Background:  There are few data about factors associated to survival and hepatic decompensation in HIV/hepatitis C virus (HCV)-co-infected patients with liver cirrhosis.

Methods:  In 331 HIV/HCV-co-infected patients with cirrhosis we compared the association of survival (defined as time to death, hepatocarcinoma, or liver transplant) with: age, sex, time since cirrhosis or HIV diagnosis, alcohol use, CD4 (nadir, baseline and <100 at baseline), HIV viremia, suppressed HIV replication, history of anti-HCV treatment, HCV genotype, sustained viral response to anti-HCV treatment, concomitant chronic hepatitis B, history of cirrhosis decompensation, Child-Pugh score (CPS) and HAART (at baseline, continuous/interrupted during follow-up). For the 251 patients without prior hepatic decompensation we analyzed variables associated with the occurrence of first decompensation. Univariate/Multivariate Cox proportional hazard models were used.

Results:  Subjects’ characteristics were:  male 78%, median age 44, median follow-up 18 months, 87% on HAART at baseline (50% received un-interrupted HAART during follow-up), median CD4 384, 74% with undetectable HIV RNA, median time since HIV/cirrhosis diagnosis 16/3 years, 6% had concomitant chronic HBV, 27% HCV genotype 2/3, 30% had history or excessive alcohol intake, 58% were receiving anti-HCV treatment (27% with sustained virological reaction). During follow-up 62 endpoints occurred. Factors significantly associated to decreased survival (univariate) were:  male gender, alcohol abuse, CD4 <100 (baseline), unsuppressed HIV replication (baseline), not having received anti-HCV treatment, no sustained virological reaction to anti-HCV treatment, history of decompensation, CPS-B and C, not receiving HAART at baseline and non-continuous HAART during follow-up. In the multivariate analysis 5 independent factors were associated with decreased survival: CD4 nadir (p = 0.014; HR 1003); CPS-B (p = 0.0001; HR 10.1); CPS-C (p = 0.0001; HR 19); unsupressed HIV replication at baseline (p = 0.045; HR 1.9) and not continuous HAART during follow-up (p = 0.001; HR 11.3). For patients with compensated cirrhosis at baseline the only variable associated with development of decompensation was a CPS-B (HR 7.4, p = 0.0001, 95%CI 2.5 to 22.2).

Conclusions:  CPS-B and C are significantly associated with decreased survival in HIV/HCV-co-infected patients with cirrhosis. Maintaining HIV viral suppression and continuous HAART use are associated with prolonged survival. CPS-B is significantly associated with the short-term risk of first hepatic decompensation.