Effectiveness and predictors of treatment discontinuation of long-acting cabotegravir/rilpivirine in virologically suppressed people with HIV: real-life data from the Icona Cohort

Background: Phase 3 studies have shown long-acting (LA) cabotegravir/rilpivirine to be effective and tolerable as maintenance therapy in people with HIV (PWH). However, real-life data on their effectiveness are limited. Methods: All PWH enrolled in the Icona Cohort who started LA cabotegravir/rilpivirine with HIV-RNA < 50 copies/mL were included. Times to treatment discontinuation (TD) and to virological failure (VF50, two consecutive HIV-RNA >50 copies/mL or one >1000 copies/mL followed by ART switch) were estimated by the Kaplan–Meier method. Cox regression models, adjusted for age, sex and mode of HIV transmission and stratified by the centre, were employed. Results: Overall, 583 PWH started LA cabotegravir/rilpivirine. Six VF50 were observed, with a 1 year estimated cumulative probability of virological failure of 1.2% (95% CI, 0.5%–3.0%). Resistance-associated mutations for rilpivirine and cabotegravir were detected in 3/4 and 4/4 participants with VF50, respectively, for which the genotypic resistance test was performed. The 1 year cumulative probability of TD was 11.4% (95% CI, 8.6%–14.9%), mainly caused by toxicity/adverse events (73.2%). Multivariable analysis identified heterosexual intercourse and IV drug use as significant risk factors for TD compared with MSM. Conclusions: This analysis demonstrated the short-term effectiveness of cabotegravir/rilpivirine in a real-life setting showing minimal incidence of virological failure but a notable probability of discontinuation due to toxicity or adverse events.