Background
Integrase strand transfer inhibitor (INSTI)-based antiretroviral therapy (ART) has been associated with disproportionate weight gain among people with HIV (PWH), primarily in treatment-naïve populations. Few studies have explored weight and cardiometabolic outcomes beyond two years and even fewer have examined effects on major adverse cardiovascular events (MACE).
Methods
Using target trial emulation, we estimated the causal effect of switching to an INSTI-based regimen, versus remaining on a non-INSTI-based regimen, on incident obesity, diabetes, hypertension, metabolic syndrome and MACE in an ART-experienced population in the REPRIEVE cardiovascular primary prevention trial. We designed a series of sequential target trials and used inverse probability of treatment weights estimated at the start of each trial to adjust for key confounders including sex, age, race, enrollment region, ASCVD risk score, statin use, ART duration and BMI. Participants not on an INSTI at REPRIEVE entry were eligible and followed until the event of interest, loss to follow-up, death (competing event) or censoring at 5 years. Hazard ratios were estimated using weighted cox-proportional cause-specific hazard models, with data pooled across all trials (N=73) and 95% confidence intervals (CI) obtained via non-parametric bootstrapping (500 samples).
Results
Among 5174 participants eligible for at least one trial, 2686 (52%) switched to an INSTI-based regimen, predominantly dolutegravir (82%). Average follow-up time was similar among switchers and non-switchers (3.8 years) and average time to event following a switch to an INSTI ranged from 3.5 – 4 years depending on the outcome. Among those who switched, 7% developed obesity, 4% diabetes, 6% hypertension, 12% metabolic syndrome and 2% experienced a MACE event. Comparing those who switched to an INSTI to those who did not, we estimated an increased hazard of developing obesity (HR: 1.34, 95% CI: 1.14-1.52), diabetes (HR: 1.41, 1.13-1.70), hypertension (HR: 1.44, 1.17-1.65), and metabolic syndrome (HR: 1.18, 1.04-1.33) but not MACE (HR: 1.06, 0.67-1.53). Adjustment for NRTI choice did not meaningfully impact results.
Conclusions
Leveraging a longitudinal global cohort of PWH, we found an increased risk of weight gain and cardiometabolic complications following a switch to an INSTI over 5 years of follow-up. Given the observed elevated risk profile, long-term observation of PWH switching to INSTIs will be critical to assess for future development of MACE.