Background: The risk of liver enzyme elevation (LEE) after different ritonavir-boosted protease inhibitors (PI/r) has not been fully assessed in real-life settings and in populations with high rates of hepatitis C virus (HCV) coinfection.
Methods: Patients introducing a new PI/r between 1998 and 2012 were included, if transaminases and HCV antibody (Ab) were assessed before treatment initiation. Time to grade 3 and 4 LEE were assessed using univariable and multivariable conditional Cox analyses, stratified by HCV serostatus.
Results: A total of 6193 HIV-infected patients (3242 HCV-Ab negative and 2951 HCV-Ab positive) were included. Incidence of grade 3 LEE was 1.05, 7.66, and 8.08 per 100 patient-years of follow-up among HCV-Ab negative, HCV-Ab-positive and HCV-RNA-positive patients, respectively. Among HCV-Ab-negative patients, no differences were detected between different PI/r. Use of darunavir/ritonavir was not associated with LEE among HCV-coinfected patients. Atazanavir/ritonavir use was associated with grade 3 LEE but only among HCV-Ab-positive patients (versus LPV/r, hazard ratio: 1.39; 95% confidence interval: 1.1 to 1.75). This risk was not confirmed in a subanalysis restricted to HCV-RNA-positive patients (versus LPV/r, hazard ratio: 1.16; 95% confidence interval: 0.87 to 1.55). Other independent predictors of grade 3 LEE among HCV-Ab-positive patients were older age, male gender, being treatment naive, nonnucleoside reverse transcriptase inhibitor coadministration, increased aspartate aminotransferase at baseline, overweight, positive HCV-RNA, and advanced estimated liver fibrosis.
Conclusions: Occurrence of hepatotoxicity was a rare finding among HCV-Ab-negative patients and was not influenced by the type of PI/r. In particular, the use of darunavir/ritonavir, previously linked with severe cases of hepatotoxicity, was not associated with a greater risk of LEE, irrespective from HCV serostatus.