Criteria Grid Hepatitis C Research Studies, Tools, and Surveillance Systems. Part A. Basic Science Clinical Science Public Health/Epidemiology

Transcription

1 Criteria Grid Hepatitis C Research Studies, Tools, and Surveillance Systems Best Practice/Intervention: Date of Review: March 23, 2015 Reviewer(s): Christine Hu Category: Shi KQ. et al. (2012) Interleukin-28B polymorphisms on the SVR in the treatment of naive chronic hepatitis C with pegylated interferon-alpha plus ribavirin: a metaanalysis. Gene, 507(1):27-35 Part A Basic Science Clinical Science Public Health/Epidemiology Social Science Programmatic Review Best Practice/Intervention: Focus: Hepatitis C Hepatitis C/HIV Other: Level: Group Individual Other: Target Population: HCV patients Setting: Health care setting/clinic Home Other: Country of Origin: China Language: English French Other: Is the best practice/intervention a meta-analysis or primary research? Has the data/information been used for decisionmaking (e.g. program funding developments, policies, treatment guidelines, defining research priorities and funding)? Do the methodology/results described allow the reviewer(s) to assess the generalizability of the Part B YES NO N/A COMMENTS Meta-analysis; systematic review to estimate the effects of IL-28B SNPs locus on sustained virologic response in naïve chronic hepatitis C patients receiving Peg- IFN-alpha and ribavirin. Findings from the analysis were not used for decision-making. Authors suggested future studies on new therapeutic approaches which can tailor chronic HCV clinical management based on IL-28B genotyping.

2 results? Are the best practices/methodology/results described applicable in developed countries? Findings can be extended to various countries. Are the best practices/methodology/results described applicable in developing countries? The research study/tool/data dictionary is easily accessed/available electronically Is there evidence of cost effective analysis with regard to interventions, diagnosis, treatment, or surveillance methodologies? If so, what does the evidence say? Please go to Comments section Are there increased costs (infrastructure, manpower, skills/training, analysis of data) to using the research study/tool/data dictionary? How is the research study/tool funded? Please got to Comments section YES NO N/A COMMENTS Purchase required for full access from The study was supported grants from the Scientific Research Foundation of Wenzhou, Zhejiang Province, China; Health Bureau of Zhejiang Province; Research Foundation of Education Bureau of Zhejiang Province and Project of New Century 551 Talent Nurturing in Wenzhou; Natural Science Foundation of Shandong Province; Independent Innovation Foundation of Shandong University Is the best practice/intervention dependent on external funds? Other relevant criteria: - probability of achieving SVR in rs CC carriers was more than four times than non-carriers in CHC genotype 1/4 patients - rs had no influence on the SVR rate in patients with CHC genotype 2/3

3 Are these data regularly collected? Are these data regularly collected at and/or below a national level? Are these data collected manually or electronically? Has this research been published in a juried journal? Does the evidence utilize the existing data/surveillance information or has it generated new data and/or information? WITHIN THE SURVEILLANCE SYSTEM FOR REVIEW RESEARCH REPORTS Search of articles published before June, 30 th, 2011 Electronically: PubMed, EMBASE, and the Cochrane Library Gene Existing data

4 Gene 507 (2012) Contents lists available at SciVerse ScienceDirect Gene journal homepage: Interleukin-28B polymorphisms on the SVR in the treatment of naïve chronic hepatitis C with pegylated interferon-α plus ribavirin: A meta-analysis, Ke-Qing Shi a,b, Wen-Yue Liu c,a, Xian-Feng Lin c,a, Yu-Chen Fan d, Yong-Ping Chen a,b, Ming-Hua Zheng a,b, a Department of Infection and Liver Diseases, Liver Research Center, the First Affiliated Hospital of Wenzhou Medical College, Wenzhou , China b Institute of Hepatology, Wenzhou Medical College, Wenzhou , China c School of the First Clinical Medical Sciences, Wenzhou Medical College, Wenzhou , China d Department of Hepatology, Qilu Hospital of Shandong University, Jinan , China article info abstract Article history: Accepted 17 July 2012 Available online 25 July 2012 Keywords: Interleukin-28B Single nucleotide polymorphisms Hepatitis C virus Systematic review Sustained virological response There are accurate but inconclusive data on the association between single nucleotide polymorphisms (SNPs) of interleukin (IL)-28B and sustained virological response (SVR) in chronic hepatitis C (CHC). This meta-analysis aimed to derive a more precise estimation of the effects of IL-28B SNPs locus (rs and rs ) on SVR in naïve CHC patients receiving pegylated interferon alpha (PEG-IFN-α) plus ribavirin. Literature search was conducted up to June, 2011, in PubMed, EMBASE and Cochrane Database of Systematic Reviews. A total of 36 studies involving cases with CHC receiving PEG-IFN-α plus ribavirin met the inclusion criteria. Analyses were stratified either by ethnicity or genotype of hepatitis C virus. In genotype 1/4 patients, rs CC was associated with high SVR in CHC patients (Caucasian: odds ratio (OR), 4.567; 95% confidence interval (CI), ; Asian: OR, 4.033; 95%CI, ; African American: OR, 4.297; 95%CI, ; Hispanics: OR, 4.350; 95%CI, ) but had no effect in genotype 2/3. In Caucasian (genotype 1/4: OR, 2.542; 95%CI, ; genotype 2/3: OR, 1.363; 95%CI, ) and Asian (genotype 1/4: OR, 5.214; 95%CI, ; genotype 2/3: OR, 1.785; 95%CI, ), rs TT was associated with high SVR in both genotype 1/4 and 2/3. Meta-regression showed that in Caucasians with CHC genotype 1/4, gender male might contribute to the effect of rs on SVR but advanced fibrosis might weaken this effect. Furthermore, in Asians with CHC genotype 1/4, high baseline viral load and advanced fibrosis might also undermine the effect of rs on SVR. This meta-analysis suggested that IL-28B rs CC and rs TT were associated with high SVR rate in CHC genotype 1/4. In CHC genotype 2/3, rs TT carriers also had higher SVR. Crown Copyright 2012 Published by Elsevier B.V. All rights reserved. 1. Introduction An estimated 3% of the worldwide population have chronic hepatitis C virus (HCV) infection. Treatment goals in chronic hepatitis C (CHC) patients are to delay or prevent progression of fibrosis and to prevent the Abbreviations: CHC, chronic hepatitis C; CI, confidence interval; GWAS, genome-wide association study; HCV, hepatitis C virus; HIV, human immunodeficiency virus; IL, interleukin; OR, odds ratio; PEG-IFN-α, pegylated interferon-alpha; RVR, rapid virological response; SNP, single nucleotide polymorphism; SVR, sustained virological response. Potential conflict of interest: No. Grant support: This work was supported by grants from the Scientific Research Foundation of Wenzhou, Zhejiang Province, China (H , Y ), Health Bureau of Zhejiang Province (2010KYB070), Research Foundation of Education Bureau of Zhejiang Province (Y ) and Project of New Century 551 Talent Nurturing in Wenzhou, Natural Science Foundation of Shandong Province (ZR2010HQ040), Independent Innovation Foundation of Shandong University (IIFSDU, 2010TS013). Corresponding author at: Department of Infection and Liver Diseases, Liver Research Center, the First Affiliated Hospital of Wenzhou Medical College, Institute of Hepatology, Wenzhou Medical College, No. 2 Fuxue lane, Wenzhou , China. Tel.: ; fax: address: blueman1320@163.com (M.-H. Zheng). development of cirrhosis. Current treatment of chronic HCV infection is based on factors predicting sustained virological response (SVR) and includes pegylated interferon (PEG-IFN) and ribavirin (McHutchison et al., 2009). In the past decade, baseline predictors for SVR usually included alanine aminotransferase, gamma-glutamyltransferase, fibrosis, insuline resistance, steatosis, HCV genotype, baseline viral load and so on (Berg et al., 2003; Shirakawa et al., 2008). Recently, interleukin (IL)-28B single nucleotide polymorphism (SNP) has been identified to exert strong efforts on SVR in CHC patients. Advances in genome research, especially high throughput technologies for DNA sequencing and genomics, have contributed to understanding of the genetic component for decision making in CHC treatment (Clark et al., 2011). Recently, a genome-wide association study (GWAS) of more than 1600 individuals (Caucasian, American-African and Hispanics) with CHC in a clinical trial with PEG-IFN-α plus ribavirin treatment identified that a SNP on chromosome 19q13, rs C/T, was strongly associated with SVR. This SNP maps 3 kb upstream of the IL28B gene, which encodes the type III interferon IFN-λ3 (Ge et al., 2009). The carrier of CC genotype had a 2-fold or higher rate (depending on ethnicity) of SVR compared with non-carrier. Simultaneously, Tanaka /$ see front matter. Crown Copyright 2012 Published by Elsevier B.V. All rights reserved.

5 28 K.-Q. Shi et al. / Gene 507 (2012) et al. (2009) conducted a GWAS analysis in Japanese patients with CHC genotype 1, and two genetic signals in the chromosome 19q13 region were identified to be strongly related to SVR (rs and rs ). These findings were further robustly confirmed in several cohorts with different ethnicities worldwide and also apply to patients infected with HCV genotype 2 or 3 (Bochud et al., 2011; Galmozzi et al., 2010; Grebely et al., 2010; Honda et al., 2010; Hsu et al., 2011; Kawaoka et al., 2011; Mangia et al., 2010; McCarthy et al., 2010; Montes-Cano et al., 2010; Rauch et al., 2010; Sarrazin et al., 2011; Suppiah et al., 2009). Nevertheless, despite the fact that SVR was more frequently observed among rs CC or rs TT carriers with CHC genotype 1, magnitude and strength of the effect of the variant widely varied among populations of similar ethnic background, odds ratios (ORs) of rs ranging from 3.86 (Mangia et al., 2011) to 8.17 (Fattovich et al., 2011) for Caucasians, and ORs of rs ranging from 3.43 (Hayes et al., 2011) to 7.41(Kurosaki et al., 2011) for Asians. Conversely, some studies found that IL28B SNPs rs and rs were not associated with SVR in patients with CHC genotype 2/3 (Scherzer et al., 2011; Yu et al., 2011). Up to now, it is still difficult to gain one consistent conclusion in multiple studies. In the present study, this meta-analysis was aimed to derive a more precise estimation of the effects of IL-28B SNPs locus (rs and rs ) on SVR in naïve CHC patients receiving PEG-IFN-α plus ribavirin. Furthermore, we also intended to search for potential factors which might affect the effect of IL-28B SNPs on SVR using meta-regression analysis (de Boer et al., 2009). 2. Materials and methods 2.1. Search strategy An electronic search was conducted on PubMed, EMBASE, and the Cochrane Library for articles published prior to June, 30th, 2011, including the following search terms: (IL-28B, IL28B, interleukin 28, interferon λ) and (gene or variants or polymorphism or alleles or mutation) and (hepatitis C). Language was limited to English only. The references reporting the genetic variation of IL-28B and SVR of PEG-IFN-α plus ribavirin treatment in individual patients with HCV mono-infection were identified. The reference lists of identified studies and review articles were also examined to identify additional studies Selection criteria Studies were included in the current analysis if they evaluated the association between sustained virological response (SVR) after pegylated interferon alpha plus ribavirin therapy and interleukin (IL)-28B single nucleotide polymorphism (SNP) in treatment naïve patients with chronic hepatitis C virus (HCV) mono-infection. Information about the number of subjects in each ethnicity, HCV genotype, IL-28B genotype and SVR could be extracted. Authors were contacted to provide supplemental data, if data were not available in the publication. Exclusion criteria were enrollment of patients co-infected with human immunodeficiency virus, HCV-infected liver transplant recipients, other antiviral therapies involved (including Telaprevir), and studies with overlapping patient samples were excluded to only include the study with the larger number of patients Definition Advanced fibrosis was identified as METAVIR, Batts and Ludwig, or Scheuer F3-4, or Ishak F4-6. SVR was absence of detectable HCV-RNA 6 months after the cessation of therapy Data extraction The authors (Ke-Qing Shi and Yu-Chen Fan) reviewed all titles and abstracts independently to determine the eligibility criteria and retrieve the articles. Disagreement was resolved by discussion between the two authors. If the two authors could not reach an agreement, another author (Ming-Hua Zheng) was consulted to resolve the dispute, and a final decision was made by a majority vote. In order to show the relationship between IL-28B SNPs and SVR, the most strongly and independently associated SNPs were selected in the analysis (rs and rs ). For each study, the following information was abstracted according to a fixed protocol: study design, geographical location, study type, demographic information (age, sex, and ethnicity), HCV genotypes, numbers of SVR and non-svr, DNA extraction and genotyping methods, frequency of genotypes, and factors associated with SVR. Different ethnicity descents were categorized as Caucasian, Hispanics, African American and Asian. When studies included subjects of more than one ethnicity, genotype data were extracted separately. In order to address which genetic model best explains the effect of the IL-28B SNPs on CHC treatment-induced viral clearance, the dominant model (CC versus CT and TT in rs ; TT versus TG and GG in rs ) and two co-dominant models (CC versus TT, and CC versus CT in rs ; TT versus GG, and TT versus TG in rs ) were all evaluated according to HCV genotype and ethnicity Quality assessment The mean (±standard deviation) impact factor of the journals in which they had been published was used to evaluate the study quality of the studied articles Statistical analysis Strength of association of SVR with IL-28B SNP in patients with HCV mono-infection was assessed by summary odds ratios (ORs) and corresponding 95% confidence interval (CI) that were estimated by fixed and random effects meta-analysis, respectively. Considering the possible sources of heterogeneity, the studies were stratified by ethnicity, and HCV genotype, and the analysis was repeated separately for each group. Since genotypes 2/3 are more sensitive to treatment than genotype 1/4 (Di Martino et al., 2011), subgroup analysis was stratified as genotype 1/4 and genotype 2/3. For each group, we estimated the between-study heterogeneity, which was evaluated with Q statistic, a p value of less than 0.1 was considered to be representative of statistically significant heterogeneity, and the I 2 statistic, the proportion of total variation contributed by between-study variation. If the I 2 was b50%, or the p value was >0.1, a fixed-effect model was used for meta-analysis. Otherwise, a random-effect model was used. Study characteristics were also identified to stratify the studies into subsets with homogeneous effects. If heterogeneity still existed, the studies were ranked according to their individual ORs, and excluded the studies with higher ORs and repeated the process until homogeneity was achieved. And then, summary OR and corresponding 95% confidence interval were recalculated after excluding the outlier studies. Further, we performed meta-regression analysis to assess the influence of study characteristics on outcomes which were the effect estimates. Further, we performed sensitivity analyses, serially excluding studies to determine the source and assess the stability of the results. Publication bias was assessed by Begg and Mazumdar's rank correlation test which reported the rank correlation (Kendall's tau) between the standardized effect size and the variances (or standard errors) of these effects. A P value of less than 0.05 was considered to be representative of statistically significant publication bias. All calculations were performed using the Comprehensive Meta-Analysis computer program version 2 (Biostat, Englewood, NJ, USA).

6 K.-Q. Shi et al. / Gene 507 (2012) Results 3.1. Characters of the articles in our analysis Overall, a total of 39 articles from 121 relevant reports met the inclusion criteria and IL-28B rs and rs were included in this meta-analysis. The flowchart of patient selection was shown in Fig. 1. StudiesbyThompson et al. (2010) and Lagging et al. (2011) were excluded because they overlapped patient samples with studies by Ge et al. (2009) and Bochud et al. (2011). Investigators of 26 studies that fulfilled the eligibility criteria but lacked detailed data were contacted privately by or fax. In these studies, Data from one study was not accessible because we did not contact the authors successfully (Li et al., 2010). Therefore, a total of 36 studies involving a total number of 10,912 patients were included in the meta-analysis. Specifically, 14 studies referred to single rs mutation (Bitetto et al., 2011; Costa et al., 2011; Darling et al., 2011; de Rueda et al., 2011; Galmozzi et al., 2010; Ge et al., 2009; Liao et al., 2011; Lindh et al., 2010; Mangia et al., 2010, 2011; McCarthy et al., 2010; Montes-Cano et al., 2010; Patel et al., 2011; Reiberger et al., 2011), 9 studies to single rs (Grebely et al., 2010; Honda et al., 2010; Hsu et al., 2011; Kurosaki et al., 2011; Rauch et al., 2010; Sakamoto et al., 2011; Suppiah et al., 2009; Tanaka et al., 2009; Yu et al., 2011), and 13 studies to both of them (Abe et al., 2011; Bochud et al., 2011; Chen et al., 2011; Dill et al., 2011; Fattovich et al., 2011; Hayes et al., 2011; Kawaoka et al., 2011; Lin et al., 2011; Moghaddam et al., 2011; Sarrazin et al., 2011; Scherzer et al., 2011; Sinn et al., 2011; Stattermayer et al., 2011). There were 17 studies of Caucasian, 14 of Asian, 2 of Hispanics and 3 of mix ethnicities (Supplementary Table 1). Up to now, the association between IL-28B SNPs and SVR was rarely reported in other populations (Supplementary Figures 2 and 3) Association between rs C/T polymorphism and SVR CHC genotype 1/4 Thirteen studies were about rs C/T polymorphism and SVR in Caucasian patients with CHC genotype 1/4. Overall, 1494 of 2952 (50.6%) patients received SVR after anti-hcv treatment. Carriers of rs CC had a significantly increased SVR rate (OR: 4.567; 95%CI: ; Pb in the random model) than non-carriers (Fig. 2A). There was evidence of heterogeneity among these studies (Q=34.58; P=0.001 for heterogeneity; I 2 =65.24). After removing one outlier study in which the patients were all HCV related cirrhosis, the heterogeneity was still present (Reiberger et al., 2011). No publication bias existed among these studies (P=0.799). The correlation between male proportion (slope: 0.034; 95%CI: ; P=0.002) and advanced fibrosis proportion (slope: 0.013; 95%CI: ; P=0.017) in the studied population and the effect of rs on SVR in Caucasian patients with CHC genotype 1/4 were observed using meta-regression analysis (Figs. 4A and B), which suggested that gender and advanced fibrosis proportion might be involved in the effect of rs on SVR in Caucasian patients with CHC genotype 1/4. Data regarding rs C/T polymorphism and SVR in Asian patients with CHC genotype 1/4 were available in 6 homogeneous studies (Q=2.99; P=0.702; I 2 =0), including 1531 patients, with 831 (54.3%) patients achieving SVR. Carriers of rs CC vs. non-carriers had a significantly increased rate of SVR (OR: 4.033; 95%CI: ; Pb in the fixed model) (Fig. 2A). There was publication bias among these studies (P=0.013). In three homogeneous studies (Q=1.12; P=0.572; I 2 =0) concerned rs C/T polymorphism and SVR in African Americans with CHC genotype 1/4, SVR in the carriers of rs CC (48.9%, 23/47) was significantly higher than non-carriers (28.3%, Fig. 1. Flowchart of study selection.

7 30 K.-Q. Shi et al. / Gene 507 (2012) 27 35

8 K.-Q. Shi et al. / Gene 507 (2012) /230) (OR: 4.297; 95%CI: ; Pb in the fixed model) (Fig. 2A). There was no publication bias among these studies (P=0.116). In addition, we found 3 homogeneous studies (Q=0.477; P=0.788; I 2 =0) about rs C/T polymorphism and SVR in Hispanic patients with CHC genotype 1/4, including 474 patients with 257 (54.2%) achieving SVR. Carriers of rs CC vs non-carriers had a significantly increased rate of SVR (OR: 4.350; 95%CI: ; Pb in the fixed model) (Fig. 2A). No publication bias existed among these studies (P=0.251) CHC genotype 2/3 Data regarding rs C/T polymorphism and SVR in Caucasian patients with CHC genotype 2/3 were available in 14 heterogeneous studies (Q=25.04; P=0.023; I 2 =48.08), including 1850 patients, with 1237 (79.7%) patients achieving SVR. Carriers of rs CC vs non-carriers had no significantly increased rate of SVR (OR: 1.241; 95%CI: ; P=0.060 in the random model) (Fig. 2B). After removing one outlier study in which the patients were all HCV related cirrhosis (Reiberger et al., 2011), the heterogeneity was still present. No publication bias existed among these studies (P=0.411). No studyrelated factor was found that influenced the effect of rs on SVR in Caucasian patients with CHC genotype 2/3. We found 4 homogeneous studies (Q=0.931; P=0.818; I 2 =0) about rs C/T polymorphism and SVR in Asian patients with CHC genotype 2/3. SVR in the carriers of rs CC (87.6%, 367/ 419) was similar to non-carriers (69.4%, 43/62) (OR: 1.407; 95%CI: ; P=0.381 in the fixed model) (Fig. 2B). No publication bias existed among these studies (P=0.219). Only one study concerned rs C/T polymorphism and SVR in African Americans with CHC genotype 2/3 was available, including only 2 patients (McCarthy et al., 2010). Therefore, we did not perform the meta-analysis in this situation. Data about rs C/T polymorphism and SVR in Hispanics patients with CHC genotype 2/3 were available in 2 homogeneous studies (Q=0.596; P=0.440; I 2 =0). SVR in the carriers of rs CC (49.3%, 37/75) was similar to non-carriers (28.6%, 16/56) (OR: 2.120; 95%CI: ; P=0.062 in the fixed model) (Fig. 2B) Association of rs T/G polymorphism and SVR CHC genotype 1/4 Data about rs T/G polymorphism and SVR in Caucasian patients with CHC genotype 1/4 were available in 7 homogeneous studies (Q=9.447; P=0.150; I 2 =36.491), including 1916 patients, with 978 (51.0%) patients achieving SVR. Carriers of rs TT vs non-carriers had a significantly increased rate of SVR (OR: 2.542; 95%CI: ; Pb in the fixed model) (Fig. 3A). No publication bias existed among these studies (P=0.429). Nine heterogeneous studies (Q=14.81; P=0.063; I 2 =45.99) about rs T/G polymorphism and SVR in Asian patients with CHC genotype 1/4 were found, including 2425 patients with 1206 (49.7%) patients achieving SVR. Carriers of rs TT vs non-carriers had a significantly increased rate of SVR (OR: 5.214; 95%CI: ; Pb in the random model) (Fig. 3A). After removing one study which had the highest individual OR (Tanaka et al., 2009), the heterogeneity was removed and the effect was still significant(or:4.297;95%ci: ; Pb in the fixed model). No publication bias existed among these studies (P=0.614). The correlation between baseline viral load (slope: 1.536; 95%CI: ; P=0.013) and advanced fibrosis proportion (slope: 0.029; 95%CI: ; P=0.002) in the studied population and the effect of rs on SVR in Asian patients with CHC genotype 1/4 were observed using meta-regression analysis (Figs. 4C and D), which suggested that baseline viral load and advanced fibrosis proportion might be involved in the effect of rs on SVR in Asian patients with CHC genotype 1/4. None study was found in other ethnicities concerned the rs T/G polymorphism and SVR in patients with CHC genotype 1/ CHC genotype 2/3 Data about rs T/G polymorphism and SVR in Caucasian patients with CHC genotype 2/3 were available in 9 homogeneous studies (Q=6.523; P=0.589; I 2 =0), including 1216 patients, with 957 (78.7%) patients achieving SVR. Carriers of rs TT vs non-carriers had a significantly increased rate of SVR (OR: 1.363; 95%CI: ; P=0.036 in the fixed model) (Fig. 3B). No publication bias existed among these studies (P=0.466). Six homogeneous studies (Q=4.500; P=0.480; I 2 =0) about rs T/G polymorphism and SVR in Asian patients with CHC genotype 2/3 were found, including 1113 patients with 966 (86.8%) patients achieving SVR. Carriers of rs TT vs non-carriers had a significantly increased rate of SVR (OR: 1.785; 95%CI: ; P=0.020 in the fixed model) (Fig. 3B). No publication bias existed among these studies (P=0.462). None study was found in other ethnicities concerned the rs T/G polymorphism and SVR in patients with CHC genotype 2/3. The two co-dominant models (CC versus TT and CC versus CT in rs ; TT versus GG and TT versus TG in rs ) also had similar effects (Tables 1 and 2). The heterogeneity was found in Caucasian patients with HCV genotype 1/4 when comparing CC and CT in rs Then, we performed meta-regression analysis to assess the influence of study characteristics on outcomes. The results indicated that male proportion (slope: 0.034; 95%CI: ; P=0.004) and advanced fibrosis proportion (slope: 0.014; 95%CI: ; P=0.015) might be involved in the effect, which was similar to those in dominant model (Supplementary Figure 1) Sensitivity analysis Sensitivity analysis was performed if data from more than three individual studies could be pooled in the subgroup. ORs were not influenced excessively by omitting any single study (data not shown) Quality of studies The mean impact factor for all the included studies was 10.30± 8.10, suggesting that the studies we have included in the analysis were published in leading journals with a high impact factor. 4. Discussion Although the molecular mechanism responsible for the genetic effects of IL-28B on HCV clearance remains obscure, IL28B SNPs has been widely demonstrated to be associated with the SVR rate in naïve CHC patients with PEG-IFN-α plus ribavirin therapy in different ethnic populations. This present meta-analysis indicated that the probability of achieving SVR in rs CC carriers was more than four times than non-carriers in CHC genotype 1/4 patients. However, rs had no influence on the SVR rate in patients with CHC genotype 2/3.This finding strongly suggested that the impact of the rs might be less evident in genotype 2/3 infected patients and the possible interpretation for this result is that genotypes Fig. 2. Forest plots of pooled OR with 95% CI for associations between rs C/T (CC vs CT+TT) and SVR stratified by ethnicity and HCV genotype. (A) CHC genotype 1/4. (B) CHC genotype 2/3. The first author of the study and the year of publication are shown after citation name; a stands for Caucasian; b stands for African American; c stands for Hispanics. The squares and horizontal lines correspond to the study specific OR and 95% CI; the diamonds represent the pooled OR and 95% CI.

9 32 K.-Q. Shi et al. / Gene 507 (2012) Fig. 3. Forest plots of pooled OR with 95% CI for associations between rs T/G (TT vs TG+GG) and SVR stratified by ethnicity and HCV genotype. (A) CHC genotype 1/4. (B) CHC genotype 2/3. The first author of the study and the year of publication are shown after citation name. The squares and horizontal lines correspond to the study specific OR and 95% CI; the diamonds represent the pooled OR and 95% CI. 2/3 is more sensitive to SVR in CHC patients than genotype 1/4 (Di Martino et al., 2011). Carriers of rs TT had a significantly increased SVR rate than non-carriers in both Caucasian and Asian population, regardless of HCV genotype. Therefore, in CHC genotype 1/4, despite the ethnic heterogeneity in IL28B genotypes and the consequent variable rates of SVR to combination treatment, ranging from 53% in African Americans to 82% in Caucasians (Ge et al., 2009), both rs and rs are good candidates for a powerful diagnostic tool to identify subjects who are more likely to acquire SVR. Furthermore, an additive genetic effect of rs on SVR was observed only in Caucasians and Hispanics with CHC genotype 1/4, while rs had no additive genetic effect on achieving

10 K.-Q. Shi et al. / Gene 507 (2012) Fig. 4. Meta-regression analysis for association between study characteristics and the effect of the rs or rs on SVR. (A) Meta-regression analysis between male proportion among the studies and the effect of the rs on SVR in Caucasians with CHC genotype 1/4. (B) Meta-regression analysis between advanced fibrosis proportion among the studies and the effect of the rs on SVR in Caucasians with CHC genotype 1/4. (C) Meta-regression analysis between baseline viral load among the studies and the effect of the rs on SVR in Asians with CHC genotype 1/4. (D) Meta-regression analysis between advanced fibrosis proportion among the studies and the effect of the rs on SVR in Asians with CHC genotype 1/4. Log odds ratio which is the effect estimate for SVR in meta-analysis is shown in Y-axis. Characteristic of studies (male proportion, advanced fibrosis proportion, or baseline viral load) is shown in X-axis. Each circle represents a study with a size proportional to the sample size. SVR. These observations suggested that SVR of PEG-IFN-α plus ribavirin therapy can be determined by the major genetic model more correctly. A possible role of ethnic differences in genetic backgrounds and the environment they lived in should be considered. Although the two main IL28B SNPs (rs and rs ) belong to a same haplotype block with same linkage disequilibrium in most of the studies (Clark et al., 2011), they may exert different influence. It is also likely that the analysis of the heterozygosity for the variant may be due to chance, because studies with small sample size in the minor homozygous from both SVR and non-svr groups may have insufficient statistical power to detect a slight effect or may have generated a fluctuated risk estimate. Considering this situation, evaluation of the effect associated with heterozygosity for the variant in our analysis should be interpreted with caution. In addition, meta-regression showed that in Caucasians with CHC genotype 1/4, male proportion was correlated positively with the effect of rs on SVR; but advanced fibrosis had a negative correlation. Negative association between the effect of rs on SVR in Asians with CHC genotype 1/4 and high baseline viral load and advanced fibrosis were also observed using meta-regression analysis. These factors have been identified as predictors to successful treatment for CHC (Shirakawa et al., 2008). Many studies have tested the association between IL-28B and baseline viral load (Darling et al., 2011; Ge et al., 2009; McCarthy et al., 2010), which might relate to the role of interferon stimulated gene expression in modulating the response to PEG-IFN (Dill et al., 2011; Honda et al., 2010). However, the present study could not explain the interaction between IL-28B and gender and fibrosis stage. Nevertheless, host genes implicated in the immune response were reported to be related with baseline viral load and advanced fibrosis, which might interact with IL-28B and contribute to PEG-IFN sensitivity (Romero-Gomez et al., 2011). Therefore, combination of IL28B SNPs and these known clinical predictors of SVR (HCV genotype, viral load, and host characteristics) will obtain high accuracy in defining the probability of achieving SVR, which could be used for tailoring regimen dose and duration of treatment. In meta-analysis studies, heterogeneity may potentially restrict the interpretation of the pooled estimates. Ethnicity and HCV genotype might play a role in determining the heterogeneity among studies. Different allele frequencies among different ethnicities (rs C allele frequency in African: 23.1% 54.8%; Caucasians: 52.8% 85.7%; and Asian: 90.0% 100.0%) are a strong cause of heterogeneity (Thomas et al., 2009). The different frequency distribution of rs CC and rs TT in different CHC genotypes was also observed (genotype 2/3, 45%; genotype 1/4, 33%) (Lindh et al., 2010; McCarthy et al., 2010). So the analysis was stratified by ethnicity and HCV genotype. In fact, heterogeneity still remained in rs C/T polymorphism and SVR in

11 34 K.-Q. Shi et al. / Gene 507 (2012) Table 1 Magnitude of the effect of co-dominant models for rs on SVR. Ethnicity Model CC versus TT CC versus CT OR (95% CI) P P (Q-test) OR (95% CI) P P (Q-test) HCV genotype 1/4 Caucasian Fixed ( ) ( ) Random ( ) ( ) Asian Fixed ( ) ( ) Random ( ) ( ) African-American Fixed ( ) ( ) Random ( ) ( ) Hispanics Fixed ( ) ( ) Random ( ) ( ) Total Fixed ( ) ( ) Random ( ) ( ) HCV genotype 2/3 Caucasian Fixed ( ) ( ) Random ( ) ( ) Asian Fixed ( ) ( ) Random ( ) ( ) Hispanics Fixed ( ) ( ) Random ( ) ( ) Total Fixed ( ) ( ) Random ( ) ( ) CI, confidence interval; HCV, hepatitis C virus; OR, odds ratio; SVR, sustained virological response. Caucasian patients with CHC genotype 1/4 and CHC genotype 2/3, and rs T/G polymorphism and SVR in Asian patients with CHC genotype 1/4. Although the effects could be estimated in the different studies which are not identical by applying a random effect model, we also analyzed other study characteristics which could explain the significant heterogeneity in the comparisons: different gender distribution, disparity in age, different level of baseline viral load and different fibrosis stages. Some limitations of this study should be taken into consideration: first, studied populations were mainly from Caucasian and East Asian. Other populations in the subgroup meta-analysis of race were relatively small or none (Supplementary Figures 2 and 3). Thus, more data is required from other ethnic groups for strong statistical power. Second, the kinetics of viral load during treatment were not considered when pooling estimates in current study, although several studies found that rs CC/rs TT genotypes were related to rapid virological response (RVR) (Mangia et al., 2010, 2011; Moghaddam et al., 2011; Thompson et al., 2010; Yu et al., 2011). Because only three studies had sufficient data analyzing the SVR rate for IL-28B SNPs according to RVR, which were concerned different ethnicities, different IL-28B SNPs, and different HCV genotypes (Supplementary Table 1) (Mangia et al., 2011; Thompson et al., 2010; Yu et al., 2011). The analysis was not performed. More studies concerning the impact of IL-28B SNPs on viral kinetics during treatment and SVR are needed. Third, IL-28B SNPs contributes to only a part of differences in SVR between races. Therefore, other genetic and viral factors associated with SVR should be considered in our further study. Finally, the data of our meta-analysis is summary data from comparative studies, and individual patient-level data could not be combined, which would allow for the adjustment by other covariates and have a more precise analysis (Zheng et al., 2011). A phenomenon in data extracting was that most studies did not have a complete record of genotype counts for SVR and non-svr (data was obtained from 25 studies via personally /fax). Therefore, a reporting guideline genetic association studies is encouraged to facilitate further meta-analysis of the genetic effect with more robust estimates and to evaluate the study quality of this type of articles (Janssens et al., 2011). In conclusion, this meta-analysis demonstrates that rs CC and rs TT are both associated with increased SVR rate in CHC genotype 1/4 regardless of ethnicity. And rs TT also has higher Table 2 Magnitude of the effect of co-dominant models for rs on SVR. Ethnicity Model TT versus GG TT versus TG OR (95% CI) P P (Q-test) OR (95% CI) P P (Q-test) HCV genotype 1/4 Caucasian Fixed ( ) ( ) Random ( ) ( ) Asian Fixed ( ) ( ) Random ( ) ( ) Total Fixed ( ) ( ) Random ( ) ( ) HCV genotype 2/3 Caucasian Fixed ( ) ( ) Random ( ) ( ) Asian Fixed ( ) ( ) Random ( ) ( ) Total Fixed ( ) ( ) Random ( ) ( ) CI, confidence interval; HCV, hepatitis C virus; OR, odds ratio; SVR, sustained virological response.

12 K.-Q. Shi et al. / Gene 507 (2012) SVR rate in both Caucasians and Asians genotype 2/3 patients. In the future, new therapeutic approaches which can tailor CHC clinical management based on IL-28B genotyping are needed. Supplementary data to this article can be found online at dx.doi.org/ /j.gene References Abe, H., et al., IL28 variation affects expression of interferon stimulated genes and peg-interferon and ribavirin therapy. J. Hepatol. 54, Berg, T., et al., Prediction of treatment outcome in patients with chronic hepatitis C: significance of baseline parameters and viral dynamics during therapy. Hepatology 37, Bitetto, D., et al., Complementary role of vitamin D deficiency and the interleukin- 28B rs C/T polymorphism in predicting antiviral response in chronic hepatitis C. Hepatology 53, Bochud, P.Y., et al., IL28B polymorphisms predict reduction of HCV RNA from the first day of therapy in chronic hepatitis C. J. Hepatol., j.jhep Chen, J.Y., et al., IL28B genetic variations are associated with high sustained virological response (SVR) of interferon-alpha plus ribavirin therapy in Taiwanese chronic HCV infection. Genes Immun. 12, Clark, P.J., Thompson, A.J., McHutchison, J.G., IL28B genomic-based treatment paradigms for patients with chronic hepatitis C infection: the future of personalized HCV therapies. Am. J. Gastroenterol. 106, Costa, J.M., et al., HCV-GenoFibrotest: a combination of viral, liver and genomic (IL28b, ITPA, UGT1A1) biomarkers for predicting treatment response in patients with chronic hepatitis C. Clin. Res. Hepatol. Gastroenterol. 35, Darling, J.M., et al., Quantitation of pretreatment serum interferon-gammainducible protein-10 improves the predictive value of an IL28B gene polymorphism for hepatitis C treatment response. Hepatology 53, de Boer, A.G., Taskila, T., Ojajarvi, A., van Dijk, F.J., Verbeek, J.H., Cancer survivors and unemployment: a meta-analysis and meta-regression. JAMA 301, de Rueda, P.M., et al., Importance of host genetic factors HLA and IL28B as predictors of response to pegylated interferon and ribavirin. Am. J. Gastroenterol., dx.doi.org/ /ajg Di Martino, V., et al., Response-guided Peg-interferon plus ribavirin treatment duration in chronic hepatitis C: meta-analyses of randomized controlled trials and implications for the future. Hepatology, Dill, M.T., et al., Interferon-induced gene expression is a stronger predictor of treatment response than IL28B genotype in patients with hepatitis C. Gastroenterology 140, Fattovich, G., et al., IL28B polymorphisms, IP-10 and viral load predict virological response to therapy in chronic hepatitis C. Aliment. Pharmacol. Ther. 33, Galmozzi, E., et al., A tetra-primer amplification refractory mutation system polymerase chain reaction for the evaluation of rs IL28B genotype. J. Viral Hepat., Ge, D., et al., Genetic variation in IL28B predicts hepatitis C treatment-induced viral clearance. Nature 461, Grebely, J., et al., Potential role for interleukin-28b genotype in treatment decision-making in recent hepatitis C virus infection. Hepatology 52, Hayes, C.N., et al., HCV substitutions and IL28B polymorphisms on outcome of peg-interferon plus ribavirin combination therapy. Gut 60, Honda, M., et al., Hepatic ISG expression is associated with genetic variation in interleukin 28B and the outcome of IFN therapy for chronic hepatitis C. Gastroenterology 139, Hsu, C.S., et al., Association of IL28B gene variations with mathematical modeling of viral kinetics in chronic hepatitis C patients with IFN plus ribavirin therapy. Proc. Natl. Acad. Sci. U. S. A. 108, Janssens, A.C., Ioannidis, J.P., van Duijn, C.M., Little, J., Khoury, M.J., Strengthening the reporting of genetic risk prediction studies: the GRIPS statement. PLoS Med. 8, e Kawaoka, T., et al., Predictive value of the IL28B polymorphism on the effect of interferon therapy in chronic hepatitis C patients with genotypes 2a and 2b. J. Hepatol. 54, Kurosaki, M., et al., Pre-treatment prediction of response to pegylated-interferon plus ribavirin for chronic hepatitis C using genetic polymorphism in IL28B and viral factors. J. Hepatol. 54, Lagging, M., et al., Response prediction in chronic hepatitis C by assessment of IP- 10 and IL28B-related single nucleotide polymorphisms. PLoS One 6, e Li, J.H., et al., Interferon-lambda genotype and low serum low-density lipoprotein cholesterol levels in patients with chronic hepatitis C infection. Hepatology 51, Liao, X.W., et al., Association of genetic variation in IL28B with hepatitis C treatment-induced viral clearance in the Chinese Han population. Antivir. Ther. 16, Lin, C.Y., et al., IL28B SNP rs is a critical predictor for on-treatment and sustained virologic response in patients with hepatitis C virus genotype-1 infection. PLoS One 6, e Lindh, M., Lagging, M., Norkrans, G., Hellstrand, K., A model explaining the correlations between IL28B-related genotypes, hepatitis C virus genotypes, and viral RNA levels. Gastroenterology 139, Mangia, A., et al., An IL28B polymorphism determines treatment response of hepatitis C virus genotype 2 or 3 patients who do not achieve a rapid virologic response. Gastroenterology 139, e1. Mangia, A., et al., Limited utility of IL28B in the setting of response-guided treatment with detailed on-treatment virological monitoring. Hepatology, dx.doi.org/ /hep McCarthy, J.J., et al., Replicated association between an IL28B gene variant and a sustained response to pegylated interferon and ribavirin. Gastroenterology 138, McHutchison, J.G., et al., Peginterferon alfa-2b or alfa-2a with ribavirin for treatment of hepatitis C infection. N. Engl. J. Med. 361, Moghaddam, A., et al., IL28B genetic variation and treatment response in patients with hepatitis C virus genotype 3 infection. Hepatology 53, Montes-Cano, M.A., et al., Interleukin-28B genetic variants and hepatitis virus infection by different viral genotypes. Hepatology 52, Patel, K., et al., High predictive accuracy of an unbiased proteomic profile for sustained virologic response in chronic hepatitis C patients. Hepatology 53, Rauch, A., et al., Genetic variation in IL28B is associated with chronic hepatitis C and treatment failure: a genome-wide association study. Gastroenterology 138, e1-7. Reiberger, T., et al., Portal pressure predicts outcome and safety of antiviral therapy in cirrhotic patients with hepatitis C virus infection. Clin. Gastroenterol. Hepatol. 9, e1. Romero-Gomez, M., Eslam, M., Ruiz, A., Maraver, M., Genes and hepatitis C: susceptibility, fibrosis progression and response to treatment. Liver Int. 31, Sakamoto, N., et al., Association of IL28B variants with response to pegylatedinterferon alpha plus ribavirin combination therapy reveals intersubgenotypic differences between genotypes 2a and 2b. J. Med. Virol. 83, Sarrazin, C., et al., Importance of IL28B gene polymorphisms in hepatitis C virus genotype 2 and 3 infected patients. J. Hepatol. 54, Scherzer, T.M., et al., Early virologic response and IL28B polymorphisms in patients with chronic hepatitis C genotype 3 treated with peginterferon alfa-2a and ribavirin. J. Hepatol. 54, Shirakawa, H., et al., Pretreatment prediction of virological response to peginterferon plus ribavirin therapy in chronic hepatitis C patients using viral and host factors. Hepatology 48, Sinn, D.H., et al., Association of a single nucleotide polymorphism near the interleukin-28b gene with response to hepatitis C therapy in Asian patients. J. Gastroenterol. Hepatol., Stattermayer, A.F., et al., Impact of IL28B genotype on the early and sustained virologic response in treatment-naive patients with chronic hepatitis C. Clin. Gastroenterol. Hepatol. 9, e2. Suppiah, V., et al., IL28B is associated with response to chronic hepatitis C interferon-alpha and ribavirin therapy. Nat. Genet. 41, Tanaka, Y., et al., Genome-wide association of IL28B with response to pegylated interferon-alpha and ribavirin therapy for chronic hepatitis C. Nat. Genet. 41, Thomas, D.L., et al., Genetic variation in IL28B and spontaneous clearance of hepatitis C virus. Nature 461, Thompson, A.J., et al., Interleukin-28B polymorphism improves viral kinetics and is the strongest pretreatment predictor of sustained virologic response in genotype 1 hepatitis C virus. Gastroenterology 139, e18. Yu, M.L., et al., Role of interleukin-28b polymorphisms in the treatment of hepatitis C virus genotype 2 infection in Asian patients. Hepatology 53, Zheng, M.H., Shi, K.Q., Fan, Y.C., Chen, Y.P., Meta-analysis using individual participant data is the gold standard for diagnostic studies. Hepatology 53, 1062 author reply