Aim: To evaluate the association between HLA A, B with chronic Hepatitis B by comprehensive meta-analysis.
Methods: We searched PubMed and Cochrane databases, identified relevant studies, evaluated these for quality by New Castle Ottawa scale (NOS) and further analyzed the qualified data sets. Heterogeneity analyses were performed by Cochrane's Q test and I2 tests. Pooled Odds ratio (OR) & 95% Confidence Interval (CI) were obtained by fixed effects, using Mantel-Haenszel's method for homogenous studies, and by using DerSimonian and Laird's method for heterogenous studies. Publication bias was determined by the Beggs test and Eggers test and all tests were two tailed to evaluate their significance.
Results: The meta-analyses on 1652 healthy controls and 659 Chronic Hepatitis B (CHB) patients from 8 studies from various continents revealed a HLA B*07 (p value of Odds ratio (pOR)=0.004; OR Fixed effects=0.480 with 95%CI 0.290-0.794) and B*58 (pOR=0.029; OR Fixed Effects=0.020 with 95%CI 0.381-0.949) associated protection for CHB. The identified HLA B*35 associated risk (pOR 0.009; OR Fixed effect 1.445; 95% confidence interval 1.094-1.907) however did not stand the test of random effect model.
Conclusion: While HLA B*07 and B*58 are protective against CHB. The HLA B*35 associated marginal risk need to be further validated in well-designed global study on larger cohorts, considering the population, ethnic, epidemiological and HLA diversity at the sequence level: these may throw further light to utilize these markers in predictive medicine.
Keywords: HLA A; HLA B; Hepatitis B; Meta-analysis; Predictive medicine.
Copyright © 2017 Elsevier B.V. All rights reserved.