Hepatotoxicity associated with protease inhibitor-based antiretroviral regimens with or without concurrent ritonavir

OBJECTIVE: To determine the incidence of significant liver enzyme elevations following the initiation of protease inhibitor (PI)-based antiretroviral therapy (ART) with or without pharmacokinetic boosting with ritonavir (RTV), and to define the role of chronic viral hepatitis in its development. DESIGN: Prospective, cohort analysis of 1161 PI-naive, HIV-infected patients receiving RTV-boosted (lopinavir, indinavir and saquinavir) and unboosted PI-based ART (indinavir, nelfinavir) that had at least one liver enzyme measurement before and during therapy. METHODS: The incidence of grade 3 and 4 liver enzyme elevations among persons with and without hepatitis B and/or C co-infection treated with PI-based ART were compared. Severe hepatotoxicity was defined as an increase in serum liver enzyme >/= 5-times the upper limit of the normal range or 3.5-times an elevated baseline level. RESULTS: The incidence of grade 3 or 4 elevations among PI-naive patients was: nelfinavir, 11%; lopinavir/RTV (200 mg/day), 9%; indinavir, 13%; indinavir/RTV (200-400 mg/day), 12.8%; and saquinavir/RTV (800 mg/day), 17.2%. The risk was significantly greater among persons with chronic viral hepatitis (63% of cases); however, the majority of hepatitis C virus (HCV)-infected patients treated with nelfinavir (84%), saquinavir/RTV (74%), indinavir, 86%, indinavir/RTV (90%) or lopinavir/RTV (87%) did not develop hepatotoxicity. CONCLUSIONS: Our data suggest that the lopinavir/RTV is not associated with a significantly increased risk of hepatotoxity among HCV-infected and uninfected patients compared with an alternative PI-based regimen, nelfinavir. Accordingly, other medication-related factors (e.g, efficacy and non-hepatic toxicity) should guide individual treatment decisions.