that other mechanisms, possibly related to brain function, drive SDMT performances, as previously shown in early CIS patients [bib_ref] Single-subject structural cortical networks in clinically isolated syndrome, Collorone [/bib_ref].
In NAEs, the GCIPL was thinner than in HCs eye. At present, there is no imaging gold standard for ON [bib_ref] Advanced MRI of the optic nerve, Hoch [/bib_ref]. Therefore, one hypothesis is that there was sub-clinical nerve involvement. As the GCIPL did not correlate with lesion volumes, this reduces the chances of trans-synaptic degeneration from optic radiations. As brain volumes were neither altered nor related to NAEs GCIPL, another hypothesis is that this thinner GCIPL may represent a manifestation of early MS neurodegeneration not yet detectable in the brain [fig_ref] TABLE 5: Associations between significant ΔBICAMS scores and significant ΔOCT metrics and Δvisual outcomes... [/fig_ref].
## Our study has several limitations
We lost 12 patients at follow-up. However, our missed data points were stochastic, so, using mixed effect models, we could adjust for the missing data.
We could not use a non-ON CIS cohort to assess the generalizability of our findings. However, ON is a frequent CIS onset, and we believe that our results may be of interest for neurologists and researchers assessing cognition in CIS patients.
For BICAMS, we did not account for practice effects. However, we assessed intracohort relationships between visual and cognitive metrics; the practice effect should not explain the associations between HCLA and BVMT-R changes.
The patients in our cohort had relatively high lesion volumes. However, we had excluded alternative diagnoses, such as aquaporin 4 and myelin oligodendrocyte Ab conditions. Furthermore, we explored the effect of lesion volume variability in the study. Finally, the variable time from onset to recruitment among participants was another potential limitation. However, using for the first time MRI metrics to investigate the correlations between cognition and visual metrics, it is understandably challenging to assess patients at the exact onset of the ON.
In conclusion, our findings suggest that the visual acuity should be considered when BICAMS is administered to ON patients to interpret BVMT-R scores. This can have implications in clinical trials, where cognitive scores are often used, and vision is not measured.
[fig] FIG. 1: Visual outcomes. **P , 0.0001. Results from mixed effect models adjusted for age, sex, steroids intake, and months from onset. For changes overtime, an interaction group x time was used. Av, average; BL, baseline; 6 M, six months; FM, Farnsworth-Munsell test; TES, total error score; LCLA, low-contrast letter acuity. [/fig]
[fig] FIG. 2: Optical coherence tomography metrics. **P , 0.0001 *P , 0.05. Results from mixed effect models adjusted for age, sex, steroids intake, and months from onset. For changes overtime, an interaction group x time was used. Av, average; BL, baseline; 6 M, 6 months; pRNLF, peripapillary retinal nerve fiber layer; GCIPL, combined ganglion cell and inner plexiform layers; INL, inner nuclear layer.e26 Collorone et al: J Neuro-Ophthalmol 2022; 42: e22-e31 [/fig]
[fig] FIG. 3: Cognitive outcomes in patients. *P , 0.05. A. Scatter plot of cognitive test scores for each subject and each time point (B) results from mixed effect models adjusted for age, sex, and months from onset; (C) prevalence of cognitive impairment (defined as z-score ,21.5) in the cognitive tests at baseline. BVMT-R, brief visuospatial memory test-revised; CVLTII, California verbal learning test II; SDMT, Symbol Digit Modality Test. [/fig]
[fig] FIG. 4: Scatter plots of D bvmt-r and Dvisual scores in the affected eyes. *P , 0.05 results from linear regression adjusted for age, sex, education, and months from onset. BVMT-R, Brief Visuospatial Memory Test-Revised; FM TES, Farnsworth-Munsell total error score; HCLA, high-contrast letter acuity; LCLA, low-contrast letter acuity. [/fig]
[table] TABLE 1: Demographic characteristics and brain volumes of patients and healthy controls at baseline Linear regression adjusting for age and sex. HCs, healthy controls; Gm, gray matter; ORs, optic radiations; Wm, white matter; vol., volume. [/table]
[table] TABLE 2: Clinical and MRI characteristics of patients at baseline and 6 months [/table]
[table] Table 3 ,: Fig. 1). The LCLA 1.25% improvement was predicted by thicker baseline GCIPL in the AEs (RC [95% CIs]: 0.33 letter/ mm, [0.07, 0.58], P = 0.015) (Table 4;Fig. 1).Baseline BICAMS scores correlated neither with BPF, brain, and lesion volumes (See Supplemental Digital Content 1,Table E1, http://links.lww.com/WNO/A466) nor with visual and OCT metrics. PVBC and lesion volume change did not correlate with DBVMT-R and DCVLTII (See Supplemental Digital Content 2, Table E1, http://links.lww.com/WNO/A466). [/table]
[table] TABLE 3: Optical coherence tomography metrics in patients and healthy controls *Mixed effect model adjusted for age, sex, steroids months from onset, and steroids comparing affected eyes and nonaffected eyes. † Mixed effect model adjusted for age, sex, and months from onset comparing patients' nonaffected eyes with the average of healthy controls eyes. ‡ Mixed effect model adjusted for age, sex, steroids, and months from onset comparing baseline with 6-month values. BL, baseline; GCIPL, combined ganglion cell and inner plexiform layers; HCs, healthy controls; pRNFL, peripapillary retinal nerve fiber layer; INL, inner nuclear layer. [/table]
[table] TABLE 4: Visual outcomes in patients [/table]
[table] TABLE 5: Associations between significant ΔBICAMS scores and significant ΔOCT metrics and Δvisual outcomes P value significance for bold entries. *Results are from linear regression models adjusted for age, sex, and education and months from onset. AE, affected eye; BICAMS, Brief Cognitive Assessment for Multiple Sclerosis; BVMT-R,Brief Visuospatial Memory Test-Revised; CIs, confidence intervals; CVLT-II, California Verbal Learning Test-II; HCLA, high-contrast letter acuity; LCLA, low-contrast letter acuity; pRNFL, peripapillary retinal nerve fiber layer; SDMT, Symbol Digit Modalities Test; TES, total error score (Farnsworth-Munsell test). [/table]
|
New Insights and Challenges Associated With IgA Vasculitis and IgA Vasculitis With Nephritis—Is It Time to Change the Paradigm of the Most Common Systemic Vasculitis in Childhood?
What are the challenges ahead and how have we responded so far when it comes to the non-granulomatous systemic vasculitis, characterized mainly by deposits of IgA immune complexes in the endothelium of small blood vessels-IgA vasculitis (IgAV)? That is the question to which we tried to answer. We summarized existing knowledge about epidemiology, pathogenesis, genetics, diagnostic tests and therapy in this somewhat neglected entity in pediatric rheumatology. Since etiopathogenesis of IgA vasculitis is complex, with factors other than galactose-deficient IgA 1 -containing immune complexes also being important, and may involve numerous interactions between environmental and genetic factors, genomics alone cannot explain the entirety of the risk for the disease. The incidence of IgAV and nephritis varies worldwide and may be a consequence of overlapping genetic and environmental factors. In addition to the role of the HLA class II genes, some studies have pointed to the importance of non-HLA genes, and modern geostatistical research has also indicated a geospatial risk distribution, which may suggest the strong influence of different environmental factors such as climate, pathogen load, and dietary factors. The application of modern geostatistical methods until recently was completely unknown in the study of this disease, but thanks to the latest results it has been shown that they can help us a lot in understanding epidemiology and serve as a guide in generating new hypotheses considering possible environmental risk factors and identification of potential genetic or epigenetic diversity. There is increasing evidence that an integrative approach should be included in the understanding of IgA vasculitis, in terms of the integration of genomics, proteomics, transcriptomics, and epigenetics. This approach could result in the discovery of new pathways important for finding biomarkers that could stratify patients according to the risk of complications, without an invasive kidney biopsy which is still the gold standard to confirm a diagnosis of nephritis, even if biopsy findings interpretation is not uniform in clinical practice. Ultimately, this will allow the development of new therapeutic approaches, especially important in the treatment of nephritis, for which there is still no standardized treatment.
# Introduction
IgA vasculitis (IgAV) is a non-granulomatous systemic vasculitis, histologically characterized by infiltration of the walls of the blood vessels, mainly arterioles, capillaries and venules, by neutrophils with deposits of immune complexes containing predominantly IgA [bib_ref] EULAR/PRINTO/PRES criteria for Henoch-Schonlein purpura, childhood polyarteritis nodosa, childhood Wegener granulomatosis and..., Ozen [/bib_ref] [bib_ref] Pediatric vasculitis, Barut [/bib_ref] [bib_ref] Vasculitis: decade in review, Demir [/bib_ref]. The endothelium of small blood vessels in the skin, synovial membrane, gut, and kidneys is usually involved [bib_ref] EULAR/PRINTO/PRES criteria for Henoch-Schonlein purpura, childhood polyarteritis nodosa, childhood Wegener granulomatosis and..., Ozen [/bib_ref].
Even though IgAV is the most common form of vasculitis in childhood [bib_ref] Childhood vasculitis, Ozen [/bib_ref] , it is somewhat neglected in pediatric rheumatology, as it is mostly perceived as a self-limited disease lasting up to 4 weeks, with care for these patients scattered between nephrologists, rheumatologists, dermatologists, and gastroenterologists [bib_ref] Different histological classifications for Henoch-Schönlein purpura nephritis: which one should be used?, Jelusic [/bib_ref].
It is important to keep in mind that despite the favorable prognosis for most pediatric patients, various acute and chronic complications are possible. Among the acute complications of IgAV, the most frequent are those related to the gastrointestinal system, including bleeding, intussusception, and bowel perforation as the most serious ones [bib_ref] Henoch-Schönlein purpura in children, Trnka [/bib_ref]. The most important chronic complication and the main cause of morbidity and mortality among children suffering from IgAV is the renal aspect of the disease (IgAV nephritis, IgAVN), which therefore represents the main prognostic factor. IgAVN occurs in 20-60% of children suffering from IgAV, and among them chronic renal failure has been reported in 1-15% [bib_ref] Henoch-Schönlein purpura nephritis in children, Davin [/bib_ref] [bib_ref] Risk of long term renal impairment and duration of follow up recommended..., Narchi [/bib_ref] [bib_ref] Long-term followup of childhood Henoch-Schönlein nephritis, Goldstein [/bib_ref] [bib_ref] Long-term prognosis of Henoch-Schönlein nephritis in adults and children. Italian Group of..., Coppo [/bib_ref] [bib_ref] Clinical outcome of Schönlein-Henoch purpura nephritis in children, Schärer [/bib_ref].
We will try to show how new insights has been gained regarding this disease, including epidemiology, pathogenesis, genetics, diagnostic trials, and therapy, as well as to point out the fact that many questions and dilemmas concerning IgAV remain unanswered.
## New insights in the epidemiology of igav
It is known that the incidence of IgAV varies worldwide ranging from 3 to 55.9 cases per 100,000 children [fig_ref] FIGURE 1 |: Distribution of incidence of IgAV around the world [/fig_ref] , while the prevalence varies between 6.1 and 20.4 per 100,000 children [bib_ref] Incidence of IgA vasculitis in children estimated by four-source capture-recapture analysis: a..., Piram [/bib_ref] [bib_ref] Advances in our understanding of the patogenesis of Henoch-Schönlein purpura and the..., Park [/bib_ref] [bib_ref] Ten-year nationwide population-based survey on the characteristics of children with Henoch-Schönlein purpura..., Shim [/bib_ref]. The fact that IgAV is not equally present in different parts of the world is also well-known, since it has the highest occurrence in East Asians, intermediate in Europeans, and the lowest in individuals of African ancestry [bib_ref] Incidence of IgA vasculitis in children estimated by four-source capture-recapture analysis: a..., Piram [/bib_ref]. Nevertheless, only recently a study on spatial variability of the incidence of IgAV and IgAVN using modern geostatistical methods has been performed [bib_ref] Geospatial clustering of childhood IgA vasculitis and IgA vasculitis-associated nephritis, Sapina [/bib_ref]. This research is important for several reasons. First, there is a lack of application of geostatistical methods in the field of pediatric rheumatology at all. Most of the previous work explores potential risk factors using observational studies with classical statistical techniques resulting in reduction of information which may be used to deepen the knowledge of potential risk factors involved in the pathomechanism of disease, in this case IgAV. In other words, the geospatial analysis may provide useful information of genetic, socio-economic, and environmental risk factors while simultaneously taking into account their spatial diversity, which can be applied not only in the case of IgAV and IgAVN but also in many other diseases. Second, it was demonstrated that both IgAV and IgAVN may not be randomly distributed in space, but clustered, similar to some other non-communicable diseases, such as inflammatory polyarthritis, heart diseases, and diabetes [bib_ref] Geospatial clustering of childhood IgA vasculitis and IgA vasculitis-associated nephritis, Sapina [/bib_ref] [bib_ref] Changes in the geographic patterns of heart disease mortality in the United..., Casper [/bib_ref] [bib_ref] The existence of geographical clusters of cases of inflammatory polyarthritis in a..., Silman [/bib_ref] [bib_ref] Feasibility study of geospatial mapping of chronic disease risk to inform public..., Noble [/bib_ref].
The IgAV and IgAVN hotspots clusters appear where genetic and environmental factors overlap substantially. It can be speculated whether genetic or environmental factors are dominant in the example of IgAV and IgAVN clustering in Croatia [bib_ref] Geospatial clustering of childhood IgA vasculitis and IgA vasculitis-associated nephritis, Sapina [/bib_ref] [bib_ref] IgA vasculitis or Henoch-Schönlein purpura: genetics and beyond, Jelusic [/bib_ref]. Since IgAVN showed linear clustering in the eastern part of Croatia, which follows the course of the Drava and Danube rivers, in the vicinity of areas of Balkan endemic nephropathy, known to occur primarily due to environmental factors (aristolochic acid) [bib_ref] Genes and environment in chronic kidney disease hotspots, Friedman [/bib_ref] , the question arises as to whether the same is true for IgAVN. If this proves correct, it would be contrary to geospatial distribution of IgA nephropathy that is predominantly associated with genetic factors-different variants of innate immunity genes as well as of genes important for defense against parasitic infections [bib_ref] Geographic differences in genetic susceptibility to IgA nephropathy: GWAS replication study and..., Kiryluk [/bib_ref]. That would be another important difference between IgAVN and IgA nephropathy, for which it is still debated whether they are different diseases or just two variants of the same disease [bib_ref] What is the difference between IgA nephropathy and Henoch-Schönlein purpura nephritis?, Davin [/bib_ref].
## New insights in the pathogenesis of igav
The complexity of the etiopathogenesis of IgAV is reflected in the interaction of genetic and environmental factors, with special emphasis on infections [bib_ref] Advances in our understanding of the patogenesis of Henoch-Schönlein purpura and the..., Park [/bib_ref] [bib_ref] Is there a crossroad between infections, genetics, and Henoch-Schönlein purpura?, Rigante [/bib_ref]. The genetic background is indisputable; this is supported by the fact that the incidence and geospatial distribution of IgAV and IgAVN differ around the world and between the different ethnicities [bib_ref] EULAR/PRINTO/PRES criteria for Henoch-Schonlein purpura, childhood polyarteritis nodosa, childhood Wegener granulomatosis and..., Ozen [/bib_ref] , that the incidence of IgAV sometimes has a tendency for familial aggregation [bib_ref] Aberrant glycosylation of IgA1 is inherited in both pediatric IgA nephropathy and..., Kiryluk [/bib_ref] [bib_ref] Sibling cases of Henoch-Schönlein purpura in two families and review of literature, Zhang [/bib_ref] [bib_ref] Familial cases of Henoch-Schönlein purpura in Taiwanese Aborigines, Chen [/bib_ref] , and, finally, that genome-wide association studies (GWAS) point to the significance of common gene variants in the pathogenesis of this disorder [bib_ref] HLA-DQ and HLA-DRB1 alleles associated with Henoch-Schönlein purpura nephritis in Finnish pediatric..., Koskela [/bib_ref].
The results of the GWAS to date classify IgAV as a prototype of a disease related to human leucocyte antigen (HLA) class II loci. A first GWAS pointed to the significance of the polymorphisms in the HLA-DQA1 and DQB1 intergenic zone and at the HLA-DRB1 * 11 and DRB1 * 13 loci (30), while a more recent one showed that haplotype DQA1 * 01:01/DQB1 * 05:01/DRB1 * 01:01 was associated with susceptibility to IgAV but not with other autoimmune diseases [bib_ref] HLA-DQ and HLA-DRB1 alleles associated with Henoch-Schönlein purpura nephritis in Finnish pediatric..., Koskela [/bib_ref]. GWAS of IgAV have not detected potential susceptibility loci to IgAV outside HLA class II genes, but since there are no large GWAS for pediatric IgAV and IgAVN, and existing studies are underpowered to detect smaller allelic effects outside of the HLA region, it is possible that variants in various non-HLA genes associated with immune and inflammatory response escaped statistical detection and these are variants that previous studies have shown to be implicated in the etiopathogenesis of IgAV [bib_ref] Genetics of immunoglobulin-A vasculitis (Henoch-Schönlein purpura): an updated review, López-Mejías [/bib_ref]. The most important non-HLA genes linked with IgAV susceptibility include cytokines genes (ILRN * 2, IL18, and TGFB1), chemokines genes (MCP1), adhesion molecules genes (SELP), renin-angiotensin system (RAS) genes (Agt, ACE), and others (C1GALT1, NOS2A, eNOS, PON1, and MEFV) [bib_ref] Genetics of immunoglobulin-A vasculitis (Henoch-Schönlein purpura): an updated review, López-Mejías [/bib_ref]. For example, genetic variants located at the interleukin (IL) 18 locus could be associated with a higher risk of developing IgAVN [bib_ref] Interleukin 8 gene polymorphism is associated with increased risk of nephritis in..., Amoli [/bib_ref] , while carriage of interleukin-1 receptor antagonist polymorphism 2 (ILRN * 2) may be related to severe renal involvement and renal sequelae in patients with IgAV [bib_ref] Interleukin 1 receptor antagonist gene polymorphism is associated with severe renal involvement..., Amoli [/bib_ref]. Even though the results of these studies deliver more insight into the various molecular pathways, they are not sufficiently large, and the potential association with IgAV is not as strong as it is the case with HLA genes.
These genetic variants may be responsible for regional and ethnic differences observed in IgAV. Thus, for example, the pathogenic variants in the Mediterranean fever (MEFV) gene might predispose to IgAV in patients with familial Mediterranean fever (FMF) and may be associated with different clinical presentation of IgAV in countries where FMF is common [bib_ref] Vasculitis in familial mediterranean fever, Ozdogan [/bib_ref] [bib_ref] MEFV gene mutations in Henoch-Schonlein purpura, Altug [/bib_ref]. In these patients, there might be an increased risk of gastrointestinal complications and IgA deposits on biopsy specimens might be less frequently found, so the diagnosis of IgAV should not be excluded based on absent IgA deposits on skin biopsy in case of clinical presentation suggestive of IgAV.
It was shown that epigenetic changes, that regulate gene activity and expression, are involved in pathogenesis of IgAV [bib_ref] Aberrant histone modifications in peripheral blood mononuclear cells from patients with Henoch-Schönlein..., Luo [/bib_ref]. Luo et al. demonstrated increased global histone H3 acetylation and H3K4 methylation levels in the peripheral blood mononuclear cells of patients with IgAVN compared to healthy controls and patients with IgAV without renal involvement. Furthermore, the authors showed positive correlation of H3 acetylation and H3K4 methylation levels with disease severity. They hypothesized that abnormal levels of histone modifying enzymes can lead to changes in chromatin structure, resulting in increased gene transcription, such as the IL-4 promoter. Another important finding in this study was that in patients with IgAV, the balance between type 1 helper cells (Th1) and type 2 helper cells (Th2) cytokines is disturbed by increasing the level of Th2 specific cytokines (IL-4, IL-6, and IL-13) and decreasing the level of Th1 specific cytokines (IL-2 and IFN-γ) [bib_ref] Aberrant histone modifications in peripheral blood mononuclear cells from patients with Henoch-Schönlein..., Luo [/bib_ref].
Recently, a gene which encodes a histone demethylase involved in the epigenetic control of gene expression (KDM4C) has been implicated in the genetic predisposition of IgAV, highlighting the relevance of the epigenetic mechanisms in the development of this disease [bib_ref] Cross-phenotype analysis of Immunochip data identifies KDM4C as a relevant locus for..., Ortiz-Fernández [/bib_ref].
Multi-hit pathogenesis models for IgAV and IgAVN have been described lately [bib_ref] New insights in the pathogenesis of immunoglobulin A vasculitis (Henoch-Schönlein purpura), Heineke [/bib_ref] [bib_ref] IgA vasculitis with nephritis: update of pathogenesis with clinical implications, Hastings [/bib_ref] [fig_ref] FIGURE 2 |: Two multi-hit pathogenesis models for IgAV and IgAVN [/fig_ref]. It seems that the aberrantly glycosylated IgA1 plays a key role in the pathogenesis of IgAV, especially in patients who develop IgAVN. IgA 1 from most patients with IgAV lack galactose residues [galactose deficient IgA 1 (Gd-IgA 1 )] [bib_ref] Pathogenesis of IgA vasculitis: an Up-To-Date review, Song [/bib_ref]. It is hypothesized that in the Golgi apparatus of IgA 1 -producing immune cells aberrant glycosylation occurs due to decreased galactosyltransferase activity and that genetic predisposition and/or mucosal infection and concomitant IL-6 production cause aberrant glycosylation by altering the glycosylation machinery [bib_ref] Cytokines alter IgA1 O-glycosylation by dysregulating C1GalT1 and ST6GalNAc-II enzymes, Suzuki [/bib_ref]. Two potential genetic loci have been identified in a GWAS in adult patients with IgA nephropathy and increased serum levels of Gd-IgA 1 [bib_ref] GWAS for serum galactose-deficient IgA1 implicates critical genes of the O-glycosylation pathway, Kiryluk [/bib_ref]. These loci, C1GALT1 and C1GALT1C1, are inherited in an autosomal dominant manner and may be responsible for aberrant glycosylation in IgAV, and not only in patients with IgA nephropathy. IgA and IgG antibodies may recognize Gd-IgA 1 as an autoantigen, which leads to the formation of polymeric immune complexes (Gd-IgA 1 -IgA, Gd-IgA 1 -IgG, and Gd-IgA 1 -sCD89, where sCD89 is soluble IgA Fc alpha receptor). It is possible that these circulating immune complexes accumulate in the blood, resulting in their deposition on the endothelium of small blood vessels in the skin, gut, and kidneys. It was shown that serum levels of Gd-IgA 1 are higher in IgAVN patients compared to IgAV patients without nephritis [bib_ref] Biomarkers of IgA vasculitis nephritis in children, Pillebout [/bib_ref]. Some of these Gd-IgA 1 -IgG complexes may deposit in the kidneys, resulting in mesangial cell activation, release of inflammatory mediators, and glomerular injury in patients who develop IgAVN [bib_ref] IgA vasculitis with nephritis: update of pathogenesis with clinical implications, Hastings [/bib_ref].
However, some authors proposed a second multi-hit hypothesis to explain the systemic symptoms of IgAV and IgAVN [bib_ref] New insights in the pathogenesis of immunoglobulin A vasculitis (Henoch-Schönlein purpura), Heineke [/bib_ref]. This model is based on assumption that infection with microorganisms that have similar antigenic structures as components of human vessel walls could lead to the production of cross-reactive anti-endothelial cell antibodies (IgA 1 -AECA) under specific genetic influences. These antibodies may further induce the production of interleukin-8, which is a potent chemoattractant for neutrophils. After activation, neutrophils may cause damage of vascular endothelial cells.
In the context of the new insights, and related to the coronavirus disease 2019 (COVID-19) pandemic, it should be noted that there are several case reports describing patients with IgAV following COVID-19, and some of them were children [bib_ref] A child with COVID-19 and immunoglobulin A vasculitis, Hoskins [/bib_ref] [bib_ref] A child with Henoch-Schonlein purpura secondary to a COVID-19 infection, Alghoozi [/bib_ref] [bib_ref] Purpurona: a novel report of COVID-19-related Henoch-Schonlein purpura in a child, Jacobi [/bib_ref]. It is not yet known how the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus is involved in the pathogenesis of IgAV, whether it is a classical infectious trigger such as the previously described bacteria and viruses or because of its ability to elicit a cytokine response it may be directly involved in the pathogenesis of the disease.
None of the proposed models of IgAV pathogenesis described can explain the fact that in <10% of IgAV patients Gd-IgA 1 in serum or in biopsy specimens is not elevated, nor can they explain the observation that the disease occurs only in some people with IgA 1 glycosylation defect, while in others with elevated Gd-IgA 1 the disease does not occur [bib_ref] Henoch-Schönlein purpura nephritis, Pohl [/bib_ref] [bib_ref] Henoch-Schönlein purpura nephritis in children: incidence, pathogenesis management, Chen [/bib_ref].
## New insights in the clinical presentations of igav
The most common and characteristic signs of the disease are skin manifestations in the form of non-thrombocytopenic purpura or petechiae with lower limb predominance. They are present in all patients and are mandatory classification criterion according to the European League Against Rheumatism (EULAR), Pediatric Rheumatology International Trials Organization (PRINTO) and Pediatric Rheumatology European Society (PRES) classification criteria endorsed in Ankara in 2008 [bib_ref] EULAR/PRINTO/PRES criteria for Henoch-Schonlein purpura, childhood polyarteritis nodosa, childhood Wegener granulomatosis and..., Ozen [/bib_ref]. A possible diagnostic algorithm of the disease is shown in [fig_ref] FIGURE 3 |: Diagnostic algorithm of IgAV [/fig_ref].
Nonetheless, atypical distributions are also possible, affecting the head and neck area, involving the upper extremities more than lower extremities, sparing of the lower extremities, or with diffusely distributed lesions. Furthermore, hemorrhagic bullae, ulcerations and necrotic lesions can be seen in the most severe cases. Recently, several papers have been published on severe skin changes in patients with IgAV, questioning whether these changes are associated with more severe disease, persistent sequelae, and discussing how to treat patients with such manifestations [bib_ref] Blistering eruptions in childhood Henoch-Schönlein syndrome: systematic review of the literature, Ramelli [/bib_ref] [bib_ref] Persistence and severity of cutaneous manifestations in IgA vasculitis is associated with..., Sestan [/bib_ref]. The results of these studies are not uniform, and according to some authors the presence of ulcerations and necroses, persistent purpura (≥1 month) and older age were significant predictors of IgAVN. Also, with increasing severity and duration of cutaneous manifestations in IgAV the risk of developing IgAVN may increase, with a greater likelihood to need aggressive treatment [bib_ref] Henoch-Schönlein purpura nephritis in children: incidence, pathogenesis management, Chen [/bib_ref]. However, others did not notice such associations [bib_ref] Blistering eruptions in childhood Henoch-Schönlein syndrome: systematic review of the literature, Ramelli [/bib_ref]. All the studies conducted so far have in common that they included a very small number of IgAV patients with the most severe cutaneous manifestations, so it is necessary to wait for further results to reach conclusions that will include a larger number of patients.
The second most common feature is represented by musculoskeletal manifestations, in that up to 70-90% of patients with IgAV will have arthralgia or arthritis. According to a Korean nationwide population-based study, younger children are more at risk of developing arthritis, and children younger than 7 years of age had frequent joint symptoms [bib_ref] Ten-year nationwide population-based survey on the characteristics of children with Henoch-Schönlein purpura..., Shim [/bib_ref]. In children with IgAV arthritis is non-deforming and heals without chronic damage within a few weeks. It is interesting to note that in some studies it was observed that joint involvement and subcutaneous edema in the extremities were less frequent in patients with severe gastrointestinal involvement, thus arthralgia could be a negative predictive factor for severe gastrointestinal involvement in patients with IgAV [bib_ref] The clinical spectrum of Henoch-Schönlein purpura in children: a single-center study, Karadagşg [/bib_ref].
More than 50% of children with IgAV may develop gastrointestinal manifestations, and in about 10-20% of patients with gastrointestinal involvement serious complications such as intussusception, bowel perforation, and massive bleeding may occur [bib_ref] Henoch-Schönlein purpura in children, Trnka [/bib_ref]. There are conflicting results from literature regarding the possible association of gastrointestinal symptoms in IgAV and IgAVN. A meta-analysis from 2016 showed that gastrointestinal symptoms were strongly related to renal involvement [bib_ref] Risk factors associated with renal involvement in childhood Henoch-Schönlein Purpura: a meta-analysis, Chan [/bib_ref] , and several other studies have shown similar results [bib_ref] Risk factors of renal involvement significant proteinuria in Henoch-Schönlein purpura, Sano [/bib_ref] [bib_ref] Renal involvement in Henoch-Schönlein purpura: a multivariate analysis of initial prognostic factors, De Almeida [/bib_ref] [bib_ref] Gastrointestinal involvement and its association with the risk for nephritis in IgA..., Sestan [/bib_ref]. Patients in whom IgAV has started with gastrointestinal symptoms and older children with severe gastrointestinal symptoms (severe abdominal pain, intussusception, hematochezia, and/or massive gastrointestinal bleeding) may be a high-risk group for developing IgAVN [bib_ref] Gastrointestinal involvement and its association with the risk for nephritis in IgA..., Sestan [/bib_ref]. However, other studies have not confirmed this association [bib_ref] Obesity increases the risk of renal involvement in children with Henoch-Schönlein purpura, Zhao [/bib_ref] [bib_ref] Risk factors for renal involvement and severe kidney disease in 2731 Chinese..., Wang [/bib_ref] [bib_ref] Predictive factors of renal involvement or relapsing disease in children with Henoch-Schönlein..., Rigante [/bib_ref].
Finding associations of individual clinical elements with the prognosis of IgAV and renal involvement would be very important since it would help to identify high risk group of patients that should be follow-up closely and at shorter intervals to detect renal involvement.
## New insights in the diagnostics of igav, with special emphasis on the role of biomarkers and kidney biopsy
Renal involvement in IgAV ranges from urinary abnormalities (including hematuria or/and proteinuria) through nephritic and nephrotic syndrome to chronic renal failure. IgAVN is typically mild and most commonly manifest only with pathological urine findings. Chronic renal failure has been reported in 1-15% of children with IgAVN, and in the vast majority of cases it is diagnosed within 6 months of disease onset [bib_ref] Risk of long term renal impairment and duration of follow up recommended..., Narchi [/bib_ref] [bib_ref] Long-term followup of childhood Henoch-Schönlein nephritis, Goldstein [/bib_ref] [bib_ref] Long-term prognosis of Henoch-Schönlein nephritis in adults and children. Italian Group of..., Coppo [/bib_ref] [bib_ref] Clinical outcome of Schönlein-Henoch purpura nephritis in children, Schärer [/bib_ref].
Current problems related to the diagnosis of renal disease in IgAV can be divided into two groups. On one hand, kidney biopsy is still the gold standard for the diagnosis of IgAVN, but the big problem is the uneven interpretation of the results since several histological classifications are used in the analysis of renal biopsy findings in IgAVN, but it remains unknown which one has the strongest association with the severity and outcome [bib_ref] Different histological classifications for Henoch-Schönlein purpura nephritis: which one should be used?, Jelusic [/bib_ref]. Another problem is that there are currently no biomarkers in routine clinical use for IgAV and IgAVN which can stratify patients with respect to the risk of developing kidney disease progression and contribute to the earlier diagnosis of renal involvement [bib_ref] IgA vasculitis with nephritis: update of pathogenesis with clinical implications, Hastings [/bib_ref].
The most commonly used histological classification is that of the International Study of Kidney Disease in Children (ISKDC) [bib_ref] Prognosis of Henoch-Schönlein nephritis in children, Counahan [/bib_ref]. The advantage of this classification is that it is relatively simple and widespread in use, so it is known around the world (6). The most important limitation is that it is grounded mostly on the state of glomeruli, therefore only reflecting active inflammation and neglecting vascular and tubulointerstitial changes. The Oxford classification is increasingly used, and was revised in 2017 [bib_ref] Oxford classification of IgA nephropathy 2016: an update from the IgA Nephropathy..., Trimarchi [/bib_ref]. Crescents, the lesions on renal biopsy that have long been considered the most important outcome indicators of IgAVN, were not included in the first version of the classification. Although this histological classification is more complex and some studies have shown its potential for use in IgAVN, caution is needed. Indeed, the working group of Oxford classification does not recommend its use in IgAVN since cases of patients with this condition were not included in the validation cohort, and recent study suggests that the Oxford classification could not be fully validated in IgAVN [bib_ref] Oxford classification of IgA nephropathy 2016: an update from the IgA Nephropathy..., Trimarchi [/bib_ref] [bib_ref] Evaluation of the Oxford classification in immunoglobulin A vasculitis with nephritis: a..., Yu [/bib_ref]. A finnish group published the modified semi-quantitative classification in 2017 [bib_ref] The ISKDC classification and a new semiquantitative classification for predicting outcomes of..., Koskela [/bib_ref]. It is the most complicated, but the first to take into account the chronicity components. Although the first results are promising, a study of a longer number of patients is needed to properly validate the classification.
Since renal biopsy is an invasive procedure, and it is still unclear what is the prognostic value of individual histological elements, there is a need for less invasive procedures in diagnostics. Measurement of biomarkers in urine has many advantages: the sample is relatively easy to collect, without using invasive procedures; urine reflects changes in renal parenchyma, unlike blood that is in contact with a number of organs and organ systems. In addition, the number of different core proteins in the urine is lower than in blood [bib_ref] Current status advances in quantitative proteomic mass spectrometry, Wasinger [/bib_ref] [bib_ref] A comprehensive map of the human urinary proteome, Marimuthu [/bib_ref]. Despite the many potential biomarkers that are emerging, the most consistent finding in patients with IgAV remains an increased serum level of Gd-IgA 1 [bib_ref] Vasculitis update: pathogenesis and biomarkers, Brogan [/bib_ref]. Among urinary biomarkers, IgA and IgM performed best in the study conducted by Pillebout et al. [bib_ref] Biomarkers of IgA vasculitis nephritis in children, Pillebout [/bib_ref]. Recent systematic review of urine biomarkers in children with IgAVN showed that the most promising urinary biomarkers in predicting nephritis were kidney injury molecule-1 (KIM-1), monocyte chemotactic protein-1 (MCP-1), N-acetyl-β-glucosaminidase (NAG), and angiotensinogen (AGT) [bib_ref] A systematic review of urine biomarkers in children with IgA vasculitis nephritis, Williams [/bib_ref]. However, none of them proved yet to be an established marker of disease. Further studies are needed to verify whether preclinical markers are superior to the currently usd ones (24-h urinary protein values, urinary protein:creatinine ratio and urinary albumin concentration).
The application of metabolomics has proven to be a promising approach [bib_ref] Predictive biomarkers of IgA vasculitis with nephritis by metabolomic analysis, Demir [/bib_ref]. Metabolome is a product of proteome and is considered to be closer to the phenotype in comparison with genome, and proteome. Studies regarding metabolomics in IgAV are scarse. Demir et al. found that DHAP (18:0), prostaglandin D2/I2, porphobilinogen, 5-methyltetrahydrofolic acid, and N-Acetyl-4-O-acetylneuraminic acid/N-Acetyl-7-Oacetylneuraminic acid may serve as biomarkers for predicting kidney disease but studies with larger number of IgAV patients are necessary for validation of these findings [bib_ref] Predictive biomarkers of IgA vasculitis with nephritis by metabolomic analysis, Demir [/bib_ref].
## New insights in the assessment of disease activity and damage in igav
Data regarding vasculitis activity and damage assessment in children with IgAV are limited. Since these are key components of outcome measures in patients with vasculitis for both clinical trials and for observing individual patient disease course, it is important to validate the available assessment tools adapted for the pediatric population [bib_ref] Disease activity assessment in childhood vasculitis: development and preliminary validation of the..., Dolezalova [/bib_ref]. For determining disease activity and degree of kidney damage in patients with IgAV/IgAVN two clinical questionares can be used: the Pediatric Vasculitis Activity Score (PVAS) and Pediatric Vasculitis Damage Index (PVDI), respectively [bib_ref] Disease activity assessment in childhood vasculitis: development and preliminary validation of the..., Dolezalova [/bib_ref]. PVAS includes 64 clinical features (symptoms or signs of disease) that are divided into nine organ systems; the assessment of new or worsening items in the last 4 weeks, but not for more than 3 months, is recorded. The total number of points represents the activity of the disease and ranges from 0 to 63 points. PVDI includes 72 clinical variables divided into nine organ systems, as well as an "other" section. The duration of symptoms or signs lasting at least 3 months, occurring at any time since the disease onset, is defined as damage.
## New insights in the treatment of igav
During the self-limited nature of the disease, in the vast majority of children with IgAV specific treatment is not required. Regarding patients with severe skin manifestations, due to the lack of studies with large number of participants the optimal way of treatment is not known, although most are treated with systemic glucocorticoids, sometimes in combination with dapsone or azathioprine [bib_ref] Blistering eruptions in childhood Henoch-Schönlein syndrome: systematic review of the literature, Ramelli [/bib_ref]. Musculoskeletal manifestations are usually treated with rest and analgesia, while other treatment options are rarely necessary. A different situation occurs in children with severe gastrointestinal manifestations, renal involvement or those with other complications such as neurological, lung or multiple organ involvement. Recently, the European initiative SHARE (Single Hub and Access point for pediatric Rheumatology in Europe) developed consensus-based recommendations for diagnosis and treatment of IgAV [bib_ref] European consensus-based recommendations for diagnosis and treatment of immunoglobulin A vasculitis-the SHARE..., Ozen [/bib_ref]. It is important to emphasize that these recommendations are not intended for pediatric rheumatologists and nephrologists, but for general pediatricians and physicians who have little or no experience with severe IgAV and IgAVN patients.
In children with severe abdominal pain or gastrointestinal hemorrhage, glucocorticoids should be considered: oral or pulsed glucocorticoids if oral route is not tolerated or they have failed to respond. Mycophenolate mofetil, cyclophosphamide, intravenous immunoglobulin, rituximab, methotrexate, colchicine, and hydroxychloroquine may be considered as second-line treatments. Other supportive treatment, such as, nasogastric decompression, parenteral nutrition, and antibiotics may be required.
According to the SHARE management algorithm, IgAVN is divided into three categories, taking into account proteinuria, estimated glomerular filtration rate and percentage of crescents on renal biopsy [bib_ref] European consensus-based recommendations for diagnosis and treatment of immunoglobulin A vasculitis-the SHARE..., Ozen [/bib_ref]. Children without proteinuria or renal dysfunction usually do not need any specific therapeutic intervention. In patients with mild forms of IgAVN, defined as ≤2.5 g/day of proteinuria in 24 h urine collection with normal estimated glomerular filtration rate, first-line therapy consists of oral glucocorticoids, and in the case of persistence of proteinuria, second-line drugs may be used (e.g., azathioprine, mycophenolate mofetil, or glucocorticoid pulses). Glucocorticoids, usually parenterally and in pulsed doses, are the first choice in the treatment of moderate IgAVN, defined as <50% crescents on renal biopsy and impaired estimated glomerular filtration rate (<80 ml/min/1.73 m 2 ) or severe persistent proteinuria (>2.5 g/day of proteinuria in 24 h urine collection for more than 4 weeks). In the absence of response, second-line drugs are added: azathioprine, mycophenolate mofetil, or cyclophosphamide parenterally. Treatment of the most severe forms of IgAVN consists of two phases: the first one is induction using pulsed doses of glucocorticoids in combination with intravenous cyclophosphamide pulses, and the second phase is maintenance therapy with lower doses of glucocorticoids in combination with immunomodulators: azathioprine or mycophenolate mofetil. To prevent or limit secondary glomerular damage in patients with IgAVN who have persistent proteinuria (lasting more than 3 months), angiotensin converting enzyme inhibitors or angiotensin receptor blockers are recommended. The SHARE recommendations for IgAVN treatment are summarized in [fig_ref] FIGURE 4 |: The SHARE recommendations for IgAVN treatment [/fig_ref].
For unresponsive cases there is an option of plasma exchange that showed efficacy in one study while there is not enough evidence regarding the use of rituximab (although it has been described in case reports and case series) or intravenous immunoglobulins [bib_ref] Rituximab treatment for chronic steroid-dependent Henoch-Schonlein purpura: 8 cases and a review..., Crayne [/bib_ref].
In addition to the SHARE recommendations, there is also the Kidney Disease: Improving Global Outcomes (KDIGO) practice guideline on glomerulonephritis, which in one chapter provide recommendations for the treatment of IgAVN in children and adults [bib_ref] Targeted-release budesonide versus placebo in patients with IgA nephropathy (NEFIGAN): a double-blind,..., Radhakrishnan [/bib_ref]. Angiotensin converting enzyme inhibitors or angiotensin receptor blockers are suggested for children with persistent proteinuria 0.5-1 g/day per 1.73 m 2 , while for those with proteinuria >1 g/day per 1.73 m 2 , after a trial of angiotensin converting enzyme inhibitors or angiotensin receptor blockers, a 6-month course of glucocorticoid therapy should be considered. Children with crescentic IgAVN with nephrotic syndrome and/or deteriorating kidney function should be treated with glucocorticoids and cyclophosphamide according to KDIGO practice guideline.
The biggest problem in treating children with IgAV is the lack of high-level evidence and randomized controlled trials [bib_ref] European consensus-based recommendations for diagnosis and treatment of immunoglobulin A vasculitis-the SHARE..., Ozen [/bib_ref].
When it comes to new therapeutic options, there is room for improvement. Since the most important chronic complication of the disease is IgAVN, and according to some hypotheses the pathogenesis of IgAVN could be closely related to the pathogenesis of IgA nephropathy [bib_ref] New insights in the pathogenesis of immunoglobulin A vasculitis (Henoch-Schönlein purpura), Heineke [/bib_ref] [bib_ref] IgA vasculitis with nephritis: update of pathogenesis with clinical implications, Hastings [/bib_ref] [bib_ref] Cytokines alter IgA1 O-glycosylation by dysregulating C1GalT1 and ST6GalNAc-II enzymes, Suzuki [/bib_ref] , new targeted therapies being investigated in IgA nephropathy could soon be extended to IgAVN. Examples of such therapeutic options include budesonide, which can target Peyer's patches in the ileum where the production of Gd-IgA1 is thought to originate (77); bortezomib, a proteasome inhibitor which is a plasma cell depleting agent (it affects production of IgG autoantibodies) [bib_ref] Successful outcome using bortezomib in adult refractory IgA vasculitis: a case report, Van De Perre [/bib_ref] ; complement inhibitors such as APL-2, CCX168, LNP023, and OMS721 (79); or the spleen tyrosine kinase inhibitors [bib_ref] Spleen tyrosine kinase inhibition is an effective treatment for established vasculitis in..., Mcadoo [/bib_ref]. Currently, precision medicine has not found its place in the treatment of vasculitis [bib_ref] Vasculitis pathogenesis: can we talk about precision medicine?, Ozen [/bib_ref].
## New insights in the follow-up of patients with igav
It is proposed to follow-up patients with IgAV for at least 6-12 months even if the initial blood pressure measurements and urinalysis are normal (74), measuring regularly blood pressure and performing urinalyses to detect presence of haematuria, and quantification of albuminuria and/or proteinuria. Bearing in mind that recently published research has indicated that a certain group of patients (such as older children with the onset of gastrointestinal symptoms before other IgAV symptoms and severe GI form of IgAV, as well as those who develop ulcerations and necroses and persistent purpura) may be at higher risk for the later development of nephritis, the question arises whether some children should be monitored longer than recommended [bib_ref] Persistence and severity of cutaneous manifestations in IgA vasculitis is associated with..., Sestan [/bib_ref] [bib_ref] Gastrointestinal involvement and its association with the risk for nephritis in IgA..., Sestan [/bib_ref].
The question of how long to follow-up children who have developed IgAVN and entered disease remission is still unanswered. During 23 years of follow-up, it was shown that up to 44% of patients with severe IgAVN at onset and up to 13% with mild IgAVN at onset developed reduced renal function and/or hypertension [bib_ref] Long-term followup of childhood Henoch-Schönlein nephritis, Goldstein [/bib_ref]. Another study indicated that 70% of pregnancies in women with IgAVN with onset in childhood were complicated by hypertension and/or proteinuria [bib_ref] The adult kidney 24 years after childhood Henoch-Schonlein purpura: a retrospective cohort..., Ronkainen [/bib_ref]. These data call for caution and emphasize the need for long-term observation even in patients who went into remission.
# Conclusion
In this review, we describe how new insights have been gained regarding this disease, which will necessarily require a paradigm shift if we want to make further progress in terms of developing a non-invasive diagnosis of nephritis (which is the most common chronic complication of the disease), as well as its treatment (the main problem being the lack of high-level evidence based on randomized controlled trials). The incidence of IgAV and IgAVN varies worldwide and may be a consequence of overlapping genetic and environmental factors. Besides HLA class II genes, various non-HLA genes may also have significance in its etiopathogenesis. Factors other than Gd-IgA 1 -containing immune complexes may also be important in a multi-hit pathogenesis of this disease. Renal biopsy is still the gold standard for the diagnosis of IgAVN, but in interpretating the histologic findings it should be taken into account that tubulointerstitial changes could be very important as predictors of poor outcome, so other histologic classifications than ISKDC, such as the revised Oxford classification (MEST-C score) may be considered. The most consistent biomarker in patients with IgAV is represented by increased serum levels of Gd-IgA1, while non-invasive confirmation of nephritis is still pending. In the absence of high-level evidence concerning treatment based on randomized controlled trials, SHARE recommendations have been developed. Patients with IgAV, and especially with IgAVN, should be followed-up for long-term, even when the remission of the disease is established, because of possible complications.
# Author contributions
MJ and MS reviewed the literature and wrote much of the manuscript. TG and RC reviewed the literature and wrote parts of the manuscript, planned and oversaw the entire review, and contributed to all aspects of the manuscript. All authors contributed to the article, approved the submitted version, and agree to be accountable for the content of the work.
# Funding
This work has been supported in part by Croatian Science Foundation under the project IP-2019-04-8822.
[fig] FIGURE 1 |: Distribution of incidence of IgAV around the world. [/fig]
[fig] FIGURE 2 |: Two multi-hit pathogenesis models for IgAV and IgAVN. Modified according to Heineke et al. (40) and Hastings et al. (41). [/fig]
[fig] FIGURE 3 |: Diagnostic algorithm of IgAV. Modified according to Ozen et al. (2). [/fig]
[fig] FIGURE 4 |: The SHARE recommendations for IgAVN treatment. Modified according toOzen et al. (74). [/fig]
|
Human keratinocytes are a source for tumor necrosis factor alpha: evidence for synthesis and release upon stimulation with endotoxin or ultraviolet light
Tumor necrosis factor a (TNF-a), in addition to being cytotoxic for certain tumor cells, has turned out as a multifunctional cytokine that is involved in the regulation of immunity and inflammation. Since human keratinocytes have been demonstrated to be a potent source of various cytokines, it was investigated whether epidermal cells synthesize and release TNF-m Supernatants derived from normal human keratinocytes (HNK) and human epidermoid carcinoma cell lines (KB, A431) were tested both in a TNF-a-specific ELISA and a bioassay. In supernatants of untreated epidermal cells, no or minimal TNF-a activity was found, while after stimulation with lipopolysaccharide (LPS) or ultraviolet (UV) light, significant amounts were detected. Western blot analysis using an antibody directed against human TNF-a revealed a molecular mass of 17 kD for keratinocyte-derived TNF-a. These biological and biochemical data were also confirmed by Northern blot analysis revealing mRNA specific for TNF-a in LPS-or ultraviolet B (UVB)treated HNK and KB cells. In addition, increased TNF-a levels were detected in the serum obtained from human volunteers 12 and 24 h after a single total body UVB exposure, which caused a severe sunburn reaction . These findings indicate that keratinocytes upon stimulation are able to synthesize and release TNF-a, which may gain access to the circulation. Thus, TNF-a in concert with other epidermal cell-derived cytokines may mediate local and systemic inflammatory reactions during host defense against injurious events caused by microbial agents or UV irradiation.
T NF-a/cachectin has been originally described as a macrophage-derived factor that induces cachexia and hemorrhagic necrosis of tumors in animals, and ports cytolytic effects on several tumor cell lines [bib_ref] Human tumor necrosis factor : precursor structure, expression and homology to lymphotoxin, Pennica [/bib_ref] [bib_ref] Macrophages as a source of tumoricidal activity (tumor-necrotizing factor), Mannel [/bib_ref]. In the meantime, it became evident that TNF-a exhibits a variety ofactivities, including release of prostaglandins and collagenase [bib_ref] Cachectin/tumor necrosis factor stimulates collagenase and prostaglandin E2 production by human synovial..., Dayer [/bib_ref] , and activation ofneutrophils [bib_ref] Activation of human polymorphonuclear neutrophil functions by interferon-gamma and tumor necrosis factors, Shalaby [/bib_ref] , eosinophils, and macrophages [bib_ref] Potentiation oflymphokine-induced macrophage activation by tumor necrosis factor-a, Heidenreich [/bib_ref]. Moreover, TNF-a increases MHC class I antigen [bib_ref] Recombinant human tumor necrosis factor increases mRNA levels and surface expression of..., Collins [/bib_ref] and intercellular adhesion molecule (ICAM)r expression on many cells (8), and induces the release of other cytokines. Through these multiple effects, TNF-a appears to be an essential mediator in inflammatory and immunologic reactions during host defense.
In addition to their barrier function, epidermal cells and, in particular, keratinocytes have been recently recognized to 1 Abbreviations used in this paper. GM-CSF, granulocyte/macrophage CSF; HNK, normal human keratmocytes ; ICAM, intercellular adhesion molecule, LC, Langerhans cell; UVB, ultraviolet B. exhibit the capacity to release a variety of soluble mediators, including 111, Hr6, IIr8, CSF, transforming growth factors, and platelet-derived growth factor, that influence the growth, differentiation, and function of epidermal, dermal, and immunocompetent cells (9-11) . The constitutive production of thesefactors usually is very low but can be induced significantly by various injurious agents, such as endotoxin, viral particles, tumor promotors, or UV light (9, 11). Consequently, these cytokines are not only involved in the mediation oflocal inflammatory reactions within the epidermis, but may also enter the circulation and thus cause systemic effects. Therefore, the present study was performed to investigate whether human keratinocytes in addition to these mediators can also release the multifunctional cytokine TNF-a/cachectin.
# Materials and methods
Cells and Cell Lines. Long-term cultures of normal human keratinocytes (HNK) were obtained from normal human foreskin as The journal of Experimental Medicine " Volume 172 December 1990 1609-1614 TNF alpha (po/nl) . Human epidermoid carcinoma cell fine cells (KB) releaseTNF-a upon stimulation with LPS (100 pg/ml) . Supernatants of untreated or LPS-treated KB cells were harvested at different time points after stimulation and tested for TNF-ca activity by the use of a TNF-a ELISA.
described previously and maintained in KGM (Clonetics, San Diego, CA) (12) . Hydrocortisone was omitted from the culture medium 48 h before testing. The human epidermoid carcinoma cell lines KB and A431 (American Type Culture Collection, Rockville, MD) were maintained in monolayer cultures at 37°C in a humidified atmosphere containing 5% C02. To investigate factor production, cells were plated at a density of 5 x 105/ml in 24-well plates (Costar, Cambridge, MA), and supernatants were harvested at different time points (1, 3, 6, 12, 24, or 48 h), centrifuged at 1,000 g, and stored at -20°C. To stimulate cytokine production, cells were In Vivo Exposure. After informed consent was obtained, four human volunteers were exposed to a single total body UVB exposure equivalent to four minimal erythema doses as previously described (14) . As a light source, an unfiltered bank of six blue light lamps (Dr. Hoelmle, Munich, FRG) was used, emitting a linear spectrum between 250 and 600 nm with emission peaks at 300 and 360 nm, respectively. Blood was drawn from the antecubital vein into sterile heparinized tubes immediately before UV exposure, as well as 12 and 24 h after UV treatment.
Assayfor TNFa. TNF-a content of epidermal cell culture supernatants was assayed by determining the cytotoxicity of TNF-a against sensitive murine IrM cells (15) . Briefly, IM cells were seeded at a concentration of 4 x 10' cells/well in 100 Al culture medium and incubated for 22-24 h at 37°C in a 5% CO2 atmosphere to establish a dense monolayer. 0.1 ml of culture supernatants to be tested was added in the presence of actinomycin D (2 pg/ml) for 18 h. Viability of cells was measured by staining with 0.5% crystal violet . After removal of excess dye and drying, absorbance was determined at 540 nm using a microplate reader (3550; Bio-Rad Laboratories, Richmond, CA).
In addition, a commercially available ELISA was used for measuring human TNF-a (Endogen, Boston, MA). This kit is based on the use of a mouse anti-TNF-a mAb and a rabbit polyvalent antiTNF-a antibody. It is sensitive to 10 pg/ml TNF-a and does not crossreact with human lymphotoxin (TNF-0), ID1, or IL6.
Western Blot Analysis. Supernatants to be tested were concentrated 100 times by using Amicon ultrafiltration and were subjected 9o TNF alpha activity (U/ml)
# Results
## 1610
Tumor Necrosis Factor a Production by Human Keratinocytes to SDS-PAGE (15% total acrylamide concentration) under reducing conditions [bib_ref] Binding of rI1rI0 to the third complement component and alpha2-macroglobulin upon activation..., Borth [/bib_ref]. After SDS-PAGE, proteins were electrotransferred to nitrocellulose and stained with a mouse mAb directed against recombinant human TNF-a (kindly provided by G. Adolf, Ernst-Boehringer Institute, Vienna, Austria) . Biotinylated molecular weight markers were obtained from Sigma Chemical Co. (St. Louis, MO). Northern Blot Analysis . Cultured cells were detached by treatment with ice-cold PBS and lysed in 4 M guanidine isothiocyanate . RNA was isolated by ultracentrifugation through cesium chloride, and subjected to electrophoresis in 1.0% agarose gel containing formaldehyde (2 .2 M) . After capillary transfer of RNA to nitrocellulose, membranes were prehybridized for at least 4 h with use of hybridizing solution containing 10% dextrane sulfate, 40% formamide, 20% 20x SSC, sonicated salmon sperm DNA (50 pg/ml), and yeast transfer RNA (50 pg/ml) . TNF-ac mRNA was detected by hybridization with a 0.8-kb EcoRl cDNA fragment (kindly provided by D. Pennica, Genentech, San Francisco, CA). After stringent washing conditions, blots were exposed to x-ray film at -70°C for 24 h (17) .
When epidermal cell-derived supernatants were evaluated for TNF-a production using a TNF-a-specific ELISA, no or only minimal amounts were found in conditioned medium obtained from unstimulated KB cells. However, when KB cells were treated with LPS (100 pg/ml), significant amounts of TNFLoi were detected. TNF-a release started 1 h after stimulation and peaked at x+12 h, retuming to base line after 48 h . Similar results were also obtained with HNK (data not shown) . To test if the protein released by epidermal cells is biologically active, supernatants from FINK, A431, and KB cells were evaluated in the 1rM bioassay. These supernatants again demonstrated only minimal amounts of active TNF-a until cells were treated with LPS [fig_ref] Figure 2: Epidermal cells release biologically active TNF-a [/fig_ref].
To confirm that the cytotoxic activity against IrM cells observed in HNK supernatants is due to TNF-a, Western blot analysis using a mAb directed against human TNF-ot was performed. A weak, but clearly detectable band with a molecular mass of 17 kD was observed in supernatants of LPS-stimulated HNK [fig_ref] Figure 3: Western blot analysis of keratinocyte-derived TNF-a [/fig_ref].
To test whether the induction of TNF-ci by epidermal cells is regulated at the transcriptional level, the expression of TNF-ci mRNA in HNK and KB cells was examined . Unstimulated KB cells expressed little TNF-ci mRNA [fig_ref] Figure 4: KB cells express mRNA encoding for TNFLa [/fig_ref] , lane 2), consistent with our previous biological data . However, treatment with LPS resulted in a marked increase in the level of TNF-ci mRNA [fig_ref] Figure 4: KB cells express mRNA encoding for TNFLa [/fig_ref].
Since UVB light is well known to cause a significant dermal inflammatory response in vivo and is a potent inducer of cytokine production by epidermal cells in vitro as well (18), HNK and KB cells were exposed to UVB irradiation . In com- Kick et al .
7MF alpha (pg/ml) . Stimulation of TNF-a production by UVB irradiation. HNK and KB cells were exposed to UVB light (100 J/m2). After a 24-h incubation, supernatants were tested for TNF-a using a TNF-a-specific BLISA.
Results are expressed as pg/ml t SD of three different experiments.
parison with supernatants of untreated cells, both irradiated HNK and KB cells released significantly increased amounts of TNF-a . Maximum production was observed when cells were exposed to 100 J/m 2, and supernatants were harvested after 24 h. In addition, UVexposed HNK and KB expressed significantly increased TNF-a mRNA as evaluated by Northern blot analysis using a TNF-ot-specific cDNA probe [fig_ref] Figure 3: Western blot analysis of keratinocyte-derived TNF-a [/fig_ref]. Maximum TNF-ot expression was detected when RNA was extracted 12 h after UV treatment . After extensive solar exposure, increased levels of cytokines recently have been detected in the circulation (14, 18) . Thus, serum samples from human healthy volunteers after total body UV irradiation causing severe sunburn reaction were evalu- . Northern blot analysis of UVtreated normal human keratinocytes. HNK cells were left untreated or irradiated with UVB light (100 J/m2), and total RNA was extracted at various time points and hybridized using 32P-labeled TNF-a-specific cDNA . The blot was stripped and re-hybridized with a probe for fl-actin (bottom) confirming that the RNA on the blot is intact, and that approximately equal amounts of RNA were loaded in each lane. Flours post W . Serum TNF-a levels after UV treatment . Human healthy volunteers (n -4) were treated with a single total body UVB exposure equivalent to four minimal erythema doses. Serum samples were obtained before as well as 12 and 24 h after UV irradiation. TNF-a was evaluated using the TNF-a-specific ELISA . Results are expressed as pg/ml mean ± SD.
ated for TNF-a using the TNF-a-specific ELISA. Serum samples obtained before UVB exposure contained minimal levels of TNF-a . However, significant circulating TNF-a was detected 12 and 24 h after UVB treatment .
# Discussion
The present study demonstrates that both HNK and the epidermoid carcinoma cell lines KB and A431 upon stimulation release immunoreactive TNF-a. This cytokine is produced as a biologically inactive prohormone that has to be cleaved at several sites to become the mature biologically active polypeptide [bib_ref] Mechanisms of disease, Beutler [/bib_ref] [bib_ref] Cachectin and tumor necrosis factor as two sides of the same biological..., Beutler [/bib_ref]. Therefore, in addition to the ELISA, a TNF-a bioassay (IrM test) was utilized to prove that keratinocytes release a biologically active TNF-a. Accordingly, human keratinocytes produced low amounts of biologically active TNF-a. Nevertheless, it is hard to compare quantities of TNF-a measured with the ELISA with the quantity estimated in the bioassay, since one has to be aware of the fact that crude supernatants of keratinocytes and epidermoid carcinoma cell lines usually contain several other mediators that may interfere with bioassays (11). The identity of epidermal cell-derived TNF-a protein was confirmed by Western blot analysis using a specific mAb directed against TNF-a, which demonstrated a specific band with a molecular mass of 17 M Keratinocytes constitutively secrete little TNF-a since TNF-a protein was hardly detectable in supernatants of untreated epidermal cells. However, upon stimulation with LPS or UVB, both FINK and epidermal carcinoma cells produced significant amounts of TNF-a . The results were also confirmed by Northern blot analysis detecting TNF-oc mRNA in keratinocytes after stimulation with LPS or UVB. The present data clearly demonstrate that human keratinocytes upon stimulation can function as a source of TNF-a. The quantities released, however, even after optimal induction, are much less than the amounts produced by macrophages.
Human keratinocytes have been demonstrated to release various cytokines, including IIrla andjS (21), Ilr6 [bib_ref] IFN-02, B cell differentiation factor 2, or hybridoma growth factor (IL6) is..., Kimbaue4 [/bib_ref] , Ilr8 (23), CSF, transforming growth factors (11), and plateletderived growth factor (10). Through the capacity to release these mediators, keratinocytes may play an important role in inflammatory reactions of the skin. Similar to the present findings concerning keratinocyte TNF-a release, the constitutive production of these factors both in vitro and in vivo is very low and has tobe induced by various stimuli, including bacterial or viral products, tumor promotors, UV light, or cytokines themselves. TNF-a hasbeen demonstrated to stimulate the release of other mediators, such as I1J1, IIr6, IIr8, granulocyte/macrophage (GM)-CSF, and platelet-derived growth factor in various cell types [bib_ref] Recombinant human TNF induces production of granulocytemonocyte colony-stimulating factor, Munker [/bib_ref] [bib_ref] Tumo r necrosis factor (Cahectin) is an endogenous pyrogen and induces production..., Dinarello [/bib_ref] [bib_ref] Tumor necrosis factor/Cahectin interacts with endothelial cell receptors to induce release of..., Nawroth [/bib_ref] [bib_ref] Interleukin 8 and MCAF: novel inflammatory cytokines inducible by IL 1 and..., Matsushima [/bib_ref] , and thus appears to be a key member of the cytokine cascade, playing an important role in pathologic events accompanying invasion of foreign organisms. Since the epidermis is frequently in contact with environmental microorganisms, it is not surprising that epidermal cells appear to be endowed with the capacity to release TNF-a. TNF-a functions in synergy with other cytokines, exhibiting some antiviral activity [bib_ref] In vitro anti-human immunodeficiency virus activities of tumor necrosis factor-a and interferon-gamma, Wong [/bib_ref] , and thus may be involved in the defense of the epidermis against viral particles . Moveover, TNT-a, like ILl, increases ICAM-1 expression on fibroblasts, endothelial cells, and keratinocytes (8), resulting in an enhanced adhesion oflymphocytes, which may be important in certain lymphocyte-mediated skin diseases. The role of TNF-a as a regulatory cytokine in immunological as well as inflammatory processes is further supported by the finding that TNF-a enhances T and B cell functions, induces production of other cytokines, and modulates the activity of cytotoxic T cells [bib_ref] Mechanisms of disease, Beutler [/bib_ref].
An exaggerated release of TNF-a by the host can be toxic to the host, which is best demonstrated by the crucial role ofTNF-a in the pathogenesis ofendotoxic shock and cachexia in sepsis [bib_ref] The role of Cahectin/TNF in endotoxic shock and cachexia, Cerami [/bib_ref]. TNF-a may also be involved in mediating the skin changes in GVHD. Accordingly, treatment of animals with antiTNF antibodies prevents the development of cutaneous and intestinal lesions during the acute phase ofGVHD [bib_ref] Tumor necrosis factor/Cahectin is an effector of skin and gut lesions of..., Piquiet [/bib_ref]. Recently, TNF-a has been detected in human epidermal cells of skin biopsies obtained from healthy subjects, and staining intensity was found increased after UVB exposure [bib_ref] Immunohistological detection of interleukin 1-like molecules and tumour necrosis factor in human..., Oxholm [/bib_ref]. Due to the use ofonly immunohistochemical techniques, this study did not definitely prove that keratinocytes are the real source of TNF-a, since it cannot be exclu&d that the cytokine is released by inflammatory cells in the dermis and just bound on the surface of keratinocytes . The results of our study, however, indicate that human keratinocytes appear to be the most likely source for TNF-a found within the epidermis.
According to the present data, UV light appears as an inducer ofTNF-a production by epidermal cells in vitro. Since solar exposure can cause a significant inflammatory response in the skin, TNF-a may be involved in the mediation of this local reaction directly or via induction of other cytokines [bib_ref] Recombinant human TNF induces production of granulocytemonocyte colony-stimulating factor, Munker [/bib_ref] [bib_ref] Tumo r necrosis factor (Cahectin) is an endogenous pyrogen and induces production..., Dinarello [/bib_ref] [bib_ref] Tumor necrosis factor/Cahectin interacts with endothelial cell receptors to induce release of..., Nawroth [/bib_ref] [bib_ref] Interleukin 8 and MCAF: novel inflammatory cytokines inducible by IL 1 and..., Matsushima [/bib_ref]. In addition, this study demonstrates that after extensive UV exposure, increased TNF-a levels can be detected in the circulation and thus may be responsible for systemic effects . Although keratinocytes are the primary target ofUVB light and a source for TNF-a, it cannot be determinedwhether TNF-a found in the circulation is keratinocyte derived or not . Accordingly, cells other than keratinocytes (e.g., macrophages) may be considered as the source for enhanced circulating TNF-a, since in the present study, TNF-a serum levels continue to increase out to 24 h, whereas in vitro the message for TNF-a in keratinocytes has begun to decline at that time. UV exposure suppresses the immune response signficantly, e.g., blocks contact and delayed-type hypersensitivity reaction in mice [bib_ref] Immunological unresponsiveness induced by ultraviolet radiation, Kripke [/bib_ref]. Recently, it has been demonstrated also that injection of TNF-a similar to exposure with UVB light inhibits the induction of contact hypersensitivity, suggesting that TNF-a released in enhanced amounts upon UV irradiation may be involved in the suppression of the immune response after UV exposure [bib_ref] Concerning a possible link between the effect of UVB on contact hypersensitivity..., Yoshikawa [/bib_ref].
Currently, another remarkable function ofTNF-a within References the murine epidermis has been recognized. Murine epidermal Langerhans cells (LC) kept in culture mature into potent immunostimulatory dendritic cells [bib_ref] Murine epidermal Langerhans cells mature into potent immunostimulatory dendritic cells in vitro, Schuler [/bib_ref]. This maturation is mediated by two keratinocyte-derived cytokines: GM-CSF, which maintains LC viability and increases LC function; and IIr1, which in combination with GM-CSF enhances LC function [bib_ref] Granulocyte/macrophage colony-stimulating factor and interleukin 1 mediate the maturation of murine epidermal..., Heufler [/bib_ref]. Recent studies indicate that LC culturedjust in the presence of TNF-a had not matured as measured by their T cell stimulatory capacity, but still survived, suggesting that TNF-a serves as a crucial signal in LC viability [bib_ref] Tumor necrosis factor ot maintains the viability of murine epidermal Langerhans cells..., Koch [/bib_ref]. Accordingly, expression of TNF-a message has been detected in murine (BALB/c) epidermal cells [bib_ref] Tumor necrosis factor ot maintains the viability of murine epidermal Langerhans cells..., Koch [/bib_ref].
In summary, the finding that human keratinocytes release TNF-a in addition to other cytokines further supports the crucial role of the epidermis, and in particular of the keratmocyte in the pathogenesis ofboth local and systemic inflammation after UV irradiation, as well as host defense against microbial organisms and tumors .
We thank M. Bednar for the excellent secretarial assistance and R. Frost for the expert technical assistance in preparing the figures .
[fig] Figure 2: Epidermal cells release biologically active TNF-a. Supernatants of HNK, KB, or A431 cells either untreated or stimulated with LPS (100 pg/ml) were tested in the TNF-a-specific IPM bioassay. Culture supernatants were harvested 24 h after stimulation. [/fig]
[fig] Figure 3: Western blot analysis of keratinocyte-derived TNF-a. Supernatants derived from unstimulated (lane 5) or LPS (100 ug/ml)-treated HNK (lane 4) were harvested after 24 h and run on a 12% SDS gel. Staining was performed with a mouse mAb directed against recombinant human TNF-a. As a positive control, recombinant human TNF-a was used at different concentrations (100 ng, lane 1 ; 10 ng, lane 2; 1 ng, lane 3) . [/fig]
[fig] Figure 4: KB cells express mRNA encoding for TNFLa. KB cells were stimulated with LPS (100 Ag/ml, lane 1), left untreated (lane 2), or irradiated with UVB (100 J/m2, lane 3) . After a 4-h incubation, RNA was extracted and hybridized with a cDNA probe encoding for TNF-a. 1611 [/fig]
|
Effectiveness of Ultrasound-Guided Carpal Tunnel Injection Using In-Plane Ulnar Approach
The objective of this study is to evaluate the degree of symptom improvement and the change of electrophysiological and ultrasonographic findings after sonographically guided local steroid injection using an in-plane ulnar approach in carpal tunnel syndrome (CTS).Seventy-five cases of 44 patients diagnosed with CTS were included and evaluated at baseline and at 4 and 12 weeks after injection. All patients received injection with 40 mg of triamcinolone mixed with 1 mL of 1% lidocaine into the carpal tunnel using an in-plane Ultrasound (US)guided ulnar approach, out-plane US-guided approach, and blind injection. For clinical evaluation, we used the Boston Carpal Tunnel Questionnaire (BCTQ) and electrophysiological tests. The ultrasonographic findings were also evaluated with regard to cross-sectional area and the flattening ratio of the median nerve.Subjective symptoms measured by BCTQ and median nerve conduction parameters showed significant improvement at 4 weeks in the in-plane ulnar approach group compared with the out-plane ulnar approach and blind injection. This improvement was still observed at 12 weeks. The flattening ratio and cross-sectional area of the median nerve showed a more significant decrease with the in-plane ulnar approach than with the out-plane ulnar approach and blind injection (P < 0.05).US-guided local steroid injection using an in-plane ulnar approach in the CTS may be more effective than out-plane or blind injection. (Medicine 93(29):e350) Abbreviations: BCTQ = Boston Carpal Tunnel Questionnaire, CMAP = compound motor action potential, CSA = crosssectional area, CTS = carpal tunnel syndrome, FR = flattening ratio, FSS = functional status scale, LD = long diameter, SD = short diameter, SDL = sensory distal latency, SNAP = sensory nerve action potential, SSS = symptom severity scale, US = ultrasound. Editor: Kazuo Hanaoka.Values are mean AE standard deviation. Blind ¼ blind injection group, Ulnar-I ¼ in-plane ulnar approach injection group, Ulnar-O ¼ out-plane injection group. Ã Values are mean AE standard deviation. Blind ¼ blind injection group, CMAP ¼ compound muscle action potential, DML ¼ delayed motor latency, M-U ¼ median-to-ulnar sensory nerve distal latency ratio, SDL2 ¼ sensory distal latency at 2nd finger, SNAP2 ¼ sensory nerve action potential amplitude at 2nd finger, Ulnar-I ¼ in-plane ulnar approach injection group, Ulnar-O ¼ out-plane injection group. P values were calculated by analysis of variance with post hoc Tukey-b test.Ã P < 0.05 versus baseline. y P < 0.05 versus after 4 weeks.
# Introduction
C arpal tunnel syndrome (CTS) is caused by compression of the median nerve at the wrist level and the most frequent entrapment neuropathy in the upper limb. [bib_ref] Clinical practice. Carpal tunnel syndrome, Katz [/bib_ref] It can be treated with surgical or nonsurgical methods. [bib_ref] Local injection versus surgery in carpal tunnel syndrome: neurophysiologic outcomes of a..., Andreu [/bib_ref] When nonsurgical treatment is indicated, local corticosteroid injection into the carpal tunnel can be used to reduce pain and tingling sensation. On the contrary, direct needle injury of the median nerve is frequent and leakage of the corticosteroid injectate from the carpal tunnel causes complications such as fat tissue atrophy and skin color changes. [bib_ref] The safest location for steroid injection in the treatment of carpal tunnel..., Racasan [/bib_ref] [bib_ref] Re: median nerve damage following local corticosteroid injection for the symptomatic relief..., Swan [/bib_ref] Therefore, accurate injection into the carpal tunnel is very important. [bib_ref] The safest location for steroid injection in the treatment of carpal tunnel..., Racasan [/bib_ref] Ultrasound (US)-guided injection can confirm successful injection within the carpal tunnel visually so as to reduce patient discomfort, as well as median nerve injury. Although many approaching methods to carpal tunnel injection have been described, there is no sufficient evidence of one superior technique compared to the others. [bib_ref] Intralesional therapy in carpal tunnel syndrome: a sonographic-guided approach, Grassi [/bib_ref] [bib_ref] Anatomical basis of ulnar approach in carpal tunnel injection, Kim [/bib_ref] [bib_ref] Sonographically guided carpal tunnel injections: the ulnar approach, Smith [/bib_ref] [bib_ref] A review of treatment for carpal tunnel syndrome, Wilson [/bib_ref] Currently, the most widely used CTS-injection method is the longitudinal approach, which paces the needle proximal to distal to the median nerve. [bib_ref] Sonographically guided carpal tunnel injections: the ulnar approach, Smith [/bib_ref] However, this method cannot offer the benefits of carpal tunnel content images from transverse plane and needle shaft and tip simultaneously even under US guidance. So, even an accurately placed needle within the carpal tunnel following the technique may injure the median nerve because of anatomic variations such as a bifid median nerve or a median nerve in an abnormal location. [bib_ref] Re: median nerve damage following local corticosteroid injection for the symptomatic relief..., Swan [/bib_ref] Smith et al 7 developed the ulnar approach methods for the performance of US-guided carpal tunnel injection in which the transducer is placed transversely along the wrist crease at the entrance to the carpal tunnel, and the needle is passed into the skin on the ulnar side of the proximal carpal tunnel at the level of the distal wrist crease. Technically, the ulnar approach combines the advantages of longitudinal needle visualization with the flexibility of transverse carpal tunnel imaging. In addition, this approach is easy to learn.
This study was conducted in order to evaluate the degree of symptom improvement, and the change of electrophysiological and sonographic findings after US-guided local steroid injection using an in-plane ulnar approach. We hypothesized that the inplane ulnar approach injection technique provides better outcomes compared with the out-plane approach or blind injection technique in CTS.
# Materials and methods
## Patient selection
Forty-four patients with mild to moderate idiopathic CTS with a neurophysiological confirmation were included in this study. Mild and moderate CTS were defined as slowing of the sensory conduction velocity and/or abnormal distal motor latency according to a validated CTS electrophysiological severity scale. [bib_ref] Bilateral clinicalneurophysiological assessment of median nerve in carpal tunnel syndrome patients, Padua [/bib_ref] We investigated outpatients with idiopathic CTS diagnosed by physical examination and electrodiagnosis. Participation was solicited without preselection. All patients had complaints of paresthesia or numbness in the median nerve distribution area of the hand with nocturnal worsening for a period of at least 3 months.
The exclusion criteria were symptomatic CTS because of diabetes, thyroid disease or rheumatic disease, age <18 years, pregnancy, previous treatment of CTS, previous fracture or deformities at the wrist, cervical radiculopathy, and other polyneuropathy. Approval from the Institutional Review Board of the Gachon University Gil Medical Center was obtained. All subjects participated in this study voluntarily, and informed consent on all aspects of the study was obtained.
## Study design and therapeutic intervention techniques
This is a prospective randomized single-blind clinical trial comparing in-plane ulnar approach carpal tunnel injection versus out-plane carpal tunnel injection and blind injection of 40 mg of triamcinolone in patients with idiopathic CTS in terms of efficacy and safety. The patients were maintained in the supine position, with the forearm supinated and the wrist placed in slight dorsiflexion position to prevent change in the appearance of the median nerve according to wrist position [fig_ref] FIGURE 1: Transducer position and needle approach for each technique [/fig_ref].
In blind injection, after skin antisepsis, the 26-gauge needle was inserted into the proximal carpal tunnel at the distal wrist crease just ulnar to the palmaris longus tendon. The outplane approach was performed using a perpendicularly placed transducer, and the needle was inserted into the proximal carpal tunnel at the distal wrist crease just ulnar to the palmaris longus tendon. [bib_ref] Re: median nerve damage following local corticosteroid injection for the symptomatic relief..., Swan [/bib_ref] The needle tip was identified as a moving reflector in real time as the tip passed obliquely from the skin surface to the proximal to distal carpal tunnel. In the in-plane ulnar approach, the transducer is moved ulnarly while keeping the median nerve in view [fig_ref] FIGURE 2: Transverse sonogram of the right carpal tunnel in a patient with idiopathic... [/fig_ref]. In this manner, the pisiform, ulnar nerve, and ulnar artery are brought into view [fig_ref] FIGURE 2: Transverse sonogram of the right carpal tunnel in a patient with idiopathic... [/fig_ref]. [bib_ref] Re: median nerve damage following local corticosteroid injection for the symptomatic relief..., Swan [/bib_ref] The pisiform appears as a prominent superiorly rounded hyperechoic structure on the ulnar side of the screen. The ulnar nerve lies just radial to the pisiform, and the ulnar artery lies radial to the ulnar nerve. Doppler images can confirm the position of the ulnar artery. [bib_ref] Sonographically guided carpal tunnel injections: the ulnar approach, Smith [/bib_ref] Although the optimal injectable location has not been determined, target sign which is produced by injectate as a ring form structure is used as a proper injection guideline.
A typical injectate consists of 1 mL of 40 mg/mL triamcinolone and 1 mL of 1% lidocaine, delivered in equal portions above the nerve, below the nerve, and into the subsynovial connective tissue. After completion of the injection, the distal carpal tunnel is scanned to ensure distribution of injectate throughout the proximal-to-distal extent of the carpal tunnel. [bib_ref] Sonographically guided carpal tunnel injections: the ulnar approach, Smith [/bib_ref] All injections were performed by the same physician. The USguided injections were performed using an US device (GE healthcare, Hertfordshire, UK).
## Review of clinical data
Patients were assessed at baseline, 4 weeks, and 12 weeks after the injection. The follow-up criteria were electrodiagnostic parameters, sonographic findings, and Boston Carpal Tunnel Questionnaire (BCTQ). Nerve conduction recordings were performed using a 2-channel electromyography machine, and bipolar handheld surface stimulating electrodes were used to obtain motor distal latency and compound motor action potentials (CMAPs) at abductor pollicis brevis muscle. [bib_ref] The safest location for steroid injection in the treatment of carpal tunnel..., Racasan [/bib_ref] The median nerve sensory distal latency (SDL) and sensory nerve action potential (SNAP) were evaluated at the second and fourth finger, and the ulnar nerve distal latency was evaluated at fourth digit to calculate median-to-ulnar sensory nerve distal latency ratio.
The median nerve cross-sectional area (CSA) was measured in transverse sections with ultrasonography looking for nerve enlargement at the carpal inlet and outlet. To identify the inlet, the bony landmarks of the scaphoid and pisiform were visualized, which in terms of surface markers approximates to the level of the distal wrist crease. The median nerve was then traced repeatedly from distal to proximal for identification of the maximal cross-sectional area. Once identified, the maximal CSA was evaluated and then the long diameter (LD), the short diameter (SD), and the flattening ratio (FR, SD/LD of the median nerve) were measured. [bib_ref] Anatomical basis of ulnar approach in carpal tunnel injection, Kim [/bib_ref] [bib_ref] Does measuring the median nerve at the carpal tunnel outlet improve ultrasound..., Paliwal [/bib_ref] For clinical evaluation, we used a validated questionnaire, the BCTQ at the inclusion date (baseline) and at 4-and 12-week follow-up. It is the most commonly used outcome measure of assessment for improvements in clinical symptoms and functional recovery of patients with CTS. The BCTQ evaluates the clinical symptoms (symptom severity scale, SSS), including pain, numbness, weakness, paresthesia, and clumsiness using 11 questions with the Likert scale preprinted 5 answers ranging from no complaints to very severe or continuous complaints. The functional handicap (functional status scale [FSS]) is calculated from 8 questions regarding difficulties with daily activities, including writing, holding a telephone, and so on, with a setup similar to the SSS. Each score is calculated as the mean of the responses of the individual items. A higher score indicates the worse symptom or function. [bib_ref] Sonographically guided carpal tunnel injections: the ulnar approach, Smith [/bib_ref] [bib_ref] Comparison of splinting, splinting plus local steroid injection and open carpal tunnel..., Halil Ucan [/bib_ref] Statistical Analysis Statistical analyses were performed using the Statistical Package for the Social Sciences Program (SPSS Version 10.0). The main characteristics of patients were evaluated with descriptive studies. The divergence of outcome measure between baseline and posttreatment scores in the third and sixth months for each subject was computed by a general linear model for repeated measures. The level of statistical significance was set as P < 0.05.
# Results
## Patient flowchart
Forty-four patients, 75 hands, were enrolled in this study. Demographics, symptom duration, and bilateral involvement of the patients are shown in [fig_ref] TABLE 1: Demographics and Symptom Duration of the Patients According to Injection Groups [/fig_ref]. Of the 44 patients diagnosed with CTS, the blind injection group included 15 patients, the out-plane US-guided injection group included 14 patients, and the in-plane ulnar approach US-guided injection group included 15 patients [fig_ref] FIGURE 3: Patients' flow diagram [/fig_ref].
## Electrodiagnostic findings
DML, CMAP amplitude, DSL, SNAP amplitude, and median-to-ulnar SDL ratio showed significant improvement at 4 and 12 weeks after the treatment in the in-plane ulnar approach. CMAP amplitude, SNAP amplitude, and SDL were improved at 4 and 12 weeks after out-plane US-guided injection. However, in blind injection, only SNAP amplitude showed statistical improvement after treatment.
At 4-and 12-week follow up, there were significant CMAP amplitudes showed significant improvement in the in-plane ulnar approach and out-plane US-guided injection, and the 2 methods were more efficient than blind injection. For SNAP amplitude at the second finger, significant improvements were observed in all the groups after injection. The in-plane ulnar approach was significantly superior to blind injection at 4 and 12 weeks, statistically. The median-to-ulnar SNAP amplitude ratio showed significant improvement in in-plane ulnar approach at 12 weeks after the treatment, and this method is statistically superior to blind injection or out-plane US-guided injection. There was no significant change in the median-to-ulnar ratio in blind injection or out-plane US-guided injection [fig_ref] TABLE 2: Electrophysiological Findings of the Patients According to Injection [/fig_ref].
## Sonographic findings
In the blind injection and out-plane US-guided injection group, there were no significant decreases in CSA and FR. In the in-plane ulnar approach group, CSA showed a statistically significant increase at 4 and 12 weeks, and its result was superior to that of the other 2 groups (P < 0.05). However, the FR showed no change in all the groups [fig_ref] TABLE 3: Sonographic Results of the Patients According to Injection [/fig_ref].
## Changes in bctq scores
Both the symptom (SSS) and functional scores (FSS) of the BCTQ showed significant improvement after 12 weeks in all patients (P < 0.05). Obviously, in the in-plane ulnar approach group, more rapid improvement was observed for SSS compared with other methods at 4 weeks. In the in-plane ulnar approach group, the SSS results at 4 and 12 weeks and the FSS result at 12 weeks were significantly superior to baseline [fig_ref] TABLE 4: Clinical and Functional Evaluation of Patients With BCTQ Values are mean AE... [/fig_ref].
## Complications
We checked complication after injection, including nerve insult, vessel insult, and skin lesion (eg, color change). Vessel insult was not detected in US-guided groups (out-plane USguided injection and in-plane ulnar approach) and nerve insult was not detected in in-plane ulnar approach [fig_ref] TABLE 5: Posttreatment Complication of the Patients [/fig_ref].
# Discussion
The primary purpose of this study is to compare the efficacy and safety of the in-plane ulnar approach in US-guided carpal tunnel injection with blind or out-plane US-guided injection. Although steroid injections are routinely administered for CTS, direct needle injury of the median nerve is the major complication of these injections. The safest location of the injection remains controversial. Several cases of median nerve injury during CTS injection have been reported. [bib_ref] The safest location for steroid injection in the treatment of carpal tunnel..., Racasan [/bib_ref] Racasan and Dubert 3 reported median nerve irritation following carpal tunnel injection as a frequent problem and identified the fact that the median nerve extends ulnar to the palmaris longus tendon in most hands. They reported that the safest location of injection is through the flexor carpi ulnaris tendon. [bib_ref] The safest location for steroid injection in the treatment of carpal tunnel..., Racasan [/bib_ref] In the US-guided injection, the structure and location can be seen, so that the physician can reach the carpal tunnel without causing damage to neighboring tissue, and it enables visualization of the distribution of the injected substance.
In addition, the clinician can adjust the distance between the median nerve and the needle. Thus, the physician can come as close to the nerve as possible. If we assume the location of the median nerve and the needle is shown passing from the ulnar aspect of the carpal tunnel to a position adjacent to the median nerve, the injection procedure is easier to learn and provides a greater degree of needle control.
The in-plane ulnar approach has many advantages. First, imaging the wrist in the transverse plane enables visualization of all of the carpal tunnel contents and tendon structure around the nerve, facilitating an accurate perineural injection. Racasan and Dubert 3 identified the fact that the median nerve extends ulnar to the palmaris longus tendon in most hands, the median Values are mean AE standard deviation. Blind ¼ blind injection group, CSA ¼ median nerve's cross-sectional area, FR ¼ flattening ratio (SD/LD of median nerve), LD ¼ long diameter, SD ¼ short diameter, Ulnar-I ¼ in-plane ulnar approach injection group, Ulnar-O ¼ out-plane injection group. P values were calculated by analysis of variance with post hoc Tukey-b test.
à P < 0.05 versus baseline. nerve can extend up to 13 mm beyond the ulnar side of the palmaris longus tendon. From previous studies, it appears that the safest location for insertion of a needle for injection is through the flexor carpi radialis tendon, proximal to the carpal tunnel. [bib_ref] The safest location for steroid injection in the treatment of carpal tunnel..., Racasan [/bib_ref] However, the in-plane approach allows the needle to reach the median nerve directly with visualization of the whole structure of the carpal tunnel, and there is no penetrating tendon. Second, both the needle tip and shaft can be visualized in plane relative to the transducer throughout the procedure. Transverse sonogram of the carpal tunnel in a patient with idiopathic CTS showed swelling of the median nerve. A 27gauge needle is shown passing from the ulnar aspect of the carpal tunnel to a position adjacent to the median nerve, and after positioning the needle tip next to the nerve, the local anesthetic-corticosteroid mixture is injected in order to peel the nerve off the overlying flexor retinaculum via hydrodissection. Anechoic injectate is shown surrounding the deep surface of the median nerve and separating it via hydrodissection from the more deeply positioned hyperechoic flexor tendons and associated synovium. This hydrodissection may disrupt adhesions and encircling of the superior portion of median nerve. Under realtime visualization, additional injectate is delivered with redirected needle to the inferior portion of the nerve. Then, the median nerve was separated from the underlying subsynovial connective tissue. [bib_ref] Sonographically guided carpal tunnel injections: the ulnar approach, Smith [/bib_ref] The most important advantage of the in-plane ulnar approach injection appears to be its view containing nerve and vessel with a visible needle. Accordingly, novice practitioners can approach the median nerve more easily from basic anatomy of the target tissue.
The current study does have some limitations, primarily the small patient group with female predominance and the lack of long-term follow-up. However, correlation of median nerve conduction parameters, subjective symptoms, and sonographic findings of the patients with injection methods was precisely proven. Likewise, we checked several complications after injection, including nerve insult, vessel insult, and skin lesion. Although there were no statistically significant findings, vessel insult was not detected in US-guided groups (out-plane USguided injection and in-plane ulnar approach) and nerve insult was not detected in the in-plane ulnar approach. Subsequent to this study, we recommend that additional studies would require very large sample sizes in order to achieve more reliable results.
# Conclusion
Although in-plane and out-plane US-guided carpal tunnel injection were more effective in improving electrodiagnostic, sonographic findings, and symptoms than blind injection, the in-plane ulnar approach was superior to the out-plane and blind injection in median-to-ulnar sensory nerve distal latency ratio SDL ratio, CSA, and BCTQ result. US-guided local steroid injection using an in-plane ulnar approach in the CTS can be more effective than out-plane or blind injection. Blind ¼ blind injection group, Ulnar-I ¼ in-plane ulnar approach injection group, Ulnar-O ¼ out-plane injection group.
[fig] FIGURE 1: Transducer position and needle approach for each technique. (A) In-plane ulnar approach US-guided carpal tunnel injection, (B) out-plane US-guided carpal tunnel injection, and (C) blind injection. US ¼ ultrasound. [/fig]
[fig] FIGURE 2: Transverse sonogram of the right carpal tunnel in a patient with idiopathic CTS. A 27-gauge needle is shown passing from the ulnar aspect of the carpal tunnel to a position adjacent to the median nerve. (A) After positioning the needle tip next to the nerve, the local anesthetic-corticosteroid mixture is injected in order to peel the nerve off the overlying flexor retinaculum via hydrodissection. (B) Anechoic injectate is shown surrounding the deep surface of the median nerve and separating it via hydrodissection from the more deeply positioned hyperechoic flexor tendons and associated synovium. CTS ¼ carpal tunnel syndrome. [/fig]
[fig] FIGURE 3: Patients' flow diagram. BCTQ ¼ Boston Carpal Tunnel Questionnaire, CTS ¼ carpal tunnel syndrome, EMG ¼ electromyography, I ¼ in-plane, O ¼ in-plane. [/fig]
[table] TABLE 1: Demographics and Symptom Duration of the Patients According to Injection Groups [/table]
[table] TABLE 3: Sonographic Results of the Patients According to Injection [/table]
[table] TABLE 4: Clinical and Functional Evaluation of Patients With BCTQ Values are mean AE standard deviation. BCTQ ¼ Boston Carpal Tunnel Questionnaire, Blind ¼ blind injection group, FSS ¼ functional status scale of BCTQ, SSS ¼ symptom severity scale of the BCTQ, Ulnar-I ¼ in-plane ulnar approach injection group, Ulnar-O ¼ out-plane injection group. P values were calculated by analysis of variance with post hoc Tukey-b test. [/table]
[table] TABLE 2: Electrophysiological Findings of the Patients According to Injection [/table]
[table] TABLE 5: Posttreatment Complication of the Patients [/table]
|
Change in physical activity during a weight management intervention for breast cancer survivors: Association with weight outcomes
Objective-The study examined the effects of a group-phone based weight management intervention on change in physical activity as measured via accelerometer and self-report in rural breast cancer survivors. The study also evaluated the role of physical activity on clinically meaningful cut-points for weight loss (baseline to 6 months) and weight loss maintenance (6 to 18 months).Methods-Participants were breast cancer survivors in a weight management intervention who provided valid weight and accelerometer data (N=142). We categorized participants into four groups based on weight loss ≥10% and weight regain ≥5% at 18 months.Results-Accelerometer-measured moderate-to-vigorous physical activity (MVPA) significantly increased from baseline to 6 months (+46.9 minutes). MVPA declined during maintenance; however remained significantly greater than baseline. Self-reported MVPA followed a similar pattern as accelerometer MVPA, but estimates were significantly higher. Participants in the high loss, low regain group had significantly higher MVPA at all points.Conclusions-A distance-based weight management intervention for survivors improved physical activity outcomes over 18 months. Self-reported physical activity was substantially higher than accelerometer-measured. Findings highlight the importance of device-based measurement for characterizing the magnitude of physical activity change, as well as the role of physical activity in weight management outcomes.
# Introduction
Obesity and physical activity are modifiable risk factors for breast cancer and cancer recurrence [bib_ref] Effect of obesity on survival of women with breast cancer: systematic review..., Protani [/bib_ref]. Physical activity may reduce breast cancer risk directly through its effects on chronic inflammation and sex hormones, as well as indirectly through its effects on obesity-related factors such as adiposity, insulin resistance, and adipokines. During weight loss, moderate to vigorous physical activity (MVPA) helps induce caloric expenditure to reach a prescribed caloric deficit needed for weight loss. A systematic review of weight loss studies targeting both diet and physical activity changes over 3-12 months reported a small, positive additive effect of physical activity during weight loss (pooled estimate = −1.65 kg) [bib_ref] The Impact of Interventions that Integrate Accelerometers on Physical Activity and Weight..., Goode [/bib_ref]. Whereas physical activity may have a smaller role in weight loss, physical activity is crucial for successful long-term weight loss maintenance [bib_ref] Objective physical activity and weight loss in adults: the step-up randomized clinical..., Jakicic [/bib_ref]. The American College of Sports Medicine recommends 200-300 MVPA min/week for sustained weight loss maintenance [bib_ref] American College of Sports Medicine Position Stand. Appropriate physical activity intervention strategies..., Donnelly [/bib_ref]. In this regard, Jakicic (2014) examined weight loss maintenance patterns among participants from the general population and found that those who sustained ≥10% weight loss at 6 and 18 months were more likely to complete 200+ minutes of MVPA per week [bib_ref] Objective physical activity and weight loss in adults: the step-up randomized clinical..., Jakicic [/bib_ref].
Comprehensive clinical weight management trials tailored to breast cancer survivors that targeted reduced caloric intake, increased energy expenditure, and behavioral strategies have yielded initial evidence of physical activity improvements during initial (up to 6 months) and more long-term (up to 18-24 months) intervention periods [bib_ref] Weight Loss Randomized Intervention Trials in Female Cancer Survivors, Chlebowski [/bib_ref]. The Exercise and Nutrition to Enhance Recovery and Good Health for You (ENERGY) trial included a home-based (nonsupervised) physical activity component [bib_ref] Reducing breast cancer recurrence with weight loss, a vanguard trial: The Exercise..., Rock [/bib_ref]. During weight loss (0 to 6 months), selfreported MVPA min/wk increased 150% in the intervention group (94 to 238 min/wk), and increased by 50% in the non-intervention control group (103 to 163 min/wk) [bib_ref] Reducing breast cancer recurrence with weight loss, a vanguard trial: The Exercise..., Rock [/bib_ref]. Similar improvements were found in 18-month outcomes of the LISA weight loss trial [bib_ref] Randomized trial of a telephone-based weight loss intervention in postmenopausal women with..., Goodwin [/bib_ref]. Participants randomized to the phone-based intervention had self-reported improvements from 0 to 18 months (110% increase: 120 to 250 min/wk), as did participants in the mailbased intervention (122% increase: 95 to 210 min/wk) (11).
Measuring physical activity using self-report is common in lifestyle interventions for breast cancer survivors; it affords low participant burden and can be highly scalable large trials. However, measuring physical activity using self-report questionnaires can overestimate physical activity as compared to device-based measurement with accelerometers [bib_ref] Comparison between logbookreported and objectively-assessed physical activity and sedentary time in breast..., Mazzoni [/bib_ref] [bib_ref] A comparison of direct versus self-report measures for assessing physical activity in..., Prince [/bib_ref] , possibly because these measures assess different aspects of physical activity or different constructs [bib_ref] A comparison of direct versus self-report measures for assessing physical activity in..., Prince [/bib_ref]. For example, self-reported physical activity participation may also reflect an individual's perceived ease or difficulty in physical activity participation [bib_ref] A comparison of direct versus self-report measures for assessing physical activity in..., Prince [/bib_ref]. To our knowledge, only two pilot weight loss trials breast cancer survivors reported findings on physical activity change using device-based measurement. One single-arm pilot study designed to provide extended care for weight loss maintenance via text messages indicated that accelerometer-measured MVPA significantly increased from baseline to 18 months; however the magnitude of the increase was lower (38% increase) than the aforementioned self-reported physical activity increases in other breast cancer trials [bib_ref] Efficacy of a Text Message-Delivered Extended Contact Intervention on Maintenance of Weight..., Spark [/bib_ref]. In the Lifestyle, Exercise, and Nutrition (LEAN) randomized pilot study, pedometer steps increased by 37% in the in-person treatment group and 16% in the telephone treatment group from baseline to 6 months [bib_ref] Randomized Trial Comparing Telephone Versus In-Person Weight Loss Counseling on Body Composition..., Harrigan [/bib_ref]. Importantly, participants in both LEAN treatment groups self-reported a higher increase in MVPA min/wk (86-115%). Thus, self-report of physical activity may overestimate treatment effects on physical activity among breast cancer survivors. This has important implications for addressing unanswered questions about the relative importance of weight loss versus physical activity change in prognostic factors for recurrence and survival. However, given that pedometer-measured physical activity can yield substantial measurement error [bib_ref] Measurement Effects of Seasonal and Monthly Variability on Pedometer-Determined Data, Kang [/bib_ref] and no self-report measures were reported for comparison in the accelerometer study, it is difficult to determine the degree to which self-report may overestimate physical activity in treatment samples of breast cancer survivors.
The goals of the current study were threefold. The first aim was to examine the effects of an 18-month group phone-based weight loss intervention for rural breast cancer survivors on change in physical activity from 0 to 6 months (weight loss phase) and from 6 to 18 months (weight loss maintenance phase). Our second aim was to compare estimates of physical activity change across accelerometer versus self-report measures. The third aim was to compare physical activity changes across weight loss and weight regain groups as defined by clinically meaningful cut-points.
# Methods
## Study design
This study was designed to compare continued group phone-based counseling versus mailed newsletters on weight loss maintenance following weight loss. Study details and primary outcomes were presented previously [bib_ref] Weight loss maintenance strategies among rural breast cancer survivors: The rural women..., Befort [/bib_ref] [bib_ref] Effective recruitment of rural breast cancer survivors into a lifestyle intervention, Befort [/bib_ref] , and briefly described herein. Participants (N=210) were postmenopausal female breast cancer survivors residing in rural areas of the Midwestern United States, with a BMI of 27 to 45 kg/m, age <75 years, diagnosis of Stage 0-IIIc disease, and completion of breast cancer treatment.
The intervention consisted of 1) a 6-month weight loss phase where all participants received weekly group phone sessions, followed by 2) a 12-month weight loss maintenance phase (6 to 18 months) during which participants were randomized to continued phone sessions or a newsletter condition. Participants in the phone group received biweekly phone sessions, whereas newsletter condition participants received mailed newsletters at the same frequency as the phone sessions that covered the same content as the group calls.
The primary endpoint for the main trial was weight was regain from 6-18 months among participants who lost ≥5% by 6 months. However, all participants continued in the intervention and were followed throughout the study and are reported here. The study was approved by the University's Institution Review Board and informed consent was obtained from all participants.
## Study sample
The sample for the current analyses consisted of participants who provided valid weight and accelerometer data at baseline, 6, and 18 months. depicts how we obtained our final sample of N=142.
## Intervention
During the 6-month weight loss phase, participants met via conference call for one hour weekly. Participants were instructed to follow a structured meal plan and to gradually increase their physical activity, with the goal of 225 min/wk of MVPA by week 12, consistent with recommendations [bib_ref] American College of Sports Medicine Position Stand. Appropriate physical activity intervention strategies..., Donnelly [/bib_ref]. Participants received a Physical Activity Tool Kit with two DVDs, a pedometer, and self-monitoring charts. Brisk walking was the activity of choice for most participants. Physical activity sessions addressed both education to ensure safety, monitor and regulate intensity, and increase exercise variety, as well as problemsolving activities to address common barriers, including issues regarding the built environment in rural settings (lack of sidewalks, extreme temperatures, etc), sustaining motivation, and social support. Participants submitted a weekly self-monitoring report form detailing their dietary intake, physical activity minutes at ≥10 minute bouts, and steps/day (goal of 10,000/day). During each session, participants reported whether they met nutrition and physical activity goals from the previous session.
## Assessments
Participants attended assessment visits at baseline, 6, 12, and 18 months. At each visit, participants were weighed in light clothing (shorts, t-shirt) in a fasting state using a calibrated digital scale accurate to 0.1 kg (Befour PS5700). Height was measured with a stadiometer.
Both self-report and accelerometer measures of physical activity were collected at all time points. Participants were instructed to wear a GT3X+ Actigraph Accelerometer (Fort Walton Beach, FL) for seven consecutive days at each point. The ActiGraph has been shown to provide valid assessments of activity intensity during both walking/running [bib_ref] Reexamination of validity and reliability of the CSA monitor in walking and..., Brage [/bib_ref] and daily living activities [bib_ref] Validity of four motion sensors in measuring moderate intensity physical activity, Bassett [/bib_ref]. The device does not have a display screen to minimize reactivity. Participants received verbal and written instructions accompanied by a wear time log to help encourage adherence. Devices were returned in a pre-stamped envelope. Participants' data were included if data was available for ≥10 hours/day for ≥ 4 days, an algorithm that has been demonstrated to validly estimate physical activity patterns [bib_ref] Accelerometer data reduction: a comparison of four reduction algorithms on select outcome..., Mâsse [/bib_ref]. The accelerometer outcome variable was total ≥10-minute MVPA bouted minutes per week, consistent with the intervention guidelines for counting only ≥10 minute bouts toward total weekly minutes. Prior studies found this measure to be more predictive of weight loss outcomes than nonbouted minutes [bib_ref] Objective physical activity and weight loss in adults: the step-up randomized clinical..., Jakicic [/bib_ref]. Data in counts per minute summed across 10-second epochs were downloaded and number of minutes per day at various activity levels were calculated using the cut-points suggested by Matthews et al. [bib_ref] Sources of variance in daily physical activity levels as measured by an..., Matthews [/bib_ref]. Moderate to vigorous activity (counts ≥1952 per min) bouts wherein at least 8 minutes were at/above the 1952 threshold were used to identify 10-minute MVPA bouts. Bouted MVPA minutes per valid day (≥10 hours worn) were calculated and multiplied by 7 to obtain weekly estimates.
The Paffenbarger Physical Activity Questionnaire (23) includes questions on stairs climbed, blocks walked, and other sports, leisure, and recreational activities on a typical week during the past month. All activities were assigned MET values [bib_ref] Compendium of Physical Activities: a second update of codes and MET values, Ainsworth [/bib_ref] and MVPA minutes (MET values >3) were summed to obtain weekly MVPA min/week estimates. Total MVPA min/wk was used as the self-report outcome variable to allow for direct comparison to the accelerometer measure. Self-report physical activity measures were available for 139 of the 142 participants in the sample (98%). Two participants at 6 months and one at 18 months had self-report measures that were statistical outliers and were physically implausible.
## Weight-change groupings
We classified participants into low/high weight loss (< 10% loss versus ≥ 10% loss from baseline to 6 months) and further into low/high weight regain (<5% regain versus ≥ 5% regain from 6 to 18 months). This resulted four weight-change groups (see [fig_ref] Table 2: Accelerometer and Self-Reported Physical Activity by TimeAccelerometer bouted MVPA min/week [/fig_ref] for sample sizes).
## Analyses
We first calculated descriptive statistics for the outcome measures, as well as the percentage of participants who met physical activity guidelines of ≥ 150 min/wk. Both outcomes were square-root transformed to correct for skewedness. We examined the effects of randomization condition at 6 months on MVPA change during maintenance (6-18 months) using a linear mixed model. There were no significant differences in MVPA change between the conditions; therefore the groups were combined for the main analyses.
Separate repeated measures ANOVAs were run to examine changes in accelerometer and self-reported physical activity during the weight loss phase (baseline to 6 months) across the full sample, as well as the four weight-change groups. A Bonferroni adjustment was used for multiple comparisons between groups. A series of paired t tests were used to test for significant differences between accelerometer-measured and self-reported MVPA at each time point. Separate linear mixed models using a compound symmetry correlation structure were constructed to examine changes in accelerometer and self-reported MVPA outcomes across the three time points during weight loss maintenance phase (6,12, and 18 months) for the full sample, as well as the weight-change groups. Finally, an interaction between time and weight-change group was included to examine differences between groups by time point, using a Bonferroni adjustment. We included the following variables as covariates in the models: randomization assignment, age, education, rurality (large rural vs. small/isolated rural (25)), and time since end of cancer treatment. We used Restricted Maximum Likelihood Estimation (REML) to cope with missing data. presents participant characteristics. Participants were a mean of 58.6 (SD =8.0) years old, 76% were married and 22% had a 4-year college degree. Participants were a mean of 3.6 (SD =2.5) years beyond cancer treatment excluding anti-hormone therapy. Mean baseline BMI was 33.7 (4.0). Weight loss at 6 months was 13.9% (SD= 5.74) and participants regained a mean of 4.6% (SD = 5.9) by 18 months.
# Results
Participants in the current sample did not significantly differ on age, baseline BMI, or cancer treatment-related variables (all p values > .05) from those who attended the 6-month visit but were excluded from the current analyses based on missing or invalid accelerometer data. Additionally, participants in the current sample did have significantly higher percent weight loss at 6 months (13.9% vs 10.1%, p = .001) than those who attended the 6-month visit but were excluded from the current analyses based on missing or invalid accelerometer data. However, these participants did not significantly differ in percent weight regain from 6 to 18 months (4.6% vs 6.4%, p = .083). [fig_ref] Table 2: Accelerometer and Self-Reported Physical Activity by TimeAccelerometer bouted MVPA min/week [/fig_ref] shows within-subject changes in MVPA over 18 months. Accelerometer-measured median MVPA bouted minutes significantly increased from baseline to 6 months (18.4 vs 65.3, p = .001), an increase of 350% from baseline, and significantly decreased from 6 to 18 months (65.3 vs 38.1, p = .01; 42% decrease from 6 months). MVPA at 18 months remained significantly higher than baseline (18.4 vs 38.1, p = .001; 210% increase).
## Change in physical activity
Within the weight-change groups, accelerometer MVPA significantly increased from baseline to 6 months among both high weight loss groups, but not the low loss groups. In addition, accelerometer MVPA significantly decreased from 6 to 18 months in the high loss groups, but not in the low loss groups.
Self-reported MVPA min/week significantly increased in the full sample (0 vs 227.5, p = . 001; 227% increase), as well as in all weight-change groups (p values =.001) from baseline to 6 months. From 6 to 18 months, self-reported MVPA min/week significantly decreased in the full sample (227.5 vs 150.0; a 34% decrease), as well as in all groups except the low loss, high regain group [fig_ref] Table 2: Accelerometer and Self-Reported Physical Activity by TimeAccelerometer bouted MVPA min/week [/fig_ref]. Self-reported MVPA min/week at 18 months were 150% higher than baseline. shows the percentage of participants in each weight-change group that met MVPA guidelines. Accelerometer data indicated <15% of participants in both low loss groups met MVPA guidelines of ≥ 150 min/wk at any time point. In contrast, 30-43% of participants in the high loss groups met ≥ 150 min/wk guidelines at 6 months. In the high loss, low regain group, 25% met guidelines at 12 months, which increased to 32% at 18 months.
By self-report, over 60% of participants in all weight-change groups reported meeting ≥150 min/wk at 6 months. At 12 and 18 months, 26-39% of participants in the low loss groups reported meeting guidelines compared to 47-75% in the high loss groups.
## Accelerometer vs self-reported mvpa
Self-reported MVPA was significantly higher than accelerometer-measured MVPA at 6 months (p= .001) and 18 months (p=.001) in the full sample. Findings were similar among all four weight-change groups [fig_ref] Table 2: Accelerometer and Self-Reported Physical Activity by TimeAccelerometer bouted MVPA min/week [/fig_ref].
## Physical activity across weight-change groups
Group comparisons for accelerometer-measured MVPA over time are depicted in . There were no baseline group differences in MVPA. There was a significant interaction between group and time (p= .031). At 6 months, median MVPA in both of the high loss groups was significantly higher compared to the low loss groups (p values <.05). At 12 and 18 months, the high loss, low regain group had significantly higher MVPA than all other groups. Notably, the high loss, low regain group's MVPA decreased at 12 months but stabilized by 18 months, while the high loss, high regain group's MVPA decreased at 12 months and 18 months.
# Discussion
The present study demonstrated that a distance-based weight management intervention for breast cancer survivors that targeted home-based physical activity improved physical activity outcomes over 18 months. Specifically, MVPA assessed via accelerometer significantly increased (+46.9 minutes) from baseline to 6 months in the full sample. MVPA bouted minutes did decline during weight loss maintenance ; however MVPA at 18 months was still significantly greater than baseline, suggesting some maintenance of effects. Our observed changes in MVPA (Median=46.9 min) are consistent with the increases in MVPA min/wk observed in The Weight Loss Maintenance Randomized Controlled Trial in the general population (Mean= 48 min at 6 mo) [bib_ref] Comparison of strategies for sustaining weight loss: the weight loss maintenance randomized..., Svetkey [/bib_ref]. Compared to previous weight loss intervention trials in breast cancer survivors, self-reported physical activity changes from baseline to 18 months among our participants (150%) were similar in magnitude to the LISA phone-intervention condition (110%: 120 to 250 min/wk) [bib_ref] Randomized trial of a telephone-based weight loss intervention in postmenopausal women with..., Goodwin [/bib_ref] and higher than those of the ENERGY trial (79%, 94 to 168 min/wk). Participants in our study had lower physical activity at baseline compared to these other trials, a finding that may reflect environmental, cultural, and access-related barriers to physical activity in rural settings [bib_ref] Obesity and physical inactivity in rural America, Patterson [/bib_ref] [bib_ref] Determinants of leisure time physical activity in rural compared with urban older..., Wilcox [/bib_ref].
Only a minority of our participants achieved a level of physical activity consistent with guidelines, and there were no substantial differences between the phone and mail-based maintenance interventions during the maintenance phase. Novel strategies are needed to improve adherence to physical activity recommendations in non-supervised settings, where the broadest population impact is likely to occur [bib_ref] NIH working group report: Innovative research to improve maintenance of weight loss, Maclean [/bib_ref]. Our previous qualitative findings indicated that participants appeared to underestimate the impact of environmental barriers on their physical activity, which may be a point of focus for future interventions [bib_ref] A qualitative evaluation of a group phone-based weight loss intervention for rural..., Fazzino [/bib_ref]. In a recent pilot study examining tailored text messages for promoting weight loss maintenance, Spark et al found a drop in physical activity from 6 to 12 months, but a return to 6 month physical activity levels by 18 months [bib_ref] Measurement Effects of Seasonal and Monthly Variability on Pedometer-Determined Data, Kang [/bib_ref]. This finding suggests maintenance of effects is possible in this population and warrants further attention.
Our findings add to the preliminary evidence indicating self-reported physical activity can be substantially inflated as compared to device-based measurement among breast cancer survivors [bib_ref] Comparison between logbookreported and objectively-assessed physical activity and sedentary time in breast..., Mazzoni [/bib_ref] [bib_ref] Efficacy of a Text Message-Delivered Extended Contact Intervention on Maintenance of Weight..., Spark [/bib_ref] [bib_ref] Randomized Trial Comparing Telephone Versus In-Person Weight Loss Counseling on Body Composition..., Harrigan [/bib_ref]. The discrepancy between self-report and accelerometer measures in the current study was striking; median self-reported MVPA at 6 and 18 months was more than triple accelerometer-measured MVPA. Self-report estimates were particularly inflated in the low loss groups at 6 months, and in all groups except the high loss, low regain group at 18 months, indicating bias in self-reported physical activity may be greater among participants experiencing poorer weight loss and maintenance outcomes. Thus, these findings underscore the need for objective measurement of physical activity, particularly during weight loss trials during which social desirability to report physical activity adherence may be high and may influence study conclusions regarding the role of physical activity in weight management. Our findings further highlight the importance of using objective measurement of physical activity in studies that aim to disentangle the relative effects of diet, physical activity, and weight loss on biomarker modulation and disease outcomes among breast cancer survivors.
Our findings are consistent with those of Jakicic et al (2014) who reported a similar pattern of MVPA stabilization at 12 and 18 month among participants from the general population who successfully maintained 10% weight loss at 18 months. Taken together, these findings highlight the importance of preventing or minimizing the decline in physical activity during maintenance. A stepped care approach focused on sustaining or further increasing MVPA during maintenance may have benefits for preventing weight regain long-term. Given that motivation and engagement tend to decline during maintenance, a shift in intervention focus from primarily maintaining compliance with weight management skills learned during weight loss to expanding participants' physical activity goals and participation (PA type, duration, and frequency) may provide novelty and variety that could better sustain physical activity and may thus improve weight management outcomes.
The study had several limitations. First, our analyses were limited to participants who provided valid accelerometer wear time, and this group had higher weight loss compared to those without valid accelerometer data. If we had obtained data from the excluded participants, our estimates of physical activity during the program may have been lower overall. However, the percentage of participants who provided valid accelerometer data was high (89-92%) and in line with national cohort research that used accelerometers to measure physical activity [bib_ref] Obtaining Accelerometer Data in a National Cohort of Black and White Adults, Howard [/bib_ref] , as well as with some other intervention studies in the literature [bib_ref] Effect of physical activity on weight loss, energy expenditure, and energy intake..., Delany [/bib_ref] [bib_ref] Electronic feedback in a diet-and physical activity-based lifestyle intervention for weight loss:..., Shuger [/bib_ref] , although it is difficult to compare with most weight management trials because valid wear time statistics are often not reported [bib_ref] Efficacy of a Text Message-Delivered Extended Contact Intervention on Maintenance of Weight..., Spark [/bib_ref] [bib_ref] Comparison of strategies for sustaining weight loss: the weight loss maintenance randomized..., Svetkey [/bib_ref] , a problem identified previously (34). Additionally, our sample was comprised of primarily White, older survivors living in the rural Midwest and may not generalize to other groups. Also, our MVPA outcome variables were skewed. While we coped with this statistically using transformed variables as is commonly done [bib_ref] Objective physical activity and weight loss in adults: the step-up randomized clinical..., Jakicic [/bib_ref] , this issue results from a minority of participants meeting recommended physical activity levels, a problem which requires further attention. Finally, participants reported on their physical activity during a typical week in the past month, whereas accelerometers were worn the week directly after each visit. Thus, it is possible that participants' reports of their typical physical activity over the past month might not directly align with their actual physical activity the following week. Future research should examine agreement among accelerometer and self-report measures for the same week [bib_ref] Comparison between logbookreported and objectively-assessed physical activity and sedentary time in breast..., Mazzoni [/bib_ref].
The major strength of the study is the use of both accelerometer and self-reported measures of physical activity in addition to the broad range of high and low weight loss and regain groups, which allowed for comparisons across the four corresponding subgroups.
# Conclusions
A phone-based weight management intervention for breast cancer survivors that targeted non-supervised physical activity improved MVPA min/wk at 6 months, with partial maintenance of effects at 18 months. However, the majority of participants did not meet guideline thresholds, pointing to the continued need for enhancing intervention effects on physical activity. Self-reported physical activity changes were more than triple those of the accelerometer-measured changes, suggesting that measurement is crucial for accurately specifying the magnitude of physical activity changes during a weight management intervention. The highest levels of sustained physical activity were important for clinically significant weight loss, as well as successful weight loss maintenance by 18 months.
## Study importance
- Clinical weight loss trials for breast cancer survivors have yielded initial evidence of physical activity improvements; however most have assessed physical activity using only self-report. Additionally, little is known about the role of physical activity during weight maintenance among breast cancer survivors.
## -
The present study used both accelerometer and self-reported measures of physical activity in what, to our knowledge, is the largest weight loss trial to date among breast cancer survivors reporting physical activity changes across these two measurement methods.
## -
The study evaluated the role of physical activity on clinical meaningful cutpoints for weight loss (baseline to 6 months) and weight loss maintenance (6 to 18 months) among survivors. Accelerometer measured and self-reported physical activity *Significantly different at p<.05 level **Significantly different at p<.01 level a comparator group .001
Paffenbarger self -reported MVPA min/week P value corresponding to within subjects change in linear mixed model c P value corresponding to paired t-test Note: Low loss, low regain: <10% loss at 6 months, <5% regain at 18 months Low loss, high regain: <10% loss at 6 months, ≥ 5% regain at 18 months High loss, low regain: ≥ 10% loss at 6 months, <5% regain at 18 months High loss, high regain: ≥ 10% loss at 6 months, ≥ 5% regain at 18 months
[fig] Figure 1, Figure 2: Flow [/fig]
[table] 34: Murphy SL. Review of physical activity measurement using accelerometers in older adults: Considerations for research design and conduct. Prev Med. 2009; 48:108-114. [PubMed: 19111780] [/table]
[table] Table 2: Accelerometer and Self-Reported Physical Activity by TimeAccelerometer bouted MVPA min/week [/table]
|
Role of electrocardiographic early repolarization pattern in long-term outcomes of a community-based middle-aged and geriatric ambulatory population: a prospective cohort study
In some studies, electrocardiographic early repolarization pattern (ERP) has been associated with an increased risk of death from cardiac causes. However, little is known about the prognostic significance of ERP in the middle-aged and geriatric general populations. We investigated the prevalence and long-term prognostic significance of early repolarization pattern (ERP) on electrocardiograms (ECGs) in the Healthy Aging Longitudinal Study (HALST) cohort of 4615 middle-aged and geriatric community-dwelling Han Chinese adults from Taiwan. The study subjects were followed-up for 95±22 months. A positive ERP of 0.1 mV was observed in 889 (19.3%) of the subjects. Kaplan-Meier survival curve analysis showed that ERP was not associated with all-cause and cardiovascular mortality (log-rank test, P=0.13 and 0.84, respectively). Cox regression analysis after adjusting for covariables revealed that age, blood pressure, smoking, diabetes, stroke, chronic kidney disease, and corrected QT interval (QTc) were associated with increased risk of all-cause mortality (P<0.05). Age, and stroke were risk factors associated with increased risk of cardiovascular mortality (P<0.05). However, ERP alone was not associated with all-cause or cardiovascular mortality. These findings show that ERP is common in the middle-aged and geriatric Han-Chinese individuals from the HALST cohort and is not associated with all-cause or cardiovascular mortality.
# Introduction
Sudden cardiac death (SCD) is a major health issue worldwide and accounts for 180,000 to 250,000 deaths annually in the United States . The ageadjusted incidence of SCD in the United States is 60 per 100,000 population. Ventricular tachy-arrhythmia is the major cause of SCD, and is not associated with any structural heart disease in 6 to 14% of cases . In some cases, SCD is associated with electrocardiographic abnormalities that affect ventricular repolarization, such as long or short QT syndrome . Early repolarization pattern (ERP) is characterized by elevation of the QRS-ST junction or AGING the J-point in a surface 12-lead electrocardiogram (ECG). Although ERP was considered benign, recent studies have suggested its potential association with cardiac arrhythmogenicity. In case-control studies, the presence of ERP in the inferior or lateral leads is associated with susceptibility to ventricular fibrillation and SCD in patients without structural heart disease.
ERP is a common electrocardiographic finding that affects 1% to 13% of adults and is more common in young athletic males. The age of individuals enrolled in previous studies regarding ERP ranged widely from 25 to 95 years. Sinner et al conducted a large, prospective, population-based casecohort study of individuals of Central-European descent (MONICA/KORA) and reported a high prevalence of ERP (13.1%) in individuals aged between 35-74 years and a 2-to 4-fold increased risk of cardiovascular mortality in individuals with ERP and aged between 35-54 years. reviewed ECG records of 5976 atomic bomb survivors in Nagasaki, Japan with a mean age of 47.2 years and reported that ERP was associated with elevated risk of unexpected death and decreased risk of cardiac and all-cause death. However, the long-term prognostic significance of ERP is poorly characterized in older middle-aged and elderly population.
In this study, we investigated the prevalence and prognostic value of ERP regarding cardiac and all-cause mortality in a large, multi-site, Healthy Aging Longitudinal Study (HALST) cohort consisting of older middle-aged and elderly adults belonging to the Han Chinese population in Taiwan.
# Results
## Study participants
The flowchart of enrollment and inclusion criteria of the study subjects is shown in. We initially recruited 5,380 relatively healthy and ambulatory individuals from 7 communities across Taiwan. We then excluded 755 individuals with cancer and underlying severe cardiovascular diseases (e.g., myocardial infarction or pacemaker implantation), as well as those with missing follow-up information. Finally, we included 4,615 healthy individuals in this study. The prevalence of ERP in the study cohort was 19.3% (n=889/4,615). In the ERP (+) group, 122 out of 889 individuals died during follow-up (65.51±27.12 months). In the ERP (-) group, 561 out of 3726 individuals died during follow-up (62.63±28.44 months).
The baseline clinical and demographic data of the study population is shown in. The mean age was 69.1±8.2 years and 47.8% of the study subjects were males. The average age of study participants was lower and the proportion of males and current smokers was higher in the ERP (+) group compared to the ERP (-) group. The mean values for the body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), and hypertension were significantly lower (all P<0.005) and the length of the corrected QT interval (QTc) was shorter (P<0.001) in the ERP (+) group compared to the ERP (-) group.shows the representative ECGs of few selected individuals belonging to ERP (+) and ERP (-) groups. ERP was observed in the inferior and lateral leads, both separately and simultaneously. Among the 4,615 healthy individuals, 574 (12.4%) individuals were ERP-positive in the inferior leads (ERP-inf + ), 214 (4.6%) individuals were ERP-positive in the lateral leads (ERP-lat + ), and 101(2.2%) individuals were ERP-positive in both lateral and inferior leads (ERP-lat + inf + ). The agreement between the two initial interpreters was moderate to high (k=0.97, agreement proportion= 0.98).
## Association between positive erp and all-cause and cardiovascular mortality
We observed that all-cause and cardiovascular mortality rates were similar in ERP (+) and ERP (-) groups during a mean follow-up time of 95.1±21.9 months (log-rank test, P=0.13 and P=0.84, respectively;, 3B). Since previous studies showed that ERP in the inferior leads was a risk factor for all-cause and cardiovascular mortality, we performed Kaplan-Meier survival curve analysis to investigate the association between the two types of mortality and positive ERP in the inferior leads (ERPinf + ). The results showed that all-cause and cardiovascular mortality rates were similar for individuals with and without ERP in the inferior leads (log-rank test, P=0.58, 0.66,, 3D).
## Long-term outcomes of all-cause mortality and cardiovascular mortality and erp stratified by age
Age is a known risk factor for both all-cause and cardiovascular mortality. Therefore, we performed subgroup analyses in 3 age groups (55-64, 65-74, and ≥75 years) and observed no differences in all-cause or cardiovascular mortality rates in individuals with and without ERP in each subgroup (log-rank test, all P>0.05;, 4B). Moreover, the survival rates of individuals with and without ERP in the inferior leads among the 3 age subgroups were also similar (all P>0.05;, 5B).
## Aging
## Correlation between erp and other risk factors with all-cause and cardiovascular mortality
As shown inand, multivariate Cox proportional hazards analyses after adjusting for covariables showed that ERP was not a risk factor for both all-cause and cardiovascular disease mortality (P=0.12 and 0.7, respectively). Furthermore, our study showed that age, gender, BMI, SBP, DBP, smoking, diabetes mellitus, stroke, chronic kidney disease, and QTc were risk factors for all-cause mortality (all P values <0.05) and age and stroke were risk factors for cardiovascular mortality (all P values <0.05, Supplementary. Similarly, after adjusting for covariables, multivariate analysis showed that individuals with ERP in inferior leads alone (ERP-inf + ) were not associated with increased risk of all-cause and cardiovascular mortality (both P values >0.05;. Moreover, multivariate analysis showed that patients with ERP in lateral leads alone (ERP-lat + ) and ERP in both inferior leads and lateral leads (ERP-inf + /lat + ) were associated with risk of all-cause mortality (P=0.029, 0.048, respectively;, but were not associated with the risk of cardiovascular mortality (both P values >0.05; after adjusting for covariables. Although patients with ERP-inf + /lat + were associated with increased risk of all-cause mortality, the data was not sufficient to make a strong conclusion because only 5 deaths were reported in this group during the 10-year follow-up.
Since age is a strong predictor of both all-cause and cardiovascular mortality, we further compared mortality rates by stratifying ERP (+) and ERP (-) individuals into 3 age subgroups (55-64, 65-74, and ≥75 years). Multivariate Cox proportional hazard analyses after adjusting for covariables showed that ERP was not a risk factor for both all-cause and cardiovascular mortality in all the age subgroups (all P values >0.05; Tables 2, 3,. Moreover, study subjects with ERP in inferior leads alone (ERP-inf + ), lateral leads alone (ERP-lat + ), or both inferior and lateral leads (ERP-inf + /lat + ) were not associated with increased risk of both all-cause and cardiovascular mortality in all age subgroups (all P values >0.05;ERP: early repolarization pattern by the criteria adopted in 2015; BMI: body mass index; BP: blood pressure; CKD: chronic kidney disease; *Values are presented as n (%) or mean ± standard deviation; QTc was calculated by Bazett's equations.
## Comparisons of the prevalence and clinical outcomes of erp in community-based studies worldwide
As shown in, the prevalence of ERP ranged from ~1% to 25% in Caucasians and 3.5%-24% in Japanese according to the community based population studies worldwide. However, the association between all-cause and cardiovascular mortality and EPR was not consistent among the published studies. Most studies showed that ERP was not associated with all-cause mortality but was associated with cardiovascular mortality. Our study cohort included older middle-aged and geriatric individuals ≥55 years with an average age of 69.1±8.2 years (range: 55-103 years), which was the highest age average of a study cohort when compared with previous reports. Moreover, the prevalence of ERP was 19.26% in our study cohort. This was within the prevalence range reported by other studies (0.99-24.79%), but was higher than the mean worldwide prevalence of 6.42%. Overall, our study shows that ERP is not associated with increased risk of all-cause and cardiovascular mortality in older middle aged and geriatric Han-Chinese population in Taiwan.
# Discussion
The prevalence of cardiovascular disease is expected to increase because of a constant rise in the proportion of older individuals worldwide. ERP is not a rare event and has been reported in studies related to several ethnicities, and has been associated with sudden cardiac arrest. To the best of our knowledge, this is the first study to examine the prevalence and prognostic value of ERP in a large cohort of older middle-aged and elderly individuals with a mean age ≥65 years.
Previous studies show that prevalence of ERP varies from ~1% to 25% in a general population. This wide range of ERP prevalence reflects differences in age range, proportions of male subjects, and inconsistent definition of ERP between studies. In the MONICA/KORA study that included 6,213 participants between 35-74 years, ERP prevalence of individuals above 55 years was 6.29%. In the Japanese NIPPON DATA90 study with 7,630 participants aged 30-95 years, ERP prevalence of individuals above 60 years was 2.5%. The number of participants in the older middle-aged and elderly population (aged 55-103 years) was highest in our study compared to previous studies. The prevalence of ERP in this age group was 19.2%-the highest overall ERP prevalence among all community-based studies that included study subjects with a mean age > 55 years. This may suggest that aging causes degeneration of the cardiac conduction pathwaysor increased AGING fibrosis and fat deposition within the heart of elderly patients.
In the past years, the association between ERP and mortality has been tested in general populations by several epidemiologic studies. The results are inconsistent and conflicting. The MONICA/KORA study screened 6,213 individuals aged 35-74 years and showed increased all-cause and cardiovascular mortality in individuals with ERP. The CARDIA study enrolled 5,039 biracial young adults aged 18-30 years with a follow-up time >20 years and demonstrated that ERP was not associated with allcause and cardiovascular mortality. The CHD study consisting of 10,864 individuals aged 30-59 years concluded that ERP-positive in the inferior leads was associated with increased risk of cardiac mortality, but was not related to all-cause mortality. A metaanalysis of 16 studies including 4 case-control and 12 prospective or retrospective studies (334,524 individuals from multiple ethnicities) showed that ERP was an electrocardiographic risk marker for deaths related to arrhythmia, cardiac diseases, and all-causes. However, there was considerable heterogeneity because of differences in study designs, ethnicity and potential misdiagnosis of ERP. Moreover, majority of these duties did not analyze older middleaged and geriatric individuals. In the Japanese NIPPON DATA90 study of 2,433 individuals older than 60 years, subgroup analysis showed that J-point elevation was not associated with cardiovascular death or death from coronary artery disease. Our random-sampling AGING community-based cohort study specifically enrolled older middle-aged and elderly individuals from the Han Chinese population, and prospectively followed up these individuals for 11years with less than 1% dropouts. Our analysis showed that ERP was not associated with both all-cause and cardiovascular mortality.
ECG is globally used as an inexpensive and noninvasive technique to detect electric abnormalities of the heart. Several individuals receive annual health examinations including ECG. ERP is an incidental and common finding on an ECG during a routine health examination.
Our study provides an important reference for clinicians or health care providers when they encounter asymptomatic elderly individuals with an incidental ERP and without a family history of SCD in members younger than 40 years.
The results of the association between ERP and mortality in several epidemiologic studies are inconsistent and conflicting. The possible reasons may be due to heterogeneity in observational studies, differences in study designs, as well as differences in age and ethnicity of the study subjects. Therefore, prospective long-term randomized clinical trials (RCTs) in different age groups year or 60-70 year-olds) in specific ethnic populations are required in the future to minimize all possible confounding factors. Moreover, combining ERP with other ECG risk parameters (such as QTc interval) may provide better risk stratification of large community-based cohorts compared to ERP as a single ECG risk factor.
There are several limitations in our study. Firstly, our findings may not apply to younger individuals because our study consisted of individuals above the age of 55 years. Secondly, the HALST cohort was restricted to individuals of Han Chinese ethnicity, and may not be applicable to other ethnicities. Thirdly, detailed clinical information including echocardiographic assessments, coronary angiography, and medications were not available in the HALST database because this data was obtained during screening.
In summary, our study shows that prevalence of ERP in a standard 12-lead ECG is common in relatively healthy, community-dwelling, ambulatory individuals above 55 years. Moreover, ERP alone is not associated with all-cause and cardiovascular mortality.
# Materials and methods
## Study subjects and inclusion criteria
Majority (>95%) of Taiwanese are of Han Chinese ancestry and nearly 2% are of aboriginal Austronesian ancestry. In this study, we initially included 5,380 Han Chinese individuals of the HALST study cohort and excluded all aboriginal Taiwanese subjects. This study was approved by the Ethics Committee of the National Health Research Institutes and conducted according to the principles of the Declaration of Helsinki. We obtained signed informed consent from the study subjects. The study cohort represented a random sample of the entire national population and was selected based on citizen IDs from 7 communities in the Northern, Central, Southern, and Eastern regions of Taiwan. The eligible and willing participants were enrolled from December 2008 to March 2013. At the time of enrollment, we prospectively collected three serial 12-lead ECGs and other relevant clinical and demographic information from the study subjects.
We then excluded individuals with cancer, significant heart diseases such as myocardial infarction and severe valvular diseases, ventricular conduction delay (left or right bundle branch block or QRS>120 ms), atrial fibrillation or flutter, QTc interval corrected for heart rate with Bazett's formula (QTc) of less than 340 msec (short QT interval) or more than 440 msec (long QT interval) at baseline and before arrhythmia, Brugada type 1 ECG, catecholaminergic arrhythmias, ventricular pre-excitation, and implanted pacemakers. The final study cohort consisted of 4,615 individuals, which was followed up for 95.1±21.9 months. During follow-up, 770 study subjects died of cardiac (n=87) or other causes (n=683).
All participants were prospectively followed on a regular basis according to the HALST study framework guidelines and the follow-up information was available until April 2019. We obtained death certificates from the National Taiwan Ministry of Health and Welfare and evaluated the cause using the 10th revision of the International Classification of Diseases (ICD-10). The death from cardiac causes was defined by ICD codes, I01-I02.0, I05-I09, I20-I25, I27, and I30-I52.
## Ecg recording and definition of erp
The 12-lead ECGs (Hewlett Packard, USA) were recorded using standard settings of 10 mm/mV and 25 mm/s. PR and QRS were computed automatically, whereas, QTc was computed using the Bazett's formula. ERP was defined as a J-point (QRS-ST junction) elevation of at least ≥0.1 mV from baseline in ≥2 adjacent leads with either QRS slurring or notching morphology, as described by Haissaguerre et al. and the 2015 consensus criteria. The anterior precordial leads (V1 to V3) were excluded from the analysis to avoid inclusion of patients with right ventricular arrhythmogenic dysplasia or the Brugada syndrome.
ECGs were displayed on a 24-inch computer screen in multiple formats to enable careful classification of slurring on the downslope of the R and J waves. All ECGs were independently analyzed and interpreted in a random order by two trained cardiologists who were blinded to clinical data and the follow-up status. In case of divergent results, a third blinded cardiologist reinterpreted the ECG and a preliminary ERP status was achieved by a majority vote. After the preliminary decision on ERP status, two trained cardiologists jointly reassessed all the ECGs that were considered ERPpositive and reached a final consensus decision on the ERP status.
## Long-term prognosis and follow-up
We collected clinical data regarding history of unexplained syncope, circumstances of sudden cardiac arrest, and a family history of unexplained sudden death (at <40 years of age).
The primary end point was death from cardiac causes, and the secondary end point was death from any cause before April 2019. An annual follow-up telephone interview of all the study participants was carried out by the study nurses to determine any new cardiovascular or frailty-related events after the initial enrollment. The cause of deaths was determined by examining the death certificates from the National Taiwan Institute of Health and Welfare and reviewed by 2 clinicians blinded to the ECG results. The cause of sudden death from arrhythmia was adjudicated by a committee of experienced cardiologists who were blinded to the data from the ECG analyses.
# Statistical analysis
We used the Fisher exact test to compare categorical variables and Student's t test to compare continuous data between the groups. Linear regression was used for continuous variables and logistic regression was used for dichotomous variables to determine the relationship between ERP and mortality. Multivariate analysis was performed to determine hazard ratios (HRs) and confidence intervals (CIs). The models were primarily adjusted for age and sex, and further adjusted for covariates that were selected on the basis of previous evidence of an association with death from AGING cardiovascular causes or other causes. This included continuous variables such as BMI, SBP, DBP, and QTc, and categorical variables such as smoking status (current, quit, never), hypertension, diabetes mellitus, hyperlipidemia, stroke, and chronic kidney disease. The relationship between ERP and mortality was calculated using a weighted Cox proportional hazards model. Kaplan-Meier survival curves and log rank tests were used to determine survival times of different groups of individuals. Clinical factors such as gender, age, smoking, SBP, DBP, BMI, stroke, diabetes mellitus, hypertension, hyperlipidemia, and chronic kidney disease were used for multivariable analysis. P < 0.05 was considered statistically significant. Inter-observer agreement was based on the overall proportion of agreements and the Kappa statistic across all ECGs. Statistical analyses were performed using the R software, version 4.0.
|
Leaving academia: PhD attrition and unhealthy research environments
This study investigates PhD candidates' (N = 391) perceptions about their research environment at a Dutch university in terms of the research climate, (un)ethical supervisory practices, and questionable research practices. We assessed whether their perceptions are related to career considerations. We gathered quantitative self-report estimations of the perceptions of PhD candidates using an online survey tool and then conducted descriptive and within-subject correlation analysis of the results. While most PhD candidates experience fair evaluation processes, openness, integrity, trust, and freedom in their research climate, many report lack of time and support, insufficient supervision, and witness questionable research practices. Results based on Spearman correlations indicate that those who experience a less healthy research environment (including experiences with unethical supervision, questionable practices, and barriers to responsible research), more often consider leaving academia and their current PhD position.
# Introduction
PhD candidates' attrition rates and time-to-degree-completion are, according to [bib_ref] The PhD track: Who succeeds, who drops out?, Groenvynck [/bib_ref] , important measurements of the efficiency and effectiveness of doctoral education. Higher attrition rates and longer times-to-degree go hand in hand, and lead to increased costs for PhD candidates and institutionsas well as a range of negative consequences for PhD candidates and their supervisors [bib_ref] Who Are the Doctoral Students Who Drop Out? Factors Associated with the..., Wollast [/bib_ref]. The average rate of attrition or prolongation of doctoral studies in Europe, Australia, and North America is estimated to be around 60% across multiple studies [bib_ref] Why did I drop out? Former students' recollections about their study process..., Leijen [/bib_ref]. Precise numbers are difficult to obtain for doctoral programs in Europe, but according to a recent survey, an aggregate of 34% of doctoral candidates who enrolled in European institutions in 2009 did not graduate within six years [bib_ref] Doctoral education in Europe today: approaches and institutional structures, Hasgall [/bib_ref]. The literature uses various terms for PhD attrition, such as "quit," "leave," "does not finish," or "drop-out." Throughout our paper, "quit" or "attrition" refer to not finishing one's PhD candidature, and "leave" refers to researchers leaving academia after gaining their PhD.
Many factors have been suggested as contributing to attrition and, more broadly, to intentions of leaving academia, including: funding, quality of supervision, scientific discipline, exposure to questionable research practices, institutional factors, organizational climate, involvement and socialization in the academic environment, community support, mental health, financial and nonfinancial costs, a lack of career prospects, and personal factors, such as life situations and attitudes towards doctoral studies [bib_ref] The PhD track: Who succeeds, who drops out?, Groenvynck [/bib_ref] [bib_ref] Who Are the Doctoral Students Who Drop Out? Factors Associated with the..., Wollast [/bib_ref] [bib_ref] Why did I drop out? Former students' recollections about their study process..., Leijen [/bib_ref] [bib_ref] Questionable Research Practices, Preregistration, and More-Exploring Self-Report Opinions of Swedish and Dutch..., Arlinghaus [/bib_ref] [bib_ref] Analysis of social media forums to elicit narratives of graduate engineering student..., Berdanier [/bib_ref] [bib_ref] Doctoral students' experiences leading to completion or attrition: a matter of sense,..., Devos [/bib_ref] [bib_ref] Beginning Graduate School: Explaining First-Year Doctoral Attrition, Golde [/bib_ref] [bib_ref] Too Many PhD Graduates or Too Few Academic Job Openings: The Basic..., Larson [/bib_ref] [bib_ref] Exploring the Fit between Doctoral Students' and Supervisors' Perceptions of Resources and..., Pyhä Ltö [/bib_ref] [bib_ref] What Took Them So Long? Explaining PhD Delays among Doctoral Candidates, Van De Schoot [/bib_ref] [bib_ref] The PhD Experience: A Review of the Factors Influencing Doctoral Students' Completion,..., Sverdlik [/bib_ref] [bib_ref] Factors that influence PhD candidates' success: the importance of PhD project characteristics, Van Rooij [/bib_ref]. Most studies on PhD candidates who quit academia assume that this is undesirable and call for preventive measures; an assumption reflected by the value-laden term "attrition". Some literature questions this assumption, and emphasizes that there may be good reasons at early career stages to quit or leave academia, such as the (increasingly) uncertain longer-term prospects and (decreasing) availability of job resources [see e.g., [bib_ref] Unequal expressions: emotions and narratives of leaving and remaining in precarious academia, Mckenzie [/bib_ref]. Although we acknowledge that low attrition is not necessarily desirable, our study focuses on the impact of factors connected to PhD students' research climate and their exposure to questionable research practices and questionable professional conduct, including poor supervision. Lowering attrition due to these factors is part of a university's duties of care. Funding is often cited as one of the most robust predictors of doctoral completion [bib_ref] The PhD track: Who succeeds, who drops out?, Groenvynck [/bib_ref] [bib_ref] Factors Influencing Successful Submission of PhD Theses, Wright [/bib_ref]. However, even in countries such as the Netherlands where most PhD candidates are considered employees and earn a salary, only a minority complete their PhD in the usually appointed four years and the average completion time is five years [bib_ref] What Took Them So Long? Explaining PhD Delays among Doctoral Candidates, Van De Schoot [/bib_ref]. In the absence of funding-related concerns, van Rooij and colleagues [bib_ref] Factors that influence PhD candidates' success: the importance of PhD project characteristics, Van Rooij [/bib_ref] identified as a key determinant of candidates' intention to quit their PhD, the research climate. In their paper, research climate is related to such things as experienced workload, the quality of the (academic and personal) relationship with the supervisor, a sense of belonging, and the amount of freedom PhD candidates are granted to carry out their own project. Other studies also show the effects of supervisory behavior on PhD candidates' attrition rates [bib_ref] Analysis of social media forums to elicit narratives of graduate engineering student..., Berdanier [/bib_ref] [bib_ref] Beginning Graduate School: Explaining First-Year Doctoral Attrition, Golde [/bib_ref] [bib_ref] Exploring the Fit between Doctoral Students' and Supervisors' Perceptions of Resources and..., Pyhä Ltö [/bib_ref] [bib_ref] The PhD Experience: A Review of the Factors Influencing Doctoral Students' Completion,..., Sverdlik [/bib_ref]. Experienced workload and supervision are sometimes directly related: in a Dutch nation-wide studyconducted by the national group representing PhD candidates' interests, Promovendi Netwerk Nederland (PNN), almost half (43%) of the PhDs reported that their supervisor engaged in one of nine supervisory behaviors that they labeled questionable based on previous reports of actions that put strain on PhDs. These included downplaying the workload or not recognizing work pressure (22%), contacting PhDs during weekends or evenings (17%), and pressuring PhDs to take on additional tasks (13%). The PNN study results are published in a range of reports, leading to several different citations within our paper. To unify our references to the same study, we refer to these reports here as part of the PNN study.
Factors relating to researchers' ethical and professional conduct are discussed less in the literature. Recent studies highlight that experience with questionable research practices may have a negative impact on the career of early career researchers [e.g., [bib_ref] A survey of early-career researchers in Australia, Christian [/bib_ref] ]. Questionable research practices (QRPs) are actions considered unethical by many (but not all) researchers, yet typically not considered as misconduct. QRPs thus form a gray zone between good and bad practices and commonly involve some type of misrepresentation, inaccuracy, or bias [bib_ref] The Gray Zone: Questionable Research Practices in the Business School, Butler [/bib_ref] , often specific to context, discipline, or subjective judgements. Existing taxonomies include a wide variety of QRPs [see e.g., [bib_ref] Ranking major and minor research misbehaviors: results from a survey among participants..., Bouter [/bib_ref] [bib_ref] Towards a taxonomy of research misconduct: The case of business school research, Hall [/bib_ref] [bib_ref] Scientists behaving badly, Martinson [/bib_ref] [bib_ref] Enhancing the Taxonomies Relating to Academic Integrity and Misconduct, Tauginienė [/bib_ref]. Commonly used examples include not publishing negative results, rounding down p values, granting author status to non-contributing researchers ('gift authorship'), and insufficient supervision. In a recent study conducted with PhD candidates in the Netherlands and Sweden, Arlinghaus and Kekecs [bib_ref] Questionable Research Practices, Preregistration, and More-Exploring Self-Report Opinions of Swedish and Dutch..., Arlinghaus [/bib_ref] found that 24% of their respondents considered leaving academia because of exposure to QRPs: PhD candidates' attrition considerations increased by a factor of 1.42 with each observed QRP.
The current study aims to find additional evidence for the importance of these different factors in explaining doctoral attrition. Additionally, given that the factors may be significantly correlated, we aim to contribute to the existing literature by examining these correlations and investigating the combined effects of: (1) the research climates in which PhD candidates operate; (2) their experiences of misconduct and questionable research practices; and (3) concerns related to supervisors' behavior. In the remainder of this paper, we first present the quantitative questions we used in our online survey tool. After discussing our results in terms of descriptive statistics, we then shift our focus on how PhD candidates' experiences in their immediate research environment (i.e., the research climate where they work, questionable research practices, and supervisors' conduct) correlate with their considerations of leaving academia or quitting their PhD. Rather than estimating the true prevalence of subpar and questionable practices, we are interested in PhD candidates' subjective beliefs about the quality of their research environment. We conclude that PhD candidates who perceive more questionable practices and insufficient supervision feel their research environment is poor, and the poorer their perceptions about their research environment, the more often candidates consider leaving. Since our study concerns one Dutch university, it has both the drawbacks and the benefits of a case study; the generalizability of our results is limited, but the uniqueness of the Dutch doctoral education system and the relative homogeneity of the macro-level environment (nation-and university-wide policies) aid the detection of meso-level differences.
## Context
The population studied are PhD candidates working at Eindhoven University of Technology (TU/e) in the Netherlands. They primarily engage in STEM (science, technology, engineering, and mathematics) research. STEM researchers constitute an understudied population [bib_ref] Analysis of social media forums to elicit narratives of graduate engineering student..., Berdanier [/bib_ref] that might differ substantially from non-STEM populations, as natural and laboratory sciences appear to have (much) lower attrition rates than, for example, the social sciences and humanities [bib_ref] Who Are the Doctoral Students Who Drop Out? Factors Associated with the..., Wollast [/bib_ref].
PhD candidates in the Netherlands are generally employees and earn a salary, which means "funding" is much less of a concern for them. Studies show that PhD candidates who have more stable financial resources are more likely to complete their training . In light of this, our study design allows us to focus on what we are most interested in, namely PhD candidates' experiences with their research environments. While we acknowledge limitations to generalizability (discussed in Section 5.1 below), our data provides an important starting point to examine correlations between factors that predict doctoral attrition among STEM PhD students.
According to the PNN study, despite most PhD candidates in the Netherlands having employee status, there are alternative legal arrangements such as scholarships or individually funded external positions. The typical duration of PhD contracts is 4 years. Contract hours range from 32 to 40 hours a week; scholarship and external PhD candidates are not usually bound to a formal number of working hours. The PNN study reported that PhD candidates' main obligations were research, taking courses, and teaching. To complete their PhD, candidates are expected to produce a dissertation either in the form of a monograph or a collection of articles. Policies and exact guidelines on what is accepted for a dissertation differ depending on the university, research group or department, and discipline.
At the time of our study, the TU/e website listed 1,608 PhD candidates. As a comparison, in 2018, around 5,000 candidates completed their PhD in the Netherlands. Given the university's disciplinary focus and existing support systems, along with the fact that most PhD candidates are funded, we did not expect to see high attrition rates.
# Methods
## Recruitment and participation
The planned methods, design, and analytical steps were registered in the study proposal and ethical form approved by the TU/e's Ethical Review Board (see the project's OSF repository for all supplementary materials as well as an explanation of all deviations from the steps discussed in the original proposal: https://osf.io/bqx7v/). We hoped to recruit at least 200 participants based on feasibility considerations of the entire population, and the response rates of similar, previous studies; these rates were 3% in a university-wide Open Science Community Eindhoven and Information Expertise Center survey, and 21% in a nation-wide survey. We wanted to describe the opinions of PhD students as the percentage of individuals who agree or disagree with statements, and estimate correlations between variables. After collecting 200 participants we would be able to describe frequencies with a margin of error of at most 7% (i.e., if 20% of participants would choose an answer option, the margin of error would range from 13.5% to 27.4%), and correlations with a margin of error of at most 0.14 (i.e., for a correlation of 0). Given the achieved sample size of 391, these maximum margin or errors decreased to 5% (for frequencies) and 0.1 (for correlations), which determine the granularity at which readers can interpret descriptive patterns in the data. For many descriptive statistics our interest is in how often people agree or disagree (regardless of whether they do so somewhat or strongly), or if questionable practices are observed at all, in which case accuracy in one response category is less important. Readers are advised to take the remaining uncertainty into account, especially when using our data for policy decisions.
All 1,608 PhD candidates listed as working at TU/e were invited and later reminded to participate via their official university email address (see OSF repository). To increase response rates and reduce non-response bias, email communication was designed based on best practice recommendations by Dillman and colleagues, and ten EUR 50 vouchers were raffled among the participants. Data collection took place from December 2020 to January 2021 via LimeSurvey (see OSF repository).
Duplicate responses, responses with data missing and only the demographic questions answered, were considered invalid and excluded from the data analysis. Sample size after removing invalid responses was 391 (32% women, less than 1% gender variant/non-conforming). The response rate was 24%, slightly higher than the 20% average in similar surveys [bib_ref] A Comparison of Web and Mail Survey Response Rates, Kaplowitz [/bib_ref]. Respondents had a mean age of 28.8 years, 95% confidence interval ; 62% were 25 to 29 years old; 48% of the participants were Dutch, 33% indicated belonging to an ethnic minority, and 24% to a racial minority. The sample is similar to the TU/e doctoral population in terms of age (59% between 25 and 29 years old), nationality (41% Dutch), and gender (33% women). Similarity to racial and ethnic minority status could not be established due to lack of institutional data.
## Survey design
After giving informed consent, participants were asked to complete five blocks of survey questions in the order as presented in the OSF repository. Unless stated otherwise, perceptions about frequency (ranging from 'never' to 'very often' for career considerations, and 'never' to 'almost always' for QRPs) and extent of agreement ('strongly disagree' to 'strongly agree') could be expressed on a 7-point Likert scale (for a complete list of questions, items, and scales, see OSF repository).
Demographic and general data. Participants were asked to report their age, gender, country of origin, nationality (Dutch/non-Dutch), the year they started their PhD and the year they expected to complete it, research area, and ethnic and racial minority status.
Responsible research climate.
To measure to what extent participants experienced a responsible research climate, we included six facilitators and four barriers [bib_ref] Perceptions of a Responsible Research Climate: A Multi Focus Group Study, Haven [/bib_ref]. Many studies highlight the role of macro-level (e.g., structural problems within academia) and meso-level (e.g., organizational culture and climate) influences on unethical behaviors and research quality [bib_ref] The Perverse Effects of Competition on Scientists' Work and Relationships, Anderson [/bib_ref] [bib_ref] A Decade of Empirical Research on Research Integrity: What Have We (Not)..., Bonn [/bib_ref] [bib_ref] Feline followers and "umbrella carriers"_ Department Chairs' influence on faculty job satisfaction..., Bä Ker [/bib_ref] [bib_ref] Commentary: Perverse Incentives or Rotten Apples?, Bouter [/bib_ref] [bib_ref] Relationships Between the Survey of Organizational Research Climate (SORC) and Self-Reported Research..., Crain [/bib_ref] [bib_ref] Explaining variance in perceived research misbehavior: results from a survey among academic..., Haven [/bib_ref] [bib_ref] Environmental Influences on Ethical Decision Making: Climate and Environmental Predictors of Research..., Mumford [/bib_ref] [bib_ref] Exploring Scientific Misconduct: Isolated Individuals, Impure Institutions, or an Inevitable Idiom of..., Sovacool [/bib_ref] [bib_ref] Expanding Research Integrity: A Cultural-Practice Perspective, Valkenburg [/bib_ref]. Haven and colleagues [bib_ref] Perceptions of a Responsible Research Climate: A Multi Focus Group Study, Haven [/bib_ref] distinguish "the shared meaning researchers attach to the policies, practices and behaviors they associate with a responsible research climate" from general perceptions about the organization. By asking 61 researchers in focus-group interviews to reflect on the facilitators of and barriers to responsible research climates, the authors identified six facilitators (fair evaluation, openness, sufficient time, integrity, trust, and freedom) and four barriers (lack of support, unfair evaluation policies, normalization of overwork, and insufficient supervision). Haven and colleagues [bib_ref] Perceptions of a Responsible Research Climate: A Multi Focus Group Study, Haven [/bib_ref] use the more general term "characteristics," which we decided to change to "facilitators" to simplify our terminology and better indicate the opposing nature of barriers and characteristics of responsible research climates.
We considered these ten items the most relevant for our study for two reasons: First, they were derived from interviews conducted on a sample of researchers working in the Netherlands, which increases the likelihood that our participants would recognize the concepts. Second, the items are less connected to macro-level systemic issues and organizational structures that PhD candidates might have seldom or not directly experienced. To implement these characteristics, participants were asked to what extent they experienced the facilitator or barrier within their research environments.
Questionable research practices. Participants were asked to recall encounters and perceptions on the prevalence of misconduct and QRPs in their disciplinary fields, as well as in their current work environments (research group or institution). Scientific misconduct refers to any (attempted) action that undermines academic integrity, most typically fabrication, falsification, or plagiarism. To measure QRPs, we combined the ten highest-ranking items according to frequency and impact on trust, and the ten highest-ranking ones according to frequency and impact on validity from a sixty-item QRP taxonomy [bib_ref] Ranking major and minor research misbehaviors: results from a survey among participants..., Bouter [/bib_ref]. Removing duplicates resulted in a list of 14 items, such as "Not publishing a valid 'negative' study," "Ignoring basic principles of quality assurance," or "Not reporting clearly relevant details of study methods." To these, we added three other items: "Fabricate or falsify data," "Plagiarize" (according to the most common characterization of research misconduct), and "Other behaviors that I perceive as questionable research practices."
Supervision. To examine how doctoral attrition relates to supervisory practices, participants were asked about their supervisors' conduct. In terms of fostering research integrity, supervisors play a role as ethical examples [bib_ref] What Are Ethics in Doctoral Supervision, and How Do They Matter? Doctoral..., Lö Fström [/bib_ref] , and are important for creating a responsible research climate [bib_ref] Perceptions of a Responsible Research Climate: A Multi Focus Group Study, Haven [/bib_ref]. According to professional guidelines, supervisors are expected to facilitate responsible research by providing competent supervision and mentoring. Some supervisory actions are deemed unethical [bib_ref] Ethical Issues in Doctoral Supervision: The Perspectives of PhD Students in the..., Lö Fström [/bib_ref] or have detrimental consequences in terms of research integrity considerations [bib_ref] Ranking major and minor research misbehaviors: results from a survey among participants..., Bouter [/bib_ref] [bib_ref] Perceptions of a Responsible Research Climate: A Multi Focus Group Study, Haven [/bib_ref]. Based on a series of qualitative studies, Löfström and Pyhältö [bib_ref] What Are Ethics in Doctoral Supervision, and How Do They Matter? Doctoral..., Lö Fström [/bib_ref] listed five ethical principles underlying supervisory practices: respect of autonomy, non-maleficence, beneficence, justice, and fidelity. From these ethical principles, they constructed a 16-item Ethical Issues in Supervision Scale with subcategories such as exploitation, intrusion of supervisor views and values, inadequate supervision, inequality, and unfair authorship. Failure to adhere to these ethical principles was related to poor outcomes of the doctoral process in terms of engagement, satisfaction with supervision and doctoral candidacy, burnout, and attrition intentions.
The questions in our study were developed based on these Ethical Issues in Supervision Scales [bib_ref] What Are Ethics in Doctoral Supervision, and How Do They Matter? Doctoral..., Lö Fström [/bib_ref]. Participants first answered a set of questions on "How would you characterize your experience with your supervisor? If you have multiple supervisors, think about your primary supervisor, or the person you are most in contact with regarding your supervision. Please indicate the extent to which you agree with the following statements." Eighteen items such as "I receive supervision when I need it," and "My supervisor lacks cultural competency / multicultural sensitivity," were rated in terms of extent of agreement. Participants could agree or disagree to items attached to the second set of questions based on more serious ethical violations listed by January and colleagues [bib_ref] Impressions of misconduct: Graduate students' perception of faculty ethical violations in scientist-practitioner..., January [/bib_ref] , and we asked "Do any of your supervisors engage in the following actions?" Six items such as "Have intimate relationships with a doctoral student," "Harassment (including unwanted sexual coercion)," and "Use slurs or make jokes about racial or ethnic groups" were presented. Items were rated "No," "Yes," or "Don't know / Prefer not to disclose." Career, attrition, and leave considerations. Participants were asked about their career commitment and considerations of leaving academia or quitting their current job. To tap into different aspects of career considerations, we presented two sets of questions. First, using questions on career commitment developed by Blau [bib_ref] Testing the Generalizability of a Career Commitment Measure and Its Impact on..., Blau [/bib_ref] , we asked participants to rate whether they agreed to seven statements such as "I definitely want a career for myself in academia" or "I am disappointed that I ever entered academia." Second, participants were asked: "How often have you seriously considered quitting academia / your current job?" based on questions previously presented by Spector and colleagues [bib_ref] Relation of Job Stressors to Affective, Health, and Performance Outcomes: A Comparisonof..., Spector [/bib_ref].
# Results
We present descriptive results for all relevant survey questions in Figures and provide Tables with all relevant correlations between survey questions. When interpreting the data, we discuss general patterns and highlight the most important variables (either in terms of frequencies, or in terms of the size of the correlation). We did not test any hypotheses. Given the sample size of N = 391 PhD students, correlations above 0.18, 0.21, 0.22, and 0.24 are statistically significant at the 0.05, 0.01, 0.005, and 0.001 level in a two-sided test, respectively. For the reader's convenience, we flag correlations significant at alpha levels of 0.05 and 0.001 in all correlation tables, so as to transparently communicate that there is a low probability we mistake random variation as a true effect.
## Missing data
The overall percentage of missing data, including "don't know," "doesn't apply," and "prefer not to say" answers, was 15%. Missing data pertained especially to supervision-related questions about collaboration and favoritism, and to QRP for participants who recently became PhD candidates. This might reflect the lack of experiences with collaboration and QRPs among this sub-population (both in a disciplinary and work environment context). We provide a more in-depth analysis of missing data in our OSF repository.
## Responsible research climate
Most participants had positive impressions about their research climate. Almost all agreed that they experienced fair evaluation, integrity, freedom, trust, and openness. Most agreed that they experienced sufficient time for work [fig_ref] Fig 1: Facilitators of a responsible research climate experienced in the scientific environment [/fig_ref]. Most seemed satisfied with the fairness of evaluation policies, support, and supervision, but almost half reported overwork being normalized within their research environment [fig_ref] Fig 2: Barriers to a responsible research climate experienced in the scientific environment [/fig_ref].
Both on the facilitator and barrier side, time constraints emerge as the most serious concern. One out of four respondents (25%) disagreed with having sufficient time for work; and almost 45% agreed with having experienced normalization of overwork within their scientific environment. Results of a Spearman correlation indicated a significant negative association between having sufficient time (facilitator) and normalization of overwork (barrier), (rS = -.47, p < .001). One out of five respondents agreed with experiencing a lack of support (18%) and/or insufficient supervision (19%), and these two items were strongly positively correlated (rS = .69, p < .001).
In terms of correlations between items, high intercorrelations can be perceived within both the set of 6 facilitators (α = .81) and the set of 4 barriers (α = .73). Furthermore, facilitators and barriers were negatively correlated with each other: those experiencing barriers to a larger extent reported experiencing facilitators to a lesser extent [fig_ref] Table 1: Correlations among research climate items [/fig_ref].
Research climate items were also correlated with most items measuring supervision and QRPs within the work environment. Perceptions about almost all QRPs within the work environment showed weak to moderate negative correlations with almost all facilitators, and weak to moderate positive correlations with almost all barriers (see OSF repository).
## Questionable research practices
Respondents indicated that QRPs [fig_ref] Fig 3: Estimated Prevalence of Questionable Research Practices, in the discipline [/fig_ref] were more prevalent within their disciplinary field than within their current work environment. The majority of PhD candidates (N discipline = 74%, N work = 63%) estimated that at least one QRP occurs at least very rarely within their discipline and their work environment. The three most frequently reported QRPs were stated by 54% of the participants (insufficiently reporting flaws or limitations, not reporting methods, The top three QRPs in their working environment reported by 39% and 38% of participants, were keeping inadequate notes, insufficient supervision, and insufficiently reporting flaws or limitations). In addition, publication bias (i.e., not publishing a valid 'negative' study) was the main practice reported occurring at least often, both within the discipline and the work environment.
About 20% of PhD candidates estimated that misconduct is at least a very rare occurrence within their discipline, while 2% estimated that misconduct occurs often, very often, or almost always within their discipline. About 10% reported occurrences of plagiarism or fabrication / falsification within their work environment.
## Supervisory practices
Almost all respondents experienced receiving criticism in a friendly manner (90%), that they could tell supervisors if personal matters were affecting their work (86%), and that they could negotiate central choices for their dissertation (87%), see A total of 32 respondents (8%) reported experiencing that their supervisors engaged in at least one serious transgression: racism (4%), sexism (3%), slurs/jokes about other minorities (3%), intimate relationships with PhD candidates (2%), harassment (less than 1%), and homophobia (less than 1%). Respondents reported a total of 53 such serious transgressions, with 19 PhD candidates reporting only one. Note that it is possible multiple PhD students independently reported observing the same transgression.
## Career considerations
The majority of PhD candidates did not express strong negative feelings about academia as a vocation. Responses to questions about future career paths were more mixed. For example, 50% reported they loved doing research in academia too much to give it up and 8% were disappointed that they had entered academia [fig_ref] Fig 5: Vocational commitment [/fig_ref].
Respectively, 10% and 18% of PhD candidates considered quitting their current job and leaving academia at least often . Not surprisingly, respondents who started working at the university more recently, had less frequently considered leaving their job (r S = -.25, p < . Numbers in red are rate of respondents reporting occurrences "very rarely". .001) as well as academia (r S = -.34, p < .001). Results of a Spearman correlation indicated a significant positive correlation between considerations of quitting the current job and of leaving academia (r S = .57, p < .001). This correlation is high, and significant, but it is not perfect: in some cases, PhD students wanted to leave academia but not their current job as PhD students.
## Demographic subgroup analyses
Experiences with research climates did not differ significantly based on ethnic or racial minority status and research area. There were some gender differences: female respondents (50%) were more likely to report normalization of overwork than male respondents (35%). Most of the statistically significant differences were related to nationality [fig_ref] Fig 7: Fig 7 [/fig_ref].
Similar to experiences with research climates, perceptions about supervisory practices varied between Dutch and non-Dutch PhD candidates .
Non-Dutch PhD candidates were more likely than Dutch PhD candidates to report: definitely wanting a career in academia (40% vs 24%), love doing research in academia too much to give it up (45% vs 30%), and feeling academia was the ideal vocation (11% vs 3%).
## Correlations between research climates, qrps, and supervisory practices
Perceptions about nearly all QRPs within the work environment correlated negatively with almost all facilitators, and positively with almost all barriers of responsible research climates (i.e., correlations were significant between almost all items). All significant correlations were weak (<0.4) to moderate (0.4 to 0.6) [fig_ref] Table 2: Spearman correlations for research climate items and questionable research practices within the... [/fig_ref] according to the naming convention most commonly used in psychology [bib_ref] User's guide to correlation coefficients, Akoglu [/bib_ref]. Integrity and trust (research-climate characteristics) were related to all questionable practice items. This substantiates the idea that, although QRPs might form an ethical gray zone, they are perceived to be connected to a lack of research integrity. In addition, all research climate items were significantly correlated to the only supervision-related questionable practice (i.e., insufficient supervision). Most supervision-practice items were also related to perceptions about research climates [fig_ref] Table 3: Spearman correlations for research climate items and supervisory practices [/fig_ref]. Rate of Dutch and non-Dutch PhD candidates who report experiencing sufficient time for work, unfair evaluation policies, or normalization of overwork. Note: Scales range either from strongly disagree to strongly agree (marked "disagree") or the opposite, strongly agree to strongly disagree (marked "agree"). Numbers in red represent the rate of respondents who at least "somewhat agree/disagree". https://doi.org/10.1371/journal.pone.0274976.g007
## Correlations with leave and quit considerations
In terms of responsible research climates, both leave and quit considerations correlated negatively with facilitators (e.g., trust or freedom), and positively with barriers (lack of support or insufficient supervision), see [fig_ref] Table 5: Spearman correlations for research climate items and leave/quit considerations [/fig_ref]. Most supervisory practices were also related to leave and quit considerations. For example, experiences with actions like being left without supervision or supervisors' inadequate preparedness were positively correlated with leave considerations , while fair treatment of all PhD candidates and the ability to negotiate central choices for the dissertation were negatively correlated with leave and quit considerations [fig_ref] Table 7: Spearman correlations for supervision items and leave/quit considerations [/fig_ref]. These correlations, even when significant, are weak. This indicates that, even though the correlations are related, other factors might influence the decision to quit or leave, or possibly a combination of factors. This analysis, however, falls outside the scope of this paper, and could be addressed in future research.
## Fig 8. dutch and non-dutch phd candidates' characterization of their primary supervisor's practices.
Note: Scales range either from strongly disagree to strongly agree (marked "disagree") or the opposite, strongly agree to strongly disagree (marked "agree"). Numbers in red represent the rate of respondents who at least "somewhat agree/disagree". https://doi.org/10.1371/journal.pone.0274976.g008
## Plos one
Most QRPs in the discipline were not significantly correlated with leave or quit considerations, while most QRPs within the work environment were weakly positively correlated .
# Discussion and conclusion
Training PhD candidates is an investment that requires funding, specific educational programs organized by the university, as well as supervisors' time and commitment. Negative To these, we added insights in the correlations between such factors [fig_ref] Table 2: Spearman correlations for research climate items and questionable research practices within the... [/fig_ref] and their correlations with doctoral attrition (Sections 4.7 and 4.8; [fig_ref] Table 5: Spearman correlations for research climate items and leave/quit considerations [/fig_ref]. In all the factors studied, we found some correlations of moderate strength (here defined as Spearman correlation > .4); correlations of factors with doctoral attrition were weaker (although with several items >.. This strengthens intuitions on the combined effects of factors studied more qualitatively [bib_ref] Perceptions of a Responsible Research Climate: A Multi Focus Group Study, Haven [/bib_ref] , and with small samples [bib_ref] Questionable Research Practices, Preregistration, and More-Exploring Self-Report Opinions of Swedish and Dutch..., Arlinghaus [/bib_ref]. Overall, many of our constructs are moderately correlated. Although such correlations should be interpreted with caution, establishing them by combining constructs linked to professional and ethical conduct is a substantive contribution to the literature on doctoral attrition.
In terms of facilitators and barriers for responsible research climates, the majority of PhD candidates report experiencing fair evaluation, integrity, freedom, trust, and openness. Insufficient supervision and having insufficient time are frequently reported barriers to responsible research climates, barriers for which we found positive correlations with leave and quit considerations. These results underline findings from previous research, that time constraints and supervisory practices are important influencers of doctoral satisfaction and attrition rates in the Netherlands [bib_ref] Factors that influence PhD candidates' success: the importance of PhD project characteristics, Van Rooij [/bib_ref]. Lack of time is often cited as being detrimental to research integrity, the quality of research, education, mentorship, and a healthy work-life balance [bib_ref] A Decade of Empirical Research on Research Integrity: What Have We (Not)..., Bonn [/bib_ref]. We found intercorrelations within both the set of six facilitators (α = .81) and the set of four barriers (α = .73), indicating that they tend to co-occur. We also found significant correlations between perceptions of research climate, supervisory practices, and QRPs. This correlative relationship with variables related to responsible practices confirms the validity of the responsible research climates' construct by Haven and colleagues [bib_ref] Perceptions of a Responsible Research Climate: A Multi Focus Group Study, Haven [/bib_ref]. Respondents estimate that questionable research practices are more prevalent within their disciplinary field than within their current work environment. This is in line with earlier findings, where researchers report more wrongdoings when asked about (distant) others than themselves or their close environments as described in Fanelli's [bib_ref] How Many Scientists Fabricate and Falsify Research? A Systematic Review and Meta-Analysis..., Fanelli [/bib_ref] meta-analysis. Still, a significant ratio of PhD candidates report experiencing questionable practices and even misconduct within their immediate surroundings. These outcomes are not directly related to actual prevalence estimates, as the reports were based on subjective perceptions and may not reflect actual misconduct or questionable research practices. Moreover, given their shared environment, several respondents might refer to the same perceived incident. Yet the reported incident rates warrant further investigation of the occurrence and effects of QRPs in more detail, since their impact could be large.
Based on our correlational outcomes, experiences with QRPs are positively related to experiences with barriers, and negatively to experiences with facilitators of responsible research climates. This confirms the idea that good things tend to co-occur (and bad things too): worse research climates also tend to have worse research practices. Indeed, trust and integrity are facilitators of responsible climates, but also fundamental for responsible research practices.
Leave and quit considerations were connected to most QRPs within work environments but not to those within the PhD candidates' disciplines. This may be an effect of distance: candidates are likely to be more upset by negative experiences closer to home than by factors experienced more indirectly. While our results are only correlative, they support the notion that experiences of QRPs are connected to considerations of leaving academia [bib_ref] Questionable Research Practices, Preregistration, and More-Exploring Self-Report Opinions of Swedish and Dutch..., Arlinghaus [/bib_ref].
For the correlations between supervisory practices and quit and leave considerations, we found that the PhD candidates who have more extensive experiences with unethical supervision practices reported considering leaving and quitting academia with a higher frequency. This corroborates earlier findings that claimed a connection between attrition and supervision-related experiences [bib_ref] Why did I drop out? Former students' recollections about their study process..., Leijen [/bib_ref] [bib_ref] What Are Ethics in Doctoral Supervision, and How Do They Matter? Doctoral..., Lö Fström [/bib_ref]. In addition, experiences with supervision display the strongest connections with perceptions about research climates-a result that is in line with the contention that insufficient supervision is an important barrier to responsible research climates [bib_ref] Perceptions of a Responsible Research Climate: A Multi Focus Group Study, Haven [/bib_ref].
Our results were obtained for a relatively homogeneous sample population, i.e., PhD candidates at a Dutch technical university. This partly addresses the lack of studies specific to STEM researchers [bib_ref] Analysis of social media forums to elicit narratives of graduate engineering student..., Berdanier [/bib_ref] , and suggests some insights related to confounding factors. Earlier work suggests that attrition rates in natural sciences are generally lower than in the social sciences and humanities [bib_ref] Who Are the Doctoral Students Who Drop Out? Factors Associated with the..., Wollast [/bib_ref]. In comparison, our study partly confirms this expectation: attrition considerations are indeed less prevalent in our sample (21%) than in the PNN study (42%; 53). However, first-and second-year candidates (overrepresented in our sample) may also have had less time to experience reasons for leaving (relating to research climate, QRPs, and supervision) than the candidates in the PNN study, who were more equally distributed over every stage of a PhD candidacy.
Moreover, 34% of the respondents in our survey indicated that they definitely want a career in academia, whereas the national survey reports a substantially higher percentage, 61.1. Some of this could be due to different formulations of the question (in our study: "I definitely want a career for myself in academia", rather than "In what sector would you like to work after obtaining your PhD?"). Still, the magnitude of the difference suggests that other factors might be at work. One explanation could be our candidates' stronger link to industry: STEM doctorate holders report the facilitating role of the connections they develop even during their PhD work, and the higher salaries and better working conditions offered by industry.
# Limitations
Our study involved mainly first-or second year PhD candidates working at a technical university in the Netherlands. This restricts the generalizability of our results. PhD candidates' perceptions and concerns may differ at various stages of their career; and their experiencesespecially in terms of exposure to supervisory and research practices-can be expected to change and diversify over time. Leave and quit considerations could be affected by the university's focus, since attrition rates are lower for students working in natural and laboratory sciences than for students in the humanities and social sciences. As for national characteristics: most PhD candidates in the Netherlands are salaried employees. Hence it is currently unclear whether our descriptive findings apply generally to other populations, especially in other countries or less STEM-focused environments. Still, our findings were in line with previous observations in different populations, and we expect that the correlations observed in our sample might hold in other populations, but perhaps with a varying effect in size.
We relied on volunteer self-reports. We tried to implement motivational tools within our invitation procedure to increase the response rates and diversify the incentives underlying participation. We increased the benefits of participation by providing financial incentives, asking peers for help and advice, and conveying that other PhD candidates had already helped on prior occasions (see OSF repository). Still, self-selection might have led to overrepresentation of some respondent characteristics. For example, PhD candidates who are dissatisfied could be more likely to respond; or many respondents may be motivated to participate in similar studies because of a general interest in topics related to research integrity.
The measurements we used also impose limitations. Unlike previous studies, we asked participants about their career considerations rather than their intentions. The two concepts seem intertwined but are somewhat different, thus making comparisons to studies using intentions or behaviors as attrition measurements difficult. In addition, as our focus was on subjective perceptions (for which considerations are relevant), comparisons between attrition rates in other studies and our data are problematic. Similarly, our results do not pertain to objective prevalence estimates of questionable research practices or supervisory actions. Even if our respondents' reports could be considered objective estimates, one bad action could have been perceived and reported by multiple respondents. In addition, although we carefully selected the QRP items included in our study, we did not ask participants about their understanding of each item, nor their estimates of severity. These constraints affect both the generalizability of our results and the comparability with similar studies.
The correlations observed in our dataset do not allow for claims about causality. On the one hand, the fact that barriers and facilitators of research climates had strong inter-correlations with questionable (supervisory) practices and attrition or leave considerations could mean that bad experiences are more likely to occur in poor research climates. The opposite is also plausible, in that bad research climates lead to a deterioration of individual behaviors. From this perspective, better practices could improve research climates, or research climates could facilitate better practices. On the other hand, these correlations could also be attributed to common method biases. A range of potential causes such as social desirability, common rater effects, or participants' biased mood could strengthen or weaken the correlations between measurements [bib_ref] Common method biases in behavioral research: A critical review of the literature..., Podsakoff [/bib_ref]. For example, good impressions of a supervisor might make participants blind to the transgressions the supervisor performs. Bad experiences might in turn curtail all the positive characteristics of the immediate environment. We recommend that future researchers examine the causality and then to what extent these experiences influence attrition considerations.
## Recommendations
An increasing number of measures is being implemented in many universities, including TU/ e, to improve the wellbeing of PhD candidates and reduce doctoral attrition. Many implications of our results are in line with recommendations in other studies and with policies that already have been partly implemented. These include: enhancing knowledge and skills related to the management of the doctoral candidacy and interactions within professional contexts [bib_ref] Factors that influence PhD candidates' success: the importance of PhD project characteristics, Van Rooij [/bib_ref] ; ensuring a good fit between PhD candidates and their supervisors, both personally and professionally [bib_ref] Operating Within" the Field: Doctoral Students' Views of Supervisors' Discipline Expertise, Gube [/bib_ref] ; and providing sufficient guidance while also leaving room for PhD candidates to pursue their own research interests [bib_ref] Promoting doctoral students' research self-efficacy: combining academic guidance with autonomy support, Overall [/bib_ref].
Several other implications emerged in conversations with university stakeholders about our results. Here, we summarize some of the recommendations we formulated during this process. Our study gives an indication of the link between PhD candidates' worst research integrityrelated experiences and their considerations of leaving academia or quitting their current job.
While connections between what PhD candidates experience in terms of integrity, ethicality, professional norms, and their career considerations have been made before, studies (and especially large sample quantitative studies) with an explicit focus on the matter at hand are scarce and fragmented. To gain further insights, we recommend that universities gather longitudinal data by conducting periodical surveys similar to the one presented in our paper. Future research using exit interviews should measure how often PhD candidates quit because of their experiences with unethicality, breaches of research integrity, or a general lack of responsible research. In addition, establishing the prevalence of wrongdoings and the number of connected wrongdoers would help to differentiate the impact of one outstanding offender and widespread issues. Further understanding of what and why PhD candidates perceive as a violation of professional codes of conduct or ethicality could also advance our understanding of how doctoral candidates are impacted by their experiences. Finally, we suggest periodical inquiries about PhD candidates' immediate research environments. If conducted and interpreted by university officials, such longitudinal data might provide valuable opportunities for interventions and insights into scarcely researched questions such as how and when attrition considerations develop.
Differences between the experiences of Dutch and non-Dutch PhD candidates can point to differences in the understanding of norms and expectations. Dutch candidates are more likely to feel the time pressures attached to employment, while not communicating norms explicitly could lead to inequalities in evaluation and expectations. Clarification of evaluation policies has a similar function and is particularly important for non-Dutch PhD candidates who already have to cope with cultural and language barriers. While we did not collect data on the characteristics of supervisors with whom PhD candidates had experiences relating to a lack of cultural sensitivity, both Dutch and non-Dutch supervisors and PhD candidates could benefit from a stronger expectation of building intercultural competencies. Here we also note that time pressure and workload are persistent and major focal points for all PhD candidates regardless of nationality. In this regard, we reiterate the recommendations by van Rooij and colleagues [bib_ref] Factors that influence PhD candidates' success: the importance of PhD project characteristics, Van Rooij [/bib_ref] : an important step in reducing workloads would be to emphasize that it is not normal or expected to work outside contract hours (weekends, holidays, or evenings) and then take further precautions and additional steps to ease pressure.
Our results establish an estimate of the subjective prevalence of certain unethical supervisory practices and corroborate notions of the importance of good supervision. For this reason, we agree with the Netherlands PhD network recommendations advocating a commitment on the side of research institutions to prevent unethical supervisory behavior by providing training, setting norms and expectations, and implementing some type of evaluation system for supervisors. With clear expectations, supervisors might have to take up new responsibilities or adjust their modus operandi in compliance with guidelines or according to PhD candidates' needs. These efforts should not go unnoticed. Universities could provide proper recognition of both daily and other supervisors and facilitate the spread of good practices by setting up peer discussion groups or semi-formal interactive forums for discussing supervisory practices.
Although research integrity training is available and mandatory for all doctoral candidates, no comparable training is required for supervisors. Even if training sheds light on the complexity of ethical decisions, PhD candidates have to (or are at least strongly incentivized to) work with the methods made available or accepted by their supervisors. Young, relatively inexperienced researchers might also have difficulty speaking up against, or even recognizing questionable research practices. Outcomes of recent studies indicate that an alarming percentage of PhD candidates report intentions to commit fraud driven by supervisors' norms [bib_ref] The Use of Questionable Research Practices to Survive in Academia Examined With..., Van De Schoot [/bib_ref]. Research integrity courses for or together with supervisors could help to open up discussions about disciplinary norms and set a basic level of understanding about (in)appropriate practices. We agree with the suggestion by Bouter and colleagues (2016): supervisors and PhD candidates might greatly benefit from a list of questionable research practices for training and supervisory purposes. Houkes.
## Supporting information
[fig] Fig 1: Facilitators of a responsible research climate experienced in the scientific environment. https://doi.org/10.1371/journal.pone.0274976.g001 [/fig]
[fig] Fig 2: Barriers to a responsible research climate experienced in the scientific environment. https://doi.org/10.1371/journal.pone.0274976.g002 and selective citing, see Fig 3, panel A) for their discipline. [/fig]
[fig] Fig 4: However, one out of five reported to have been left without supervision at one time (22%) and/or left without help if their supervisor could not advise them (22%). [/fig]
[fig] Fig 3: Estimated Prevalence of Questionable Research Practices, in the discipline (panel A) and the work environment (panel B). Note: Item descriptions were shortened for this figure. Full descriptions are in [/fig]
[fig] Fig 5: Vocational commitment. https://doi.org/10.1371/journal.pone.0274976.g005 Fig 6. Frequency of PhD candidates seriously considering quitting their current job and academia. https://doi.org/10.1371/journal.pone.0274976.g006 [/fig]
[fig] Fig 7: Fig 7. Rate of Dutch and non-Dutch PhD candidates who report experiencing sufficient time for work, unfair evaluation policies, or normalization of overwork. Note: Scales range either from strongly disagree to strongly agree (marked "disagree") or the opposite, strongly agree to strongly disagree (marked "agree"). Numbers in red represent the rate of respondents who at least "somewhat agree/disagree". [/fig]
[table] PLOS ONE: | https://doi.org/10.1371/journal.pone. [/table]
[table] Table 1: Correlations among research climate items. [/table]
[table] Table 2: Spearman correlations for research climate items and questionable research practices within the work environment. [/table]
[table] Table 3: Spearman correlations for research climate items and supervisory practices (A). [/table]
[table] Table 5: Spearman correlations for research climate items and leave/quit considerations. [/table]
[table] Table 7: Spearman correlations for supervision items and leave/quit considerations (negative correlation). [/table]
|
Generation of Corneal Epithelial Cells from Induced Pluripotent Stem Cells Derived from Human Dermal Fibroblast and Corneal Limbal Epithelium
Induced pluripotent stem (iPS) cells can be established from somatic cells. However, there is currently no established strategy to generate corneal epithelial cells from iPS cells. In this study, we investigated whether corneal epithelial cells could be differentiated from iPS cells. We tested 2 distinct sources: human adult dermal fibroblast (HDF)-derived iPS cells (253G1) and human adult corneal limbal epithelial cells (HLEC)-derived iPS cells (L1B41). We first established iPS cells from HLEC by introducing the Yamanaka 4 factors. Corneal epithelial cells were successfully induced from the iPS cells by the stromal cell-derived inducing activity (SDIA) differentiation method, as Pax6 + /K12 + corneal epithelial colonies were observed after prolonged differentiation culture (12 weeks or later) in both the L1B41 and 253G1 iPS cells following retinal pigment epithelial and lens cell induction. Interestingly, the corneal epithelial differentiation efficiency was higher in L1B41 than in 253G1. DNA methylation analysis revealed that a small proportion of differentially methylated regions still existed between L1B41 and 253G1 iPS cells even though no significant difference in methylation status was detected in the specific corneal epithelium-related genes such as K12, K3, and Pax6. The present study is the first to demonstrate a strategy for corneal epithelial cell differentiation from human iPS cells, and further suggests that the epigenomic status is associated with the propensity of iPS cells to differentiate into corneal epithelial cells.
# Introduction
Like embryonic stem cells, induced pluripotent stem (iPS) cells are capable of differentiating into all of the various cell lineages of an organism, and are established from somatic cells by introducing transcription factors such as Oct3/4, Sox2, and Klf4 [bib_ref] Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures..., Takahashi [/bib_ref] [bib_ref] Induction of pluripotent stem cells from adult human fibroblasts by defined factors, Takahashi [/bib_ref] [bib_ref] Induced pluripotent stem cell lines derived from human somatic cells, Yu [/bib_ref]. Therefore, iPS cells can be used as a cell source to regenerate tissues, such as retinal pigment epithelium (RPE), neurons, cardio muscle tissue, and corneal epithelium, and have great potential resolve current issues in the transplant field such as donor shortages, immune rejection, and ethical controversy.
The cornea is transparent tissue located in the anterior chamber of eye that is composed of 3 layers: corneal epithelium, stroma, and endothelium. The corneal epithelium originates from surface ectoderm during development, similar to the epidermis or lens epithelium [bib_ref] Characteristics of the human ocular surface epithelium, Kinoshita [/bib_ref]. Their stem/progenitor cells are believed to localize in the basal epithelium of the limbus located between the cornea and conjunctiva [bib_ref] Differentiation-related expression of a major 64K corneal keratin in vivo and in..., Schermer [/bib_ref] [bib_ref] Existence of slowcycling limbal epithelial basal cells that can be preferentially stimulated..., Cotsarelis [/bib_ref]. If corneal epithelial stem cells are completely absent due to limbal stem cell deficiencies, the peripheral conjunctival epithelium invades inwardly and the corneal surface is enveloped by vascularized conjunctival scar tissue, resulting in corneal opacification and blindness from this severe eye disease [bib_ref] Concept and application of limbal stem cells, Tseng [/bib_ref] [bib_ref] Tissue engineering of the cornea, Nishida [/bib_ref]. Although transplant therapy has been performed in patients with corneal epithelial stem cell deficiencies, most failed due to immune rejection [bib_ref] Limbal stem cell transplantation in chronic inflammatory eye disease, Samson [/bib_ref]. Regenerative medicine using differentiated autologous iPS cells has been proposed as a promising alternative; however, no differentiation strategy has been determined thus far.
Recently, human iPS cells have been established from various cell sources, including human dermal fibroblasts (HDF), keratinocytes, neural precursor cells, blood, pancreas, and testis [bib_ref] Efficient and rapid generation of induced pluripotent stem cells from human keratinocytes, Aasen [/bib_ref] [bib_ref] Pluripotent stem cells induced from adult neural stem cells by reprogramming with..., Kim [/bib_ref] [bib_ref] Generation of induced pluripotent stem cells from human blood, Loh [/bib_ref] [bib_ref] Generation of pluripotent stem cells from adult human testis, Conrad [/bib_ref] [bib_ref] Epigenetic memory and preferential lineage-specific differentiation in induced pluripotent stem cells derived..., Bar-Nur [/bib_ref]. However, these reports suggested that some of these iPS cells were limited in their differentiation capability, as they retained their original epigenetic characteristics and have the propensity to differentiate into the cell lineage originally used as cell source [bib_ref] Epigenetic memory in induced pluripotent stem cells, Kim [/bib_ref] [bib_ref] Memory in induced pluripotent stem cells: reprogrammed human retinal-pigmented epithelial cells show..., Hu [/bib_ref] [bib_ref] Reference Maps of human ES and iPS cell variation enable high-throughput characterization..., Bock [/bib_ref]. This also suggests that the generally used HDF-derived iPS cells may have limited capability to fully differentiate into other cell lineages, such as corneal epithelial cells. Human limbal epithelial cells (HLEC), in contrast, contain corneal epithelial stem/progenitor cells that may more easily differentiate into corneal epithelial cells. Thus, in this study, we attempted to establish iPS cells derived from HLECs and examine the ability of both HLEC-and HDF-derived iPS cells to differentiate into corneal epithelial cells.
# Materials and methods
## Establishment of ips cells from human corneal limbal epithelial cells (hlec)
HLECs were harvested from 2 adult human corneoscleral rims (Northwest Lions Eye Bank, Seattle, WA), according to the previously described method [bib_ref] Validation system of tissue-engineered epithelial cell sheets for corneal regenerative medicine, Hayashi [/bib_ref]. Isolated limbal epithelial cells containing corneal epithelial stem/progenitor cells were cultured in NIH/3T3 conditioned medium. Lentivirus vectors loaded with the Yamanaka 4 factors Oct3/4, Sox2, c-Myc, and Klf4 were used to reprogram limbal epithelial cells [bib_ref] Establishment of induced pluripotent stem cells from human neonatal tissues, Fujioka [/bib_ref] [fig_ref] Table 1: The summary of iPS cells established from HLEC [/fig_ref]. All experiments using recombinant DNA were approved by the Recombinant DNA Committees of Osaka and Tohoku University and performed according to the institutional guidelines.
## Human ips cell culture
Adult HDF-derived human iPS cell line 253G1 and 201B7 were obtained from RIKEN Bio Resource Center (Tsukuba, Ibaraki, Japan). 253G1, 201B7, and adult HLEC-derived human iPS cells (L1B41, L1C51, and L1B34) were maintained in culture using an MMC-treated mouse embryonic fibroblast (MEF) feeder layer in ES culture medium containing DMEM/F12 (Invitrogen, Carlsbad, CA) supplemented with 20% knockout serum replacement (KSR, Invitrogen), 0.1 mM 2-mercaptoethanol (2-ME, Invitrogen), 0.1 mM non-essential amino acid (NEAA, Invitrogen), and 4 ng/mL bFGF (Wako, Osaka, Japan) [bib_ref] In vitro differentiation of retinal cells from human pluripotent stem cells by..., Osakada [/bib_ref].
## Immunofluorescent staining
Human iPS cells were fixed in 4% paraformaldehyde (PFA) or cold methanol. Cells were washed with Tris-buffered saline (TBS, Takara Bio, Shiga, Japan) 3 times, and incubated with TBS containing 5% donkey serum and 0.3% Triton X-100, for 1 h to block nonspecific reactions. Cells were then incubated with following antibodies as appropriate: anti-Nanog (AF1997, R&D Systems, Minneapolis, MN), anti-Oct3/4 (AF1759, R&D Systems), anti-Sox2 (clone245610, R&D Systems), anti-SSEA-4 (clone MC-813-70, R&D Systems), anti-K14 (PRB-155P, Covance, Berkeley, CA), anti-Pax6 (AD2. [bib_ref] Differential methylation of tissue-and cancer-specific CpG island shores distinguishes human induced pluripotent..., Doi [/bib_ref]
# Microarray analysis
Total RNA was obtained from iPS cells using the RNeasy total RNA kit (Qiagen, Valencia, CA). Target preparation and microarray processing procedures were performed as described in the Affymetrix GeneChip Expression Analysis Manual (Affymetrix, Santa Clara, CA, USA). Data were analyzed by GeneSpring GX (Silicon Genetics, Redwood City, CA).
## Teratoma assay
For teratoma formation, 1610 6 cells of each human iPS cell clone (L1B41, L1C51, and L1B34) were injected into the testis of pentobarbital-anesthetized SCID mice (CLEA JAPAN, Tokyo, Japan). Teratomas were enucleated 4-8 weeks post-injection, fixed overnight in 10% formalin, and paraffin-embedded. The tissues were processed into 5-mm thick sections and then stained with hematoxylin and eosin (H&E staining).
Corneal epithelial differentiation from human iPS cells iPS cells were differentiated by the stromal cell-derived inducing activity (SDIA) method [bib_ref] Generation of dopaminergic neurons and pigmented epithelia from primate ES cells by..., Kawasaki [/bib_ref]. In brief, iPS cells were harvested by a dissociation solution containing 0.25% trypsin (Invitrogen) and 0.1% Collagenase type IV (Invitrogen), and then seeded on a gelatin-coated dish to eliminate MEF feeder cells for 1 h at 37uC. Non-attached cells (mainly iPS cell clumps) were harvested and reseeded on MMC-treated PA6 feeder layer at a density of 35 clumps per 3.8 mm 2 in GMEM (Invitrogen) media supplemented with 10% KSR, 1 mM sodium pyruvate, 0.1 mM NEAA, and 0.1 mM 2-ME (differentiation medium). Differentiation culture continued for 12-16 weeks. Real-time RT-PCR Total RNA was obtained from differentiated iPS cells at each culture period using the RNeasy total RNA kit. Reverse transcription was performed using the SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen) according to the manufacturer's suggested protocol, and cDNA was used as a template for PCR. Quantitative real-time RT-PCR was performed using the ABI Prism 7500 Fast Sequence Detection System (Applied Biosystems, Foster City, CA) according to the manufacturer's suggested protocol. Primer pairs and TaqManH MGB probes labeled with 6-carboxyfluorescein (FAM) at the 59-end and non-fluorescent quencher at the 39-end were designed with Assay-by-Design TM (Applied Biosystems, GAPDH: Hs99999905_m1, Pax6:
Hs00240871_m1, K12: Hs00165015_m1, K3: Hs00365080_m1, delta-N p63: Hs00978339_m1, and K14: Hs00559328_m1). The thermocycling program was performed as follows: an initial cycle at 95uC for 20 s, followed by 45 cycles of 95uC for 3 s and 60uC for 30 s. All assays were run in duplicate of 3 or more individual samples.
## Dna methylation assay
Genomic DNA (gDNA) was extracted from L1B41, 253G1, and 201B7 iPS cells, as well as from normal adult HDF (Takara Bio) and HLEC by FastPure DNA kit (Takara Bio Kyoto Japan). The methylation assay was carried out as previously described [bib_ref] High density DNA methylation array with single CpG site resolution, Bibikova [/bib_ref]. Each gDNA was bisulfite converted and enzymatically digested after amplification. The fragmented gDNA was applied to the HumanMethylation450 BeadChip (Illumina, San Diego, CA). The methylation level of each sample was calculated by the GenomeStudioH Methylation module (Illumina) as AVG-b corresponding to the ratio of the signal from the methylated allele to the sum of signal of the unmethylated allele and signal of the methylated allele. An AVG-b value of 0 corresponds to no methylation, and a value of 1.0 corresponds to 100% methylation at the specific CpG sites. Differentiation score (diffscore) represents expression differences between samples. Samples with a diffscore of greater than 633 correspond to p-value,0.001.
# Statistical analysis
Data are expressed as mean 6 S.E. Statistical analysis was performed using the Mann-Whitney rank-sum test or Student's ttest. All statistics were calculated using SigmaPlot 11.0 (Systat Software, Inc., San Jose, CA). . Gene expression analysis for corneal epithelium-related markers in iPS cells at various time points during differentiation by the SDIA method. While L1B41 began expressing significant levels of K12 after 6 weeks, other iPS cells began exhibiting K12 expression at approximately 10 weeks or later, and K12 expression in L1B41 was significantly higher than 253G1 at 6-12 weeks. Significant K3 expression levels were detected in L1B41 cells after 8 weeks in SDIA culture. Pax6 and delta-N p63 began to be expressed at 2 weeks and kept a similar expression pattern for several weeks in all iPS cells. K14 was highly expressed in both L1B41 and L1C51. The graph shows the mean 6 S.E. of 3-7 independent samples. *p,0.05 (L1B41 vs. 253G1, Mann-Whitney rank-sum test), N.S. = not significant. doi:10.1371/journal.pone.0045435.g002
# Results
## Transfection of yamanaka 4 factors can reprogram hlecs
Since HLECs share the same cell lineage as corneal epithelial cells, we tested whether iPS cells derived from them would have a propensity to develop into corneal epithelial cells. To do this, a line of iPS cells was first established by transfecting HLECs with a lentiviral vector containing the 4 transcription factors Oct3/4, Sox2, c-Myc, and Klf4 (Yamanaka 4 factors) typically used to reprogram and induce pluripotency from already-differentiated somatic cells. We selected the iPS-like colonies and cultured them in iPS cell culture conditions. After 3-7 cell passages, several iPS cell lines were isolated [fig_ref] Table 1: The summary of iPS cells established from HLEC [/fig_ref] ; among these, 3 HLEC-derived iPS cell clones (L1B41, L1B34, and L1C51) stably maintained their iPS-cell-like morphology throughout the cell passages, and were therefore selected for further testing [fig_ref] Figure 1: iPS cells were established from human corneal limbal epithelial cells [/fig_ref]. To examine whether the Yamanaka 4 factors induced expression of pluripotent stem cell markers in the selected HLEC-derived iPS cells, we analyzed Oct3/4, Nanog, SSEA-4, and Sox2 expression by immunostaining and found that all the 3 iPS cell lines expressed each of these pluripotent markers [fig_ref] Figure 1: iPS cells were established from human corneal limbal epithelial cells [/fig_ref]. The 3 iPS cell clones exhibited normal karyotype after several passages [fig_ref] Figure 1: iPS cells were established from human corneal limbal epithelial cells [/fig_ref]. Global expression analysis by microarray indicated that similar expression patterns were observed between HLEC-and HDF-derived iPS cells [fig_ref] Figure 1: iPS cells were established from human corneal limbal epithelial cells [/fig_ref]. Functional pluripotency was demonstrated in these HLEC-derived iPS cells, as the cultured L1B41 cells exhibited teratoma formation containing derivatives from all 3 germ layers, such as neural, gut, RPE, and cartilage tissues, 4-8 weeks after injection into mouse testis [fig_ref] Figure 1: iPS cells were established from human corneal limbal epithelial cells [/fig_ref]. Taken together, these data indicate that HLECs were successfully reprogrammed to iPS cells.
## Corneal epithelial cells can be differentiated from hlecand hdf-derived ips cells
We next determined whether both the HLEC-derived iPS cells (L1B41, L1C51, and L1B34) and the commercially obtained HDF-derived iPS cells (253G1 and 201B7) could differentiate into corneal epithelial cells by the SDIA cell-differentiation method. Real-time RT-PCR analysis during each differentiation period showed that only the L1B41 cells began expressing significant levels of K12, a cytokeratin protein specific for corneal epithelium, after 6 weeks . Similarly, K3, a corneal epithelial cell- specific keratin that pairs with K12, was significantly expressed after 8 weeks in L1B41 cells, even though the expression level appeared to be lower than K12. Although 253G1 and the other iPS cells also exhibited K12 expression after 10 weeks of culture, expression levels were significantly lower than L1B41. The expression pattern of Pax6, a marker of ocular and neural tissues, was similar among iPS cells. Additionally, L1B41 showed relatively higher levels of K14 and delta-N p63, markers of stratified epithelial cells, than 253G1 or 201B7. As K12 and K3 are the most specific corneal epithelial differentiation markers and not expressed in any other tissue besides corneal epithelium, the successful induction of K12 and K3 expression suggests that the SDIA method with long-term culture is sufficient to differentiate iPS cells into corneal epithelial cells.
To further examine whether iPS cells had differentiated into corneal epithelial cells, we measured protein expression and localization by immunofluorescent staining. Double immunostaining for K12 and Pax6, K12 and K14, and K12 and K3 in the SDIA-differentiated cells derived from L1B41 at 12 weeks showed that K12-expressing cells appeared as a colony, and many of these cells co-expressed Pax6 in the nuclei [fig_ref] Figure 3: Immunofluorescence of corneal epithelial cell colonies induced from iPS cells [/fig_ref] , and similarly coexpressed K14 [fig_ref] Figure 3: Immunofluorescence of corneal epithelial cell colonies induced from iPS cells [/fig_ref] and K3 [fig_ref] Figure 3: Immunofluorescence of corneal epithelial cell colonies induced from iPS cells [/fig_ref] in the cytoplasm.
To verify K12, Pax6, and K14 co-localization, triple immunostaining was performed in the differentiated cells derived from both L1B41 and 253G1 iPS cells. The results showed that the K12 + colony also co-expressed Pax6 as well as K14 in both cell types [fig_ref] Figure 3: Immunofluorescence of corneal epithelial cell colonies induced from iPS cells [/fig_ref]. We also assayed for expression of markers from other ocular epithelia cell types and showed that the differentiated cells derived from both L1B41 and 253G1 also contained RPE (pigmented cell) and lens epithelium (a-Crystallin + cell) in addition to the corneal epithelial cell markers [fig_ref] Figure 3: Immunofluorescence of corneal epithelial cell colonies induced from iPS cells [/fig_ref]. Statistical analysis revealed that the number of both Pax6-and K12-positive corneal epithelial colonies in L1B41 was significantly higher than in 253G1 [fig_ref] Figure 3: Immunofluorescence of corneal epithelial cell colonies induced from iPS cells [/fig_ref]. From these results, we conclude that longterm culture using the SDIA method can definitively induce corneal epithelium as well as RPE or lens epithelium from human iPS cells.
## Bmp4 treatment suppressed corneal epithelial differentiation from ips cells
BMP4 is known to promote surface ectodermal differentiation by suppressing neural differentiation. Since the corneal epithelium originates from surface ectoderm, we tested whether supplementing the SDIA method with BMP4 would improve differentiation of iPS cells to corneal epithelial cells. Unexpectedly, we observed almost no expression of Pax6 and K12 upon triple immunostaining for K12, Pax6, and K14 after BMP4 treatment in both SDIAdifferentiated L1B41 and 253G1 iPS cells [fig_ref] Figure 4: The effect of BMP4 treatment on iPS cells during corneal epithelial differentiation... [/fig_ref] ; some K14expressing cells were observed in BMP4-treated L1B41 cells. Realtime RT-PCR confirmed that BMP4 treatment during differentiation completely suppressed Pax6 and K12 expressions in differentiated from L1B41 as compared to the SDIA method alone [fig_ref] Figure 4: The effect of BMP4 treatment on iPS cells during corneal epithelial differentiation... [/fig_ref]. There was no significant effect on delta-N p63 expression by BMP4 treatment. In contrast to what we predicted, these data suggest that the addition of BMP4 suppressed corneal epithelial cell differentiation from iPS cells.
## Genome-wide dna methylation analyses
It was previously suggested that the differentiation propensity of iPS cells was affected by epigenomic status [bib_ref] Epigenetic memory in induced pluripotent stem cells, Kim [/bib_ref] [bib_ref] Memory in induced pluripotent stem cells: reprogrammed human retinal-pigmented epithelial cells show..., Hu [/bib_ref]. To examine whether the differential epigenomic status among iPS cells was related in the propensity of corneal epithelial differentiation, global DNA methylation status was examined by the HumanMethylation 450 BeadChip array, a newly designed high-density microarray used to quantify the methylation levels of over 450,000 CpG sites within the human genome. Scatter plot analysis showed that the difference in methylation status between L1B41 and 253G1 (R 2 = 0.976) was much smaller than between their original cell types, HLEC and HDF (R 2 = 0.771), even though the difference was larger than between 253G1 and 201B7 cells that were both derived from HDF cells (R 2 = 0.987) . Hierarchical clustering analysis revealed that L1B41 was differentially classified from either 253G1 or 201B7 . The result of genome-wide methylation analysis revealed that 32878 hypermethylated regions (6.8%) and 36849 hypomethylated regions (7.6%) were present in the original HLEC cells as compared to the original HDF cells . Comparing the methylation status among the 3 iPS cells (L1B41, 253G1, and 201B7) revealed a drastic decrease in the differentially methylated regions upon reprogramming the original . Genome wide DNA methylation analyses. (A) Scatter plots showed that a considerable difference in methylation status was observed between HLEC and HDF (A, R 2 = 0.772). Between iPS cells (L1B41 vs. 253G1), the differences became smaller (R 2 = 0.976), but were still larger than between 253G1 and 201B7 (R 2 = 0.987). (B) The methylation status in HLEC, HDF and iPS cells was shown in the heatmap with green indicating unmethylated (low AVG-b) and red indicating fully methylated (high AVG-b). Hierarchical clustering analysis showed that L1B41 and 253G1 or 201B7 were differentially classified. (C) Comparing the methylation frequency between HLEC and HDF, 32878 hypermethylated regions (6.8%) and 36849 hypomethylated regions (7.6%) in HLEC were detected. (D) After reprogramming, 2277 hypermethylated regions (0.47%) and 2040 hypomethylated regions (0.42%) in L1B41 was detected by comparing between L1B41 and 253G1 or 201B7. (E) A large proportion of differentially methylated regions between HLEC and HDF were located in non-CpG islands (the ratio of CpG islands in hyper or hypomethylated regions in HLEC was 6.8% and 16.1%, respectively). Among iPS cells, a larger proportion of the differentially methylated regions were located in CpG islands (the ratio of CpG islands in hyper or hypomethylated regions in L1B41 was 45.5% and 32.1%, respectively). doi:10.1371/journal.pone.0045435.g005 somatic cells (2277 hypermethylated regions [0.47%] and 2040 hypomethylated regions [0.42%] in L1B41 as compared to 253G1 and 201B7, respectively) . The ratio of CpG islands within the methylated regions between HLEC and HDF was relatively lower than in the other CpG-rich regions like the CpG-island shore or shelf (the hypermethylated regions were 6.8% and the hypomethylated regions were 16.1% in HLEC as compared to the 30.9% exhibited by all regions) . In contrast, the ratio of CpG islands became relatively higher between iPS cells (the hypermethylated regions were 45.5% and the hypomethylated regions were 32.1% in L1B41 as compared to the 30.9% exhibited by all regions) . Methylation analysis of individual genes revealed that no significant differences existed in the methylation status of the corneal epithelium-related markers K12, K3, Pax6, p63, and K14 genes among L1B41, 253G1, and 201B7 [fig_ref] Figure 6: DNA methylation analysis of corneal epithelium-related genes [/fig_ref].
# Discussion
Many studies show that cell lineages such as neurons, cardiac myocytes, or retinal cells can be differentiated from iPS or ES cells [bib_ref] The effects of cardioactive drugs on cardiomyocytes derived from human induced pluripotent..., Yokoo [/bib_ref] [bib_ref] Generation of retinal cells from mouse and human induced pluripotent stem cells, Hirami [/bib_ref] [bib_ref] Induction of midbrain dopaminergic neurons from ES cells by stromal cellderived inducing..., Kawasaki [/bib_ref]. However, no established strategy for corneal epithelial cell has been described. In this study, we attempted to differentiate corneal epithelial cells from iPS cells derived from both HLEC and HDF, and were able to demonstrate the first known strategy to induce corneal epithelial cells from iPS cells. We also present data that L1B41 established from HLEC showed higher propensity for differentiation to corneal epithelium than other iPS cells. To accomplish this, we showed that introducing the Yamanaka 4 factors could reprogram HLECs. At least 3 iPS cell clones were successfully established using this method, and they exhibited typical pluripotent stem cell characteristics, such as Nanog expression, teratoma formation capability, and a normal karyotype. Microarray analysis showed that the 3 HLEC-derived iPS cell clones exhibited similar expression patterns to the HDFderived 253G1 and 201B7 iPS clones with regard to complete suppression of differentiated corneal epithelial cell-related genes such as K12, K3, Pax6, delta-N p63, and K14. The efficiency of establishing iPS cells from HLECs was at least 0.0005% [fig_ref] Table 1: The summary of iPS cells established from HLEC [/fig_ref] , which was lower than iPS cells derived from other adult somatic tissues (0.0025-1.0%) [bib_ref] Induced pluripotent stem cell lines derived from human somatic cells, Yu [/bib_ref] [bib_ref] Efficient and rapid generation of induced pluripotent stem cells from human keratinocytes, Aasen [/bib_ref] [bib_ref] Epigenetic memory and preferential lineage-specific differentiation in induced pluripotent stem cells derived..., Bar-Nur [/bib_ref]. To the best of our knowledge, this is the first report to establish human iPS cells from HLEC.
From the iPS cells we developed, we successfully differentiated corneal epithelial cells by the SDIA method as assessed by expression of the specific K12 and K3 cytokeratin protein markers that are exclusively expressed in differentiated corneal epithelium tissue [bib_ref] Differentiation-related expression of a major 64K corneal keratin in vivo and in..., Schermer [/bib_ref] [bib_ref] Patterns of keratin expression define distinct pathways of epithelial development and differentiation, O'guin [/bib_ref] [bib_ref] N-Cadherin is expressed by putative stem/progenitor cells and melanocytes in the human..., Hayashi [/bib_ref]. Previous reports showed that the SDIA method could mainly induce neuro-ectodermal derivative cells such as dopaminergic neurons and RPEs [bib_ref] Generation of dopaminergic neurons and pigmented epithelia from primate ES cells by..., Kawasaki [/bib_ref] [bib_ref] Induction of midbrain dopaminergic neurons from ES cells by stromal cellderived inducing..., Kawasaki [/bib_ref]. Our results showed that long-term SDIA differentiation culture could induce K12 + /Pax6 + corneal epithelial cells after induction of differentiation of neural (data not shown), RPE, and lens cells [fig_ref] Figure 3: Immunofluorescence of corneal epithelial cell colonies induced from iPS cells [/fig_ref]. This is consistent with what is known to occur during ocular development, as corneal epithelial cells are thought to develop after RPE and lens cells and also are affected by these ocular tissues [bib_ref] Pax6 activity in the lens primordium is required for lens formation and..., Ashery-Padan [/bib_ref] [bib_ref] Primary defects in the lens underlie complex anterior segment abnormalities of the..., Collinson [/bib_ref]. Therefore, the data presented here along with previous data strongly suggest that the SDIA differentiation method in vitro mimics the processes of ocular cell development in vivo. As depicted in [fig_ref] Figure 7: Schema for corneal epithelial induction from iPS cell by the SDIA method [/fig_ref] , RPE appeared at the beginning of SDIAmediated differentiation after 4 weeks; subsequently, lens epithelium originated from surface ectoderm; finally, surface ectodermoriginated corneal epithelial cells were induced. Since corneal and lens epithelium originate during development from the same surface ectoderm in vivo, our data suggest that corneal epithelial cells and lens epithelial cells are also induced from the same origin cells during in vitro SDIA differentiation.
BMP signaling is known to promote surface ectodermal differentiation by suppressing neural differentiation [bib_ref] Induction of epidermis and inhibition of neural fate by Bmp-4, Wilson [/bib_ref] [bib_ref] Bone morphogenetic proteins in the nervous system, Mehler [/bib_ref]. Even our own previous study also showed that BMP4 could promote the epithelial differentiation of murine ES or iPS cells [bib_ref] Induction of putative stratified epithelial progenitor cells in vitro from mouse-induced pluripotent..., Sakurai [/bib_ref]. We therefore used BMP4 as a supplement to the SDIA method in order to further promote corneal epithelial differentiation in the present study. Interestingly, our data showed that BMP4 completely suppressed corneal epithelial differentiation as well as RPE and lens epithelial cell differentiation. Likely, this suppression resulted from BMP4 inhibiting early Pax6 up-regulation that is essential for ocular cell development. Alternatively, BMP signaling has also reported to play an essential role in lens development [bib_ref] BMP4 is essential for lens induction in the mouse embryo, Furuta [/bib_ref] [bib_ref] Bmp signaling is required for development of primary lens fiber cells, Faber [/bib_ref]. Although our data demonstrated that treatment with a dose of at least 0.5 mM BMP4 at an early time point during SDIAmediated differentiation suppressed the development of ocular cells from iPS cells, the concentration and/or timing of BMP4 treatment may be important factors in ocular cell induction or development.
The most important discovery in this study is the significant difference we observed in the propensity for corneal epithelial cell differentiation among different iPS cells, as HLEC-derived L1B41 iPS cells exhibited higher K12 expression and larger corneal epithelial cell colony numbers than the HDF-derived 253G1 iPS cells. Previous studies suggested that iPS cells partially retain the property of the original cell type; in particular, they retain the epigenomic signature of their original cell even after reprogramming [bib_ref] Epigenetic memory in induced pluripotent stem cells, Kim [/bib_ref] [bib_ref] Memory in induced pluripotent stem cells: reprogrammed human retinal-pigmented epithelial cells show..., Hu [/bib_ref] [bib_ref] Reference Maps of human ES and iPS cell variation enable high-throughput characterization..., Bock [/bib_ref]. The global methylation analysis we performed detected definitive regions that were differentially methylated between the L1B41 and 253G1 iPS cells, even though the difference was much smaller than between the respective HLEC and HDF original cell lineages. In contrast, methylation analysis for the individual K12, K3, Pax6, p63, and K14 genes between L1B41 and 253G1 or 201B7 iPS cells showed that there was no significant difference in their methylation patterns. These results suggest that the iPS cell propensity for corneal epithelial cell differentiation was due to the difference in methylation status of genes up-stream of K12 rather than the K12 gene itself. Although we could not identify the specific regions or genes, we further analyzed the CpG status of the methylated regions and found that the tissue-specific differentially methylated regions between HLEC and HDF were mainly located in the non-CpG islands, as previously reported [bib_ref] Differential methylation of tissue-and cancer-specific CpG island shores distinguishes human induced pluripotent..., Doi [/bib_ref] [bib_ref] Tissuespecific demethylation in CpG-poor promoters during cellular differentiation, Nagae [/bib_ref] [bib_ref] Genome-wide survey reveals dynamic widespread tissue-specific changes in DNA methylation during development, Liang [/bib_ref]. Interestingly, the ratio of deferentially methylated regions existing in CpG islands among iPS cells were higher than between HLEC and HDF, indicating that the methylation status in CpG islands was important in the inherent propensity for corneal epithelium differentiation. Although we demonstrated that corneal epithelial cells were able to differentiate from iPS cells, our data indicated that not all of the HLEC-derived iPS cells preferentially differentiated into corneal epithelial cells by the SDIA method. This suggests the possibility that the differences observed in methylation status between iPS cells resulted from differences not only between the original cell types but also the between the cell clones or even those acquired during the reprogramming process. These possibilities suggest that choosing the most appropriate iPS cell clone for corneal epithelial differentiation may be possible by analyzing the methylation status of the specific region that contributes the most to the propensity for corneal epithelial differentiation. In fact, while the gene expression among L1B41, 253G1, and 201B7 was similar, the methylation status showed some differences between L1B41 and 253G1 or 201B7.
Alternatively, some recent reports indicated that, in addition to DNA methylation, chromatin modification (such as H3K4 or H3K27 methylation) might also be important for cell differentiation [bib_ref] Distinct epigenomic landscapes of pluripotent and lineage-committed human cells, Hawkins [/bib_ref] [bib_ref] Epigenetic control of neural precursor cell fate during development, Hirabayashi [/bib_ref]. Therefore, further investigation into analyzing histone modifications and finding the specific CpG regions that regulate K12 expression will be necessary to clarify the relationship between the propensity of iPS cells to differentiate into corneal epithelial cells and epigenomic status.
In summary, the present study is the first to show that corneal epithelial cells can be generated from human iPS cells. Our study not only contributes important new discoveries for the basic research fields of corneal epithelial development and regulation of K12 expression, but also introduces a strategy to develop corneal epithelial cells that has great potential in clinical regenerative medicine to treat damaged corneal epithelium.
[fig] Figure 1: iPS cells were established from human corneal limbal epithelial cells (HLECs). (A) Three iPS-like cell colonies (L1B41, L1C51, and L1B34) were cloned after reprogramming of HLEC using Yamanaka 4 factors. (B) Immunofluorescent analysis showed that all 3 isolated iPS cell lines expressed the pluripotent stem-cell markers Nanog, Oct3/4, Sox2, and SSEA4. (C) Karyotype analysis showed no obvious aberration in the 3 iPS cell clones derived from HLEC. (D) Global expression analysis among iPS cells by microarray showed that the 3 HLEC-derived iPS cells exhibited similar expression to HDF-derived iPS cells. (E) HLEC-derived L1B41 was able to form a teratoma that contained the tissue originating from the 3 germ layers in the testis of SCID mice. Scale bar: (A) 200 mm, (B) 100 mm. doi:10.1371/journal.pone.0045435.g001 [/fig]
[fig] Figure 3: Immunofluorescence of corneal epithelial cell colonies induced from iPS cells. To examine corneal epithelial differentiation, double or triple immunostaining was performed. (A) Pax6 (green) and K12 (red) co-expressing colonies were observed in L1B41 cells differentiated for 12 weeks by the SDIA method. (B, C) Similarly, K12-expressing (red) colonies also co-expressed K14 and K3 (green) in L1B41 cells. (D) Furthermore, triple immunostaining showed that all 3 corneal epithelial differentiation markers co-localized in L1B41 (K12: red, Pax6: green, K14: blue). (F) Even in 253G1 iPS cells, corneal epithelial cell colonies were detected by immunostaining after approximately 15 weeks or later during the differentiation period (K12: red, Pax6: green, K14: blue). (E, G) The SDIA method was also able to induce RPE and lens epithelium (a-Crystallin + cells) in both iPS cell lines. The number of corneal epithelial colonies (Pax6 and K12 double-positive) in L1B41 and 253G1 induced by SDIA method for 14 weeks were counted. (H) Corneal epithelial colonies in L1B41 were significantly higher than in 253G1. The graph showed the mean 6 S.E. of 5 or 7 independent samples, respectively. Scale bar: 100 mm, *p,0.05 (t-test). doi:10.1371/journal.pone.0045435.g003 [/fig]
[fig] Figure 4: The effect of BMP4 treatment on iPS cells during corneal epithelial differentiation by the SDIA method. (A, B) 0.5 mM BMP4 was added during the first 2 weeks of differentiation by the SDIA method. Triple immunostaining showed that BMP4 treatment inhibited corneal epithelial induction from both iPS cell lines (K12: red, Pax6: green, K14: blue). (C) Real-time RT-PCR showed that Pax6 and K12 expression was completely suppressed by BMP4 treatment in L1B41 iPS cells. In contrast, BMP4 showed no significant effect on delta-N p63 expression. The graph showed the mean 6 S.E. of 3-7 independent samples, respectively. *p,0.05 (BMP4 2 vs. BMP4 + , Mann-Whitney rank-sum test), N.S. = not significant. Scale bar: 100 mm. doi:10.1371/journal.pone.0045435.g004 [/fig]
[fig] Figure 6: DNA methylation analysis of corneal epithelium-related genes. Methylation analysis of individual genes was performed. Methylation frequency in K12, K3, Pax6, p63, and K14 genes were not statistically different among L1B41, 253G1, and 201B7 (significance level; p,0.001 [ = diffscore .33 or ,233]). doi:10.1371/journal.pone.0045435.g006 [/fig]
[fig] Figure 7: Schema for corneal epithelial induction from iPS cell by the SDIA method. (A) The basic SDIA method induced the serial differentiation of neuronal cell, RPE, lens and corneal epithelial cells (CE) from iPS cells in the same culture conditions. (B) Prolonged culture periods of the iPS cells in the SDIA culture method could also induce CE following RPE or lens epithelial cell induction. The efficiency and kinetics of appearance for corneal epithelial induction was different between 253G1 and L1B41. Early BMP4 treatment in SDIA method suppressed the differentiation of ocular cell lineages from iPS cells as well as Pax6 expression. doi:10.1371/journal.pone.0045435.g007 [/fig]
[table] Table 1: The summary of iPS cells established from HLEC. [/table]
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.