PREDICTORS OF CLINICALLY RELEVANT HEPATOTOXICITY (HT) IN HIV-HCV CO-INFECTED INDIVIDUALS

CL Cooper, MA Parbhakar, JB Angel
Division of Infectious Diseases, Ottawa Hospital, University of Ottawa, Ottawa, Ontario

Objectives: To determine risk factors for clinically relevant HT in HIV-HCV co-infected individuals.
Methods: A retrospective chart review of co-infected subjects followed at our clinic between January 1994 and April 2001 was conducted. Subjects were included if they had received their first protease inhibitor (PI)-based HAART regimen at our site and adequate documentation was available. Kaplan-Meier and Cox regression analysis were performed to evaluate rate and risk of HT, defined as symptomatic hepatomegaly on examination, or acute transaminitis (i.e.³40 U/ml over baseline level) associated with right upper quadrant pain and/or constitutional symptoms attributed to liver disease, requiring the discontinuation of therapy.
Results: 66 subjects were eligible for evaluation. At one year, 8 (12%) of subjects discontinued therapy as a result of drug-related hepatotoxicity, 6 (9%) were lost to follow-up, and 24 (36%) discontinued antiretroviral therapy because of other drug-related toxicity, primarily gastrointestinal symptoms (n=10) and noncompliance (n=5). Eight of 10 individuals discontinuing therapy as a result of gastrointestinal concerns were on Ritonavir containing therapy. Over the entire follow-up period, 12 subjects stopped therapy due to HT. The rate, by Kaplan-Meier analysis, did not differ for Ritonavir inclusion or exclusion, or number of PIs contained in the HAART regimen. Univariate Cox regression analysis did not indicate that sex, co-infection with hepatitis B virus (n=7), baseline CD4 count, baseline transaminase level, number of PI included in therapy, or inclusion of Ritonavir (n=34), Indinavir (n=12), stavudine (n=31), zidovudine (n=27), or lamivudine (n=52) in therapy influenced the risk of hepatotoxicity leading to the discontinuation of HAART. Age was a significant predictor by univariate and multivariate analysis however the reduction in risk per year was minimal.
Conclusion: In our experience, RTV and stavudine-containing ART are no more hepatotoxic than other PI-based therapy. Long term HAART is difficult to maintain in this population.