BACKGROUND:

Highly active antiretroviral therapy (HAART) has improved the prognosis of HIV infection. However, replication of hepatitis C virus (HCV) is not inhibited by HAART, and treatment-related hepatotoxicity is common. To clarify the effect of HAART in HIV/HCV-coinfected patients, we studied liver-related mortality and overall mortality in 285 patients who were regularly treated during the period 1990-2002 at our department.

METHODS:

Survival was analysed retrospectively by Kaplan-Meier and Cox's regression analyses after patients (81% haemophiliacs) had been stratified into three groups according to their antiretroviral therapy (HAART n=93, available after 1995; treatment exclusively with nucleoside analogues n=55, available after 1992; or no treatment, n=137).

FINDINGS:

Liver-related mortality rates were 0.45, 0.69, and 1.70 per 100 person-years in the HAART, antiretroviral-treatment, and untreated groups. Kaplan-Meier analysis of liver-related mortality confirmed the significant survival benefit in patients with antiretroviral therapy (p=0.018), and regression analysis identified HAART (odds ratio 0.106 [95% CI 0.020-0.564]), antiretroviral treatment (0.283 [0.103-0.780]), CD4-positive T-cell count (0.746 [0.641-0.868] per 0.05x10(9) cells/L), serum cholinesterase (0.962 [0.938-0.986] per 100 U/L), and age (1.065 [1.027-1.105] per year) as independent predictors of liver-related survival. Severe drug-related hepatotoxicity was seen in five patients treated with nucleoside analogues alone and 13 treated with HAART. No patient died from drug-related hepatotoxicity.

INTERPRETATION:

In addition to improved overall survival, antiretroviral therapy significantly reduced long-term liver-related mortality in our patients. This survival benefit seems to outweigh by far the associated risks of severe hepatotoxicity.