BACKGROUND

In the REPRIEVE trial of statin therapy in people with HIV, pitavastatin reduced major adverse cardiovascular events (MACE) among those with low-to-moderate risk of cardiovascular disease (CVD). We aimed to investigate associations between former and current use of antiretroviral therapy (ART) on entry into the REPRIEVE trial and the development of MACE.

METHODS

This longitudinal cohort analysis was a prespecified secondary analysis of the REPRIEVE trial, a double-blind, placebo-controlled, multicentre, phase 3 randomised trial conducted at 137 sites in 12 countries. REPRIEVE enrolled people with HIV aged 40-75 years, currently on ART, with a CD4 count of more than 100 cells per μL and low-to-moderate CVD risk, and randomly assigned them to receive pitavastatin or placebo. For this secondary analysis, participants' history of ART use, including lifetime exposure to selected agents, was collected at baseline. The primary outcome of interest was time-to-first MACE. Stratified Cox proportional hazards models were used to estimate the relative hazards of MACE associated with ART exposures. Effects of no previous exposure, former exposure, and current exposure to ART at entry to the study were compared using models unadjusted and adjusted for entry risk factors and ART regimen at entry. All analyses were conducted in the intention-to-treat population. The REPRIEVE trial is registered at ClinicalTrials.gov, NCT02344290, and is complete.

FINDINGS

Between March 26, 2015, and July 31, 2019, 7769 participants were enrolled into the REPRIEVE trial. 2419 (31·1%) of 7769 participants were assigned female at birth and 5350 (68·9%) were assigned male at birth. 3208 (41·3%) of 7769 participants were Black or African American, 2704 (34·8%) were White, 1138 (14·6%) were Asian, and 719 (9·3%) were of other races. Participants had a median age of 50·0 years (IQR 45·0-55·0), LDL cholesterol concentration of 106 mg/dL (86-128), 10-year atherosclerotic cardiovascular disease risk score of 4·5% (2·1-7·0), and CD4 cell count of 621 cells per μL (448-827). 5867 (97·8%) of 5997 participants for whom data on this measure were available had an HIV-1 viral load of less than 400 copies per mL. The median duration of ART use at entry was 9·6 years (5·3-14·8). Overall, 1702 (21·9%) of 7769 participants reported previous exposure to abacavir, 6681 (86·0%) to tenofovir disoproxil fumarate, 3832 (49·3%) to thymidine analogues (zidovudine or stavudine), and 3683 (47·4%) to protease inhibitors. At study entry, 984 (12·6%) participants were using abacavir, 4743 (61·0%) were using tenofovir disoproxil fumarate, 756 (9·7%) were using thymidine analogues, and 1990 (25·6%) were using protease inhibitors. In adjusted analyses, former exposure (hazard ratio 1·62, 95% CI 1·14-2·30) and current exposure (1·41, 1·01-1·96) to abacavir was associated with a higher hazard of MACE than in participants who were never exposed. Associations between former or current exposure to other ART agents and MACE were not consistently apparent.

INTERPRETATION

Previous and current use of abacavir increases the hazard of MACE among people with HIV at low-to-moderate CVD risk, suggesting that abacavir should be avoided and previous exposure considered when assessing the risk of MACE in this population.

FUNDING

National Institutes of Health, Kowa Pharmaceuticals America, Gilead Sciences, and ViiV Healthcare.