HEPATITIS C IS AN INDEPENDENT PREDICTOR OF MORTALITY AMONG A POPULATION-BASED COHORT OF ANTIRETROVIRAL NAÏVE INDIVIDUALS INITIATING TRIPLE-COMBINATION THERAPY

P Braitstein, V Montessori, JSG Montaner, B Yip, MT Schechter, MV O'Shaughnessy, PR Harrigan, RS Hogg
BC Centre for Excellence in HIV/AIDS, Vancouver, British Columbia

Objective: To determine whether hepatitis C (HCV) is predictive of mortality among individuals initiating HAART in a population-based program.
Methods: The HIV/AIDS Drug Treatment Program distributes, at no cost to eligible individuals, all ART in the province of British Columbia (BC), Canada. Individuals were eligible for this analysis if their first ever ART was two NRTI's plus a PI, or two NRTI's plus a NNRTI between 08/96 and 07/00, and they had documented antibody (Ab) test results for HCV. Adherence was measured by prescription refills. Mortality data were obtained by linking with Vital Statistics of BC, and censored in 02/02. Cumulative mortality rates were estimated using Kaplan-Meier methods, and Wilcoxon Rank-Sum was used to calculate statistical significance. Cox proportional hazards regression was used to calculate adjusted hazard ratios (AHR) and 95% confidence intervals (CI) for death controlling for potential confounders. Bivariate statistics used parametric and non-parametric methods, and are expressed as unadjusted odds ratios (UOR) with 95% CI.
Results: Of the 1416 ART-naïve individuals who initiated a triple-combination, 552 have recorded HCV-Ab test results (39%), of whom 235 (43%) were HCV-antibody positive. At time of censoring, 68 (12.3%) individuals died of non-accidental causes, including 40 (17%) of HCV-pos individuals, and 28 (9%) HCV-neg. The product limit estimate (SE) of the cumulative mortality rate at 48 months for HCV-pos individuals was 21.7% (0.03) and for HCV-neg 10.3% (0.02) (p=0.002). Bivariately associated with non-accidental deaths were baseline CD4 count <50 cells/mm³ (UOR 3.5;95%CI:1.9-6.4), between 50 and 199 cells/mm³ (UOR 2.1;95%CI:1.2-3.6), AIDS diagnosis (UOR 3.0; 95%CI:1.6-5.4), having a physician with >5 HIV-pos patients (UOR 0.48;95%CI:0.27-0.86), >75% adherent to ART (UOR 0.49;95%CI:0.29-0.85), being HCV-Ab pos (UOR 2.1;95%CI:1.3-3.6), having a higher log HIV RNA (5.3vs.5.1,p=0.005), and median age (40 vs.37 yrs, p<0.001). Not associated with non-accidental deaths were gender or history of IDU. After adjustment, HCV seropositivity remained independently predictive of non-accidental mortality (AHR 2.4, 95% CI: 1.3 - 4.3).
Conclusions: Our data suggest that even after controlling for important sociodemographics and clinical characteristics, HCV is predictive of mortality among a population-based cohort of HIV-infected antiretroviral naïve individuals initiating triple-combination ART.

NEXT ABSTRACT >