Impact of Gluteal Injection Site on Pharmacokinetics and Tolerability of Antiretroviral Therapy with Long-Acting Cabotegravir and Rilpivirine

Abstract

The anatomical site of gluteal administration may influence the pharmacokinetics and tolerability of long-acting cabotegravir (CAB) plus rilpivirine (RPV), but direct comparisons between dorsogluteal (DG) and ventrogluteal (VG) injections are limited. In this prospective intra-individual crossover study, 152 adults with HIV receiving injectable CAB and RPV were switched from DG to VG administration. A total of 842 trough and 131 one-month postinjection plasma samples were analyzed. VG delivery was associated with higher one-month post-dose concentrations for both drugs (median CAB 1280 vs. 834 ng/mL, p < 0.001; RPV 148 vs. 126 ng/mL, p = 0.043). At trough, CAB concentrations were lower with VG injections (573 vs 664 ng/mL, p < 0.001), whereas RPV levels did not differ significantly (131 vs 123 ng/mL, p = 0.461). Although uncommon, a higher proportion (6.9% vs. 1.5%, p = 0.001) of CAB trough samples fell below the protein-adjusted inhibitory concentration 90 threshold in the VG group. VG administration resulted in lower variability in pooled CAB and RPV trough concentrations, both intra-patient (standard deviation ratio 0.92; p = 0.041) and inter-patient (0.80; p = 0.001), compared with DG injections. Detectable HIV-1 RNA (≥50 copies/mL) was rare and comparable between injection sites, with no confirmed virologic failures. Participants reported better tolerability with VG injections, which were associated with fewer local symptoms and were preferred by 67% of respondents. Ventrogluteal administration is associated with more consistent pharmacokinetics and improved tolerability, supporting its use as the preferred site for long-acting CAB/RPV. Although CAB trough levels were modestly lower and occasionally fell below predefined thresholds with VG injection, this did not impact virologic outcomes, reinforcing its clinical suitability in most settings.