800    

Background: Since the advent of highly active antiretroviral therapy (HAART) for patients with human immunodeficiency virus (HIV), several studies have investigated the effect of hepatitis C virus (HCV) infection on HIV progression, but with conflicting results.  Few studies had sufficient events to consider mortality alone and some used data from a single institution. We estimated the effect of HCV on survival using the large cohort of HIV patients cared for at over 100 US Veterans Affairs (VA) facilities and controlling for numerous potentially confounding factors.

Methods:  The study cohort consisted of veterans on the VA National Immunology Case Registry (ICR) who met the following criteria: first VA prescription for a HAART-defining potent antiretroviral medication between January 1997 and February 2003, VA results for a CD4 count and HIV viral load in the year prior to that first prescription and VA results for a HCV antibody test at any time.  We used the Cox proportional hazards model to study the effect of HCV serostatus on survival from the initiation of HAART.  The model included patient demographic characteristics, CD4 cell count and HIV viral load, psychiatric and substance abuse diagnoses, history of AIDS-defining opportunistic infection, receipt of pre-HAART antiretroviral therapy, receipt of HCV medication, facility size and region, and calendar year.  For a subset of patients, we also controlled for history of injection drug use from reported CDC exposure category.

Results:  Of the 12,216 patients in the study cohort, 4668 (38%) were HCV seropositive.  The mean observation time was 3.5 years.  2,087 deaths were reported through March 2003.  The unadjusted death rate for HCV seropositive patients was 6.4 deaths per 100 person years compared with 4.0 for HCV seronegative patients (p<0.0001).  In the main multivariate Cox model, the hazard ratio for death for HCV seropositive patients relative to HCV seronegative patients was 1.52 (95% CI 1.37 to 1.68, p <0.0001).  In the model also controlling for injection drug use, the comparable hazard ratio was 1.28 (95% CI 1.13 to 1.45, p0.0001).

Conclusions:  Controlling for numerous potentially confounding factors, HCV seropositivity was independently associated with increased risk of death in a large cohort of HAART-treated HIV-infected veterans. Given the success of HAART in extending the lives of HIV patients, HCV has become an important predictor of mortality in the HIV-infected population.

Keywords: Hepatitis C Virus; Mortality; Survival Analysis