No abstract

In a well-characterized real-life clinical cohort of chronically HBV-infected patients in various disease phases, older age, and inactive HBsAg carrier status were the only predictors of HBsAg clearance, whereas anti-HBV therapy only marginally increased annual incidence of HBsAg loss.

The poor response to the combined antiviral therapy of pegylated alfa-interferon and ribavarin for hepatitis C virus (HCV) infection may be linked to mutations in the viral envelope gene E1E2 (env), which can result in escape from the immune response and higher efficacy of viral entry. Mutations that result in failure of therapy most likely require compensatory mutations to achieve sufficient change in envelope structure and function. Compensatory mutations were investigated by determining positions in the E1E2 gene where amino acids (aa) covaried across groups of individuals. We assessed networks of covarying positions in E1E2 sequences that differentiated sustained virological response (SVR) from non-response (NR) in 43 genotype 1a (17 SVR), and 49 genotype 1b (25 SVR) chronically HCV-infected individuals. Binary integer programming over covariance networks was used to extract aa combinations that differed between response groups. Genotype 1a E1E2 sequences exhibited higher degrees of covariance and clustered into 3 main groups while 1b sequences exhibited no clustering. Between 5 and 9 aa pairs were required to separate SVR from NR in each genotype. aa in hypervariable region 1 were 6 times more likely than chance to occur in the optimal networks. The pair 531–626 (EI) appeared frequently in the optimal networks and was present in 6 of 9 NR in one of the 1a clusters. The most frequent pairs representing SVR were 431–481 (EE), 500–522 (QA) in 1a, and 407–434 (AQ) in 1b. Optimal networks based on covarying aa pairs in HCV envelope can indicate features that are associated with failure or success to antiviral therapy.