806    

Background:  GBV-C co-infection has been associated with delayed progression of HIV infection. We studied the effect of GBV-C co-infection on HIV-1 disease progression in a prospective cohort of homosexual men and controlled, besides prognostic factors at baseline, for CD4 counts during follow-up.

Methods:  Since 1984, 326 homosexual men with an imputed date of HIV seroconversion enrolled in the Amsterdam Cohort Studies with 3 to 6 monthly follow-up visits. The first sample available after seroconversion and the last sample before 1996 were tested for GBV-C RNA and anti-E2 antibodies. Censor date was December 31, 2002. The effect of GBV-C RNA or anti-E2 antibody on AIDS and death were studied using  proportional hazards models. Hazard ratios (HR) were adjusted for age at seroconversion, CCR5Δ32 genotype, HAART as calendar time (period before vs after 1996), HIV RNA and CD4 count at 1 year after seroconversion. GBV-C RNA and E2-ab were also studied as time-dependent variables and corrected for 6-month CD4 counts during follow-up.

Results:  Median follow-up time was 8.9 years. Median time of first GBV-C test after seroconversion was 1.9 years (IQR 1.9 years) and between first and last test 7.4 years. In the first sample, 137 (42%) tested positive for GBV-C RNA and 134 (41%) had detectable anti-E2-antibody. Ten subjects gained and 79 lost GBV-C RNA during follow-up. Participants with GBV-C RNA at first test were younger, had lower CD4 counts 1 year after seroconversion and had an increased risk of developing AIDS (HR_adj 1.4, 95%CI:  1.1 to 1.9). After adjusting for CD4 counts during follow-up HR_adj became 1.0 (95%CI:  0.7 to 1.3). Participants who lost GBV-C RNA had a significantly higher risk of AIDS compared with participants with continuous absence of GBV-C RNA during follow-up (HR_adj 2.8, 95%CI:  2.0 to 4.0). After correcting for CD4 counts during follow-up, HR_adj decreased to 1.6 (95% CI:  1.1 to 1.2). Anti-E2-antibody or GBV-C RNA persistence, gain or continuous absence had no effect on HIV progression. Analysis for survival time to death produced similar results. CCR5 32 genotype and HAART did not influence the effect of GBV-C on HIV progression.

Conclusions:  Rather than a positive effect of GBV-C RNA on HIV progression our study shows that loss of GBV-C RNA has a negative effect on progression to AIDS or death. This effect of GBV-C RNA seems to be associated with changes in CD4 count during follow-up. Hence, we hypothesize that the persistence of GBV-C might depend on sufficient CD4 cell numbers and that the CD4 cell decline associated with HIV progression rather is a cause than a consequence of absent GBV-C RNA.

Keywords: GBV-C co-infection; HIV disease progression; seroconverters