Background. Liver deterioration has been reported in HIV-Hepatitis C Virus (HCV) co-infected patients starting HAART. We analysed the impact of HAART and immune restoration on liver histological status in HIV-HCV co-infected patients.

Methods. Eligible patients had a HCV RNA load above the limit of detection, no cirrhosis or liver failure, no HCV treatment within the 12 last months (mos). They were enrolled at initiation of HAART (mo 0), and followed at mo 1, 3, 6, 9, and 12, in a prospective multi-centre cohort (TRIVIR-ANRS HC EP1). A liver biopsy was performed at mo 0 and mo 12. Histological progression (HP) was defined as an increase of ³ 2 units in the Knodell score and ³ 1 unit in the Metavir score between mo 0 and mo 12. Potential factors associated with HP were studied: sex, age, body mass index, HIV and HCV transmission category, AIDS, CD4 cell count, HIV and HCV viral load, HCV genotype, alcohol use and type of antiretroviral treatment of HAART at baseline, immunologic response (IR; a double CD4 cell count or a 100/mm³ increase between M0 and M12), onset of transaminases elevation (TE; alanine aminotransferase level ³ 5-fold the upper limit of normal value).

Results. Among the 33 patients (pts) initially enrolled, 25 (21 men) had both liver biopsies: 5 (20%; 95% confidence interval: 4%–36%) had HP. Median age was 38.1 yrs. At baseline, median CD4 cell count was 278/mm³, median levels of HIV and HCV RNA were 4.4 log₁₀ copies/mL and 6.2 IU/mL, respectively. Four (4) pts (16%) were alcohol consumers. Nine (9) pts (36%) were started on protease inhibitor, 14 (56%) on non nucleoside reverse transcriptase inhibitor and 2 (8%) on Abacavir, in addition to 2 nucleoside reverse transcriptase inhibitors. During follow-up, 13 pts (52%) had IR, 19 pts (76%) had a HIV RNA load < 1.70 log₁₀ copies/mL at M12, and there was no change in HCV RNA load. No relationship was found between IR and HP (23% of HP in pts with IR, 17% in pts without IR, p = 1.00) or other baseline characteristics and HP. However, TE occurred in 5 pts (20%) and 1 case was considered drug-related, 2 were alcohol-related and 2 were HCV-related. TE was significantly associated with HP (80% of HP in pts with TE, 5% in pts without TE, p = 0.002).

Conclusion. These data suggest that IR following the initiation of HAART is not associated with a progression of liver lesions. However, careful monitoring of liver enzymes and alcohol suppression are highly recommended in HIV-HCV co-infected pts.