Abstract Body

Background: Incomplete immune reconstitution (IR) on HAART is possibly mediated by persistent inflammation and is associated with increased all-cause mortality and incidence of AIDS and HIV-associated non-AIDS complications. We explored if statin use was associated with faster IR and greater CD4 gains on HAART. Methodology: We used the US Veterans’ Cohort (1995-2011) to identify patients with ≥1 viral load (VL) and CD4 measurement before and after HAART initiation within the VA system. HAART use and cumulative statin use rate (SUR) were computed from pharmacy refill data and defined as the % of time during the follow-up (f/u) period in which patients were exposed to HAART or any statin, respectively. To examine the association of SUR and time to IR we carried out 3 Cox proportional hazards analyses in subgroups of patients defined by baseline (BL) and threshold CD4 to achieve IR: a) <350 to ≥500; b) <200 to ≥500; and c) <200 to ≥700 cells/μL. We controlled for baseline CD4 and VL, age, race, gender, hepatitis C co-infection, clinical AIDS before HAART, and year of HAART initiation as baseline covariates as well as virologic suppression status and HAART adherence as time-updated covariates. Controlling for the same covariates we then used generalized estimating equations to characterize the contribution of statin use to quarterly average CD4 gains after HAART initiation. Results: A total of 21,132 patients were eligible for analysis. Median f/u time was 7.6 years, median BL CD4 count was 249 cells/μL. 6,957 patients (33%) were exposed to statins after HAART initiation. HAART use rate and SUR were not directly correlated. The table below outlines the proportion of patients achieving IR, the median time to IR and the hazard ratios (HR) for IR for the 3 multivariate Cox models. The contribution of 100% SUR to CD4 gains by generalized estimating equations was + 97 cells/μL (95% CI 89-106, p<0.0001) while 100% HAART use contributed +175 cells/μL (95% CI 169-181, p<0.0001). The following covariates were associated with faster IR or greater CD4 gains in all multivariate models: time-updated virologic suppression status, baseline CD4 count and baseline VL, younger age, female gender, black race, and year of HAART initiation. Conclusions: In our cohort, statin use was associated with substantially greater CD4 gains and faster IR. This effect may be mediated by reductions in persistent inflammation and immune activation and warrants further study.