Lucilene Araujo de Freitas, Carolina Coutinho, Débora Castanheira, Ronaldo Ismério, Pedro Leite, Iuri Costa Leite, Marcelo Cunha, Josias Freitas, Toni Araújo, Laylla Monteiro, Brenda Hoagland, Mayara S.T. Silva, Marcos Benedetti, Cristina Pimenta, Beatriz Grinsztejn, Valdilea G. Veloso, Thiago S. Torres
INTRODUCTION: Access to health services may affect PrEP uptake, adherence, and persistance, especially among persons affected by structural inequities. We assessed the HIV incidence and factors associated with PrEP non-adherence among men who have sex with men (MSM) and transgender women in Brazil, focusing on potential racial disparities. MATERIAL AND METHODS: ImPrEP was a prospective, single-arm, open-label, PrEP implementation study that enrolled 9509 MSM and transgender women in Brazil, Mexico and Peru (February 2018 ‒ June 2021). Participants received oral PrEP with tenofovir disoproxil fumarate and emtricitabine at the enrollment visit and quarterly thereafter. This analysis was restricted to data from 14 HIV/STI clinics in 11 cities from Brazil. We calculated the HIV incidence per 100 person-years using the Poisson model according to race. We used adjusted logistic regression models to identify factors associated with PrEP non-adherence (medication possession rate < 0.6) for White and Black/Pardo participants separately. TRIAL REGISTRATION: UTN U1111-1217-6021 RESULTS: Of 3928 participants, 1868 (47.6%) self-identified as White, 1410 (35.9%) Pardo, 595 (15.1%) Black, 42 (1.1%) Asian, and 12 (0.3%) Indigenous. PrEP non-adherence was higher among Black (26.2%) and Pardo (24.2%) compared to White (18.7%) participants (p < 0.0001). Among participants with PrEP non-adherence, HIV incidence was higher among Black (2.16 [95% CI: 0.54‒8.63]) than Pardo (1.49 [95% CI: 0.48‒4.62]) and White (1.00 [95% CI 0.25‒4.01]). In multivariable models, participants aged 18‒30 years, self-identifying as transgender women, and reporting lower number of sex partners had increased odds of PrEP non-adherence among White and Black/Pardo participants. Lower education and transactional sex were associated with increased odds of PrEP non-adherence among Black/Pardo, but not among White participants. CONCLUSIONS: Higher HIV incidence and PrEP non-adherence among racially and ethnically marginalized groups, such as Black and Pardo Brazilians, highlights the impact of structural racism on health outcomes. The implementation of public policies to reduce racial and social inequities in HIV prevention must be prioritized in Brazil.