Abstract
Background HIV pre-exposure prophylaxis (PrEP) data in men who have sex with men (MSM) in west Africa are essential to guide its large-scale implementation. We assessed the uptake of event-driven and daily PrEP, HIV incidence, and changes over time in sexual behaviours and prevalence of bacterial sexually transmitted infections (STIs) in MSM in Burkina Faso, Côte d'Ivoire, Mali, and Togo. Methods We did a prospective cohort study from Nov 20, 2017, to April 14, 2020, in four community-based clinics in Abidjan (Côte d'Ivoire), Bamako (Mali), Lomé (Togo), and Ouagadougou (Burkina Faso). Participants were MSM aged 18 years or older at substantial risk of HIV infection. Participants could choose between event-driven (2+1+1 dosing) and daily oral PrEP (tenofovir disoproxil fumarate 300 mg plus emtricitabine 200 mg), switch regimen, and discontinue or restart PrEP. We compared HIV incidence in this study with that of the same cohort before the availability of PrEP (CohMSM). Statistical analysis included the Kaplan-Meier method and mixed-effects regression models. This study is registered with ClinicalTrials.gov, NCT03459157. Findings We followed up 598 participants for a total of 743•6 person-years. At enrolment, 445 (74%) of 598 participants chose event-driven PrEP and 153 (26%) of 598 chose daily PrEP. 60 (13%) of 445 and 65 (42%) of 153 partici pants switched PrEP regimen at least once (p<0•0001). 159 participants (27%) were lost to follow-up. Overall HIV incidence was 2•3 per 100 person-years (95% CI 1•3-3•7; adjusted incidence rate ratio 0•21, 95% CI 0•12-0•36 compared with CohMSM). Adherence was optimal in 802 (41%) of 1946 measures with event-driven PrEP and in 394 (71%) of 554 measures with daily PrEP (p<0•0001). Coverage of sex acts with PrEP only and PrEP and condom decreased during follow-up (p=0•039 if PrEP only; p=0•0025 if PrEP and condom). The frequency of condomless anal sex remained stable (p=0•96). The number of male sexual partners (p<0•0001) and number of sex acts with casual male partners (p=0•0014 for 1-4 sex acts in previous 4 weeks; p=0•030 for ≥5 sex acts) decreased. The prevalence of gonorrhoea, chlamydia, and syphilis remained stable. Interpretation PrEP availability helped prevent HIV infection and did not lead to an increase in risky sexual behaviours or other STIs. PrEP should be urgently implemented in west Africa. Retention in care and PrEP adherence require special attention to ensure PrEP reaches its full prevention potential. Funding ANRS and Expertise France.