Trajectories of PrEP use among men who have sex with men: a pooled analysis of two prospective, observational cohort studies.

ABSTRACT

INTRODUCTION: Daily and event-driven oral pre-exposure prophylaxis (PrEP) reduce the risk of HIV acquisition. PrEP use can vary over time, yet little is known about the trajectories of PrEP use irrespective of the chosen PrEP regimens among men who have sex with men (MSM). METHODS: Using data from a mobile, web-based diary application collected daily from 17 August 2015 until 6 May 2018, we analysed PrEP use and sexual behaviour in two large cohorts, AMPrEP (Amsterdam, the Netherlands) and Be-PrEP-ared (Antwerp, Belgium). In both cohorts, participants could choose between daily and event-driven oral PrEP every 3 months. We used group-based trajectory modelling to identify trajectories of PrEP use over time and their determinants. In addition, we estimated the incidence rate of chlamydia, gonorrhoea and syphilis within these trajectories. RESULTS: We included 516 MSM (n = 322 AMPrEP; n = 194 Be-PrEP-ared), of whom 24% chose event-driven PrEP at PrEP initiation. Participants contributed 225,015 days of follow-up (median = 508 days [IQR = 429-511]). Four distinct PrEP use trajectories were identified ≤2 tablets per week ("low frequency," 12% of the total population), 4 tablets per week ("variable," 17%), "almost daily" (31%) and "always daily" (41%). Compared to participants with "low frequency" PrEP use, participants with "variable" (odds ratio [OR] = 2.18, 95% confidence interval [CI] = 1.04-4.60) and "almost daily" PrEP use were more often AMPrEP participants (OR = 2.64, 95% CI = 1.27-5.49). "Almost daily" PrEP users were more often employed (OR = 6.76, 95% CI = 2.10-21.75) and were younger compared to participants with "low frequency" PrEP use. In addition, the number of days on which anal sex occurred was lower among participants with "low frequency" PrEP use compared to the other groups (all p<0.001). Compared to "low frequency" PrEP users, the incidence rates of chlamydia and gonorrhoea were higher for participants with "almost daily" and "always daily" PrEP use. CONCLUSIONS: We uncovered four distinct PrEP use trajectories, pointing to different patterns of PrEP use in practice beyond the two-regimen dichotomy. These trajectories were related to sexual behaviour and rates of sexually transmitted infection. Tailoring PrEP care according to different PrEP use patterns could be an important strategy to improve efficient PrEP delivery.