Factors associated with long-term HIV pre-exposure prophylaxis engagement and adherence among transgender women in Brazil, Mexico and Peru: results from the ImPrEP study.

INTRODUCTION: The HIV epidemic continues to disproportionately impact Latin-American transgender women (TGW). We assessed factors associated with long-term pre-exposure prophylaxis (PrEP) engagement and adherence among TGW enrolled in the Implementation of PrEP (ImPrEP) study, the largest PrEP demonstration study in Latin America. METHODS: HIV-negative TGW aged ≥18 years reporting 1+eligibility criteria in the 6 months prior to enrolment (e.g. sex partner known to be living with HIV, condomless anal sex [CAS], transactional sex or having a sexually transmitted infection [STI]) who could safely take PrEP were enrolled. Follow-up visits were conducted at 4 weeks and then quarterly. We conducted logistic regression to identify factors associated with long-term PrEP engagement (3+ follow-up visits in 52 weeks) and complete self-reported adherence (no missed pills in the past 30 days) during follow-up. For both outcomes, we constructed multivariable models controlling for country, socio-demographics, sexual behaviour, substance use, STIs and self-reported adherence at 4 weeks (long-term engagement outcome only). RESULTS: From March 2018 to June 2021, ImPrEP screened 519 TGW, enrolled 494 (Brazil: 190, Mexico: 66 and Peru: 238) and followed them for 52 weeks. At baseline, 27.5% of TGW were aged 18-24 years, 67.8% were mixed-race and 31.6% had >secondary education. Most, 89.9% reported CAS, 61.9% had >10 sex partners and 71.9% reported transactional sex. HIV incidence was 1.82 cases per 100 person-years (95% confidence interval [CI]: 0.76-4.38). Almost half of TGW (48.6%) had long-term PrEP engagement, which was positively associated with reporting complete adherence at week 4 (aOR:2.94 [95%CI:1.88-4.63]) and was inversely associated with reporting CAS with unknown-HIV partner (aOR:0.52 [95%CI:0.34-0.81]), migration (aOR:0.54 [95%CI:0.34-0.84]), and being from Mexico (aOR:0.28 [95%CI:0.14-0.53]). Self-reported adherence was associated with TGW aged >34 (aOR:1.61 [95%CI:1.10-2.34]) compared to those aged 25-34 and those with >secondary education (aOR:1.55 [95%CI:1.10-2.19]) and was lower among TGW from Peru (aOR:0.29 [95%CI:0.21-0.41]) or reporting PrEP-related adverse effects (aOR:0.63 [95%CI:0.42-0.92]). CONCLUSIONS: Although TGW were willing to enrol in ImPrEP, long-term PrEP engagement and complete self-reported adherence were limited, and HIV incidence remained relatively high. A successful HIV prevention agenda should include trans-specific interventions supporting oral PrEP and exploring long-acting PrEP strategies for TGW.