Integrating long-acting injectable cabotegravir for HIV prevention into the Brazilian public health system (ImPrEP CAB Brasil): a hybrid type-2 implementation-effectiveness cohort study.

BACKGROUND: Daily oral pre-exposure prophylaxis (PrEP) requires sustained adherence, which remains challenging in real-world settings. Although long-acting injectable cabotegravir (CAB-LA) could improve prevention coverage, implementation evidence in Latin America is scant. We evaluated the feasibility of integrating same-day CAB-LA delivery within Sistema Único de Saúde (SUS), the Brazilian public health system. METHODS: ImPrEP CAB Brasil was a multicentre, hybrid type-2 implementation-effectiveness, open-label cohort study done at 14 PrEP public health services in six Brazilian cities. Eligible participants were cisgender men, non-binary people assigned male at birth, and transgender men and women who have sex with men, aged 18-30 years, who tested HIV negative, and were PrEP-naive. Participants could choose either CAB-LA (CAB-LA choice cohort) or oral PrEP with tenofovir disoproxil fumarate and emtricitabine (oral PrEP choice cohort), We included a comparison cohort of people initiating PrEP in SUS (oral PrEP SUS cohort). The main outcomes were HIV incidence and follow-up days covered by PrEP. For CAB-LA, we evaluated on-time injection adherence, switching, loss to follow-up, and safety. Participants were given the option to switch PrEP modalities or remain off PrEP during follow-up. Adverse events of special interest in participants using CAB-LA were assessed at every study visit. The study was registered at ClinicalTrials.gov (NCT05515770) and is ongoing. FINDINGS: Participants were enrolled from Oct 17, 2023, to Sept 16, 2024, and followed up until Oct 16, 2025. 3096 participants were included in this analysis: 1201 (38·8%) in the CAB-LA choice cohort, 246 (7·9%) in the oral PrEP choice cohort, and 1649 (53·3%) in the oral PrEP SUS cohort. Most participants were cisgender men (2893 [90·9%]), Pardo or Black (1663 [53·7%]), and completed secondary education or higher (2517 [81·3%]). Median age was 26 years (IQR 23-28). HIV incidence was 0·20 per 100 person-years (95% CI 0·04-0·57) for the CAB-LA cohort (three new HIV infections), 0·87 (0·11-3·16) for the oral PrEP choice cohort (two new HIV infections), and 1·48 (0·88-2·33) for the oral PrEP SUS cohort (18 new HIV infections). 590 091 (90·8%) follow-up days were covered by PrEP in the CAB-LA cohort versus 69 491 (47·3%) in the oral PrEP choice cohort and 328 025 (35·5%) in the oral PrEP SUS cohort (p<0·0001). 107 (8·8%) of 1220 participants were lost to follow-up in the CAB-LA cohort, and were mainly people who were non-White, had lower levels of education, and had moderate or severe depressive symptoms. On-time injection adherence was high (92%). Injection-site reactions were the most frequent adverse event (4212 events in 1067 [87·5%] participants) and were generally mild to moderate. 80 serious adverse events were reported: 43 adverse events of interest (one related to CAB-LA) and 37 other events (none related to CAB-LA). INTERPRETATION: Same-day CAB-LA delivery effectively reduced HIV incidence and increased PrEP coverage, supporting the implementation of long-acting options within SUS in Brazil. FUNDING: Unitaid, WHO, and Ministry of Health of Brazil. TRANSLATIONS: For the Portuguese and Spanish translations of the abstract see Supplementary Materials section.