Background:  Hepatotoxicity after initiation of cART has been described in HIV/hepatitis C virus (HCV) patients and has been associated with use of protease inhibitors (PI) and nevirapine (NVP). This risk may be diminished in the modern cART era, and we evaluated the association between period of cART initiation, HCV status, and subsequent hepatotoxicity.

Methods:  Data were analyzed from 748 naïve individuals enrolled in the British Columbia Drug Treatment Program who initiated cART between 1997 to 2009 with ≥12 months of laboratory follow up, and eligible with baseline normal alanine aminotransferase (ALT). Three periods of initiation were defined:  Period 1, 1 January 1997 to 31 December 1999; Period 2, 1 January 2000 to 31 December 2003; and Period 3, 1 January 2004 to 31 December 2009. Hepatotoxicity was defined as an ALT >5x baseline ALT. Incidence rates of hepatotoxicity were determined for each time period according to HCV status. Poisson regression estimated the factors associated with hepatotoxicity, while adjusting for age, gender, baseline intravenous drug use (IDU) status, alcohol use, HCV status, cART at time of hepatotoxicity (NNRTI vs PI), number of regimen switches within first year, and period of cART initiation.

Results:  Overall, 91% were male, 24% were IDU, 26% HCV⁺ with median age 42 (interquartile range [IQR] 36 to 49). There were 107 (14%) individuals who initiated cART in Period 1, 208 (28%) in Period 2, and 433 (58%) in Period 3. Median baseline ALT was 33 IU/L (IQR 26 to 43), and time to hepatotoxicity was 10 months (IQR 9 to 11). Incidence of hepatotoxicity was 38/100 person-years (PY), 30/100 PY, 25/100 PY in Periods 1 to 3 for HCV⁺ compared to 22/100 PY, 0.91/100 PY, and 2/100 PY for HCV^– (p <0.001). Baseline HCV status (adjusted relative risk [aRR] 12.01; 95%CI 6.43 to 22.46), period of initiation (aRR 2.56; 95%CI 1.15 to 5.69, Period 1 vs Period 3, aRR 0.73; 95%CI 0.45 to 1.19, Period 2 vs Period 3), IDU (aRR 1.63, 95%CI 1.01 to 2.65), and cART regimen (1.64; 95%CI 1.11 to 2.44 for NNRTI vs PI) were associated with hepatotoxicity in Poisson regression analysis. Alcohol use (aRR 0.73, 95%CI 0.50 to 1.08) and regimen switch in first year of therapy (aRR 1.11; 95%CI 0.85 to 1.44) were not associated with hepatotoxicity. In sub-analysis the incidence of hepatotoxicity on PI cART was 7/100 PY, on NVP was 16/100 PY vs 9/100 PY for efavirenz (p 0.57 and 0.07, respectively).

Conclusions:  Incidence of hepatotoxicity has diminished in the modern cART era, but remains significantly higher in those with HCV co-infection. Early HCV therapy may serve to decrease the risk of hepatotoxicity following cART initiation.