Antigen-specific immune responses and liver histology in HIV and hepatitis C coinfection

Abstract

Objective: To test the hypothesis that antigen-specific interferon (IFN)gamma responses are correlated with milder liver disease in subjects coinfected with HIV-1 and hepatitis C virus (HCV).

Design: Cellular immune responses were studied in a cohort with HIV/HCV coinfection (n = 107) who underwent liver biopsy.

Methods: We measured HCV-specific and recall responses in peripheral blood mononuclear cells using IFNgamma and interleukin (IL)-10 ELISpots, and correlated these immune responses with liver histology. The relationship of immunologic, virologic and clinical variables to inflammation and fibrosis was modeled using recursive partitioning.

Results: There were significant negative correlations between inflammatory scores and IFNgamma production in response to the HCV proteins core, NS5 and summed HCV responses. Lower fibrosis scores were also correlated with higher IFNgamma production in response to NS5 and summed HCV proteins. Higher IFNgamma production in response to Candida was significantly associated with lower inflammatory and fibrosis scores. In multivariable models, factors associated with severe fibrosis were lower IFNgamma responses to Candida and summed HCV proteins. Factors associated with severe inflammation were detectable HIV viral load and lower HCV viral load, while predictors of mild inflammation included undetectable HIV viral load and higher IFNgamma response to Candida.

Conclusions: In this cohort of subjects coinfected with HIV and HCV, antigen-specific IFNgamma responses are correlated with milder inflammation and fibrosis. Immunological responses best predicted severity of fibrosis, while clinical variables and recall antigen responses best predicted severity of inflammation.

Fig. 1. CART for predicting the outcome of severe fibrosis

Forty-two percent of the overall cohort had severe fibrosis (HAI E scores > 2). The CART shows predictors that identified subgroups with higher or lower frequencies of severe fibrosis. All percentages indicate classified subjects with severe fibrosis compared to the total subjects in that classified subgroup.

Fig. 2. CART for predicting the outcome of severe inflammation

Forty-five percent of the overall cohort had severe inflammation (HAI A–D scores > 5). The CART shows predictors that identified subgroups with higher or lower frequencies of severe inflammation. All percentages indicate classified subjects with severe inflammation compared to the total subjects in that classified subgroup.