Background:
In trials, long-acting (LA) injectable antiretroviral therapy (ART) with cabotegravir plus rilpivirine (CAB+RPV) was shown to be non-inferior to oral ART regimens in virologically suppressed (viral load [VL] <50 copies/mL) individuals. We assessed real world effectiveness after a switch to CAB+RPV or a new oral ART regimen.
Methods:
From the OPERA cohort, ART-experienced, suppressed adults with HIV switching to CAB+RPV or a new oral ART regimen between 21Jan2021-31Dec2022 were followed through 30June2023. Confirmed virologic failure (VF; 2 VL ≥200 copies/mL or 1 VL ≥200 copies/mL + regimen change) was assessed among those with ≥1 follow-up VL. Logistic regression models were fit to assess the risk of VF by regimen, adjusted for age (linear & quadratic terms), sex, race, injection drug use (IDU), history of AIDS-defining events (ADE), CD4 count (linear & quadratic terms), comorbid conditions, and prior regimen class. In those receiving CAB+RPV injections, age, sex, race, US region, IDU, history of ADE, CD4 count (per 100 cells/μL), comorbid conditions, prior regimen class, and BMI were evaluated as potential predictors of VF.
Results:
In OPERA, 1362 virologically suppressed adults switched to CAB+RPV injections and 2783 switched to a new oral ART regimen. CAB+RPV users were younger (aged ≥50 years: 29% vs. 41%), had been on their prior regimen for a shorter period (20 vs. 37 months), were more likely to switch from an INSTI (74% vs. 68%), but had similar median CD4 counts at initiation (686 [IQR 496-902] vs. 700 [524-913] cells/μL) compared to oral ART users. Risk of VF out of individuals with a follow-up VL (CAB+RPV: n=1236; oral ART: n=2432) did not statistically differ (adjusted OR: 0.64; 95% CI: 0.41, 1.02). Only baseline CD4 marginally predicted VF; every 100 CD4 cells/μL increase was associated with 15% lower risk of VF (Fig 1). Of the 25 CAB+RPV VF, 40% went to INSTI oral therapy, 40% remained on CAB+RPV, 16% went to multi-core agents, and 1 remained off therapy. Of the 19 with VL, 95% achieved <200 and 79% <50 after VF. In contrast, of the 78 oral VF, 69% stayed on the same ART, 23% went to INSTI regimen, 3 remained off therapy, and the remainder went on a variety of other regimens. Of the 43 with VL, 84% achieved <200 and 72% <50 after VF.
Conclusions:
In routine clinical care in the US, the risk of VF did not differ between virologically suppressed adults switching to CAB+RPV injections or oral ART regimens. Lower CD4 count at initiation was the only predictor of VF with CAB+RPV.
