AEGiS-14IAC: Hepatitis G virus (HGV) infection does not prolong survival of patients with early-stage HIV disease: importance of baseline HIV viral load as a predictor of mortality.

14th International AIDS Conference

Barcelona, Spain - July 7-12, 2002

Hepatitis G virus (HGV) infection does not prolong survival of patients with early-stage HIV disease: importance of baseline HIV viral load as a predictor of mortality.

Int Conf AIDS 2002 Jul 7-12; 14:(abstract no. WeOrC1378)

Brust D, Jagannatha S, Herpin B, Miller K, Metcalf J, Lau D, Alter H, Lane HC, Falloon J, Fauci A
NIH, Bethesda, United States

BACKGROUND: Two recent studies reported that HGV infection improved the outcome of HIV disease; however, neither study evaluated baseline HIV viral load, a known predictor of outcome. To address this issue, we analyzed the impact of HGV infection on survival in a cohort originally enrolled in an HIV treatment protocol from 1988-91.

METHODS: 180 subjects with early disease (CD4 counts > 500 cells/mm³ and no history of OIs) were equally randomized to receive either daily zidovudine (ZDV), interferon (INF), or ZDV and INF to assess the impact of these regimens on HIV. At study end, subjects continued to follow-up. Baseline HIV viral load and HGV RNA were performed on frozen samples. Subject survival was determined by chart review and a Death Index database search. To determine the impact of HGV status on survival, a Kaplan-Meier (KM) analysis and a Cox proportional hazards model (which accounted for such variables as CD4 count, HIV viral load and treatment arm) were performed.

RESULTS: 37% of the cohort was HGV infected. Baseline HIV viral load and CD4 count were not different between HGV-positive and negative subjects. Median follow-up was 9.6 years. At study termination, 49% of patients were alive, 27% dead, and 24% lost-to-follow-up. A KM analysis indicated that HGV co-infected participants had increased mortality when compared to HGV-negative subjects (p = 0.036). However, when other factors associated with HIV survival were accounted for, the only robust predictor of survival was baseline HIV viral load (p = 0.008).

CONCLUSIONS: In this study, HGV co-infection tended to worsen, rather than improve, mortality. Our analysis underscores the importance of including baseline HIV viral load when modeling survival. Additionally, our results suggest that the proposed protective effect of HGV co-infection may not be universally applicable to all HIV-infected people, and in particular, may be linked to disease stage.

Keywords: AEGIS, GB virus C, HIV Infections, CD4 Lymphocyte Count, HIV, Survival, Survival Rate, Proportional Hazards Models, Greece, Human, mortality

020707
WeOrC1378