Abstract
ABSTRACTBackgroundHIV pre-exposure prophylaxis (PrEP) may change serosorting patterns. We examined the influence of serosorting on the population-level HIV transmission impact of PrEP, and how impact could change if PrEP users stopped serosorting.MethodsWe developed a compartmental HIV transmission model parameterized with bio-behavioural and HIV surveillance data among men who have sex with men in Canada. We separately fit the model with serosorting and without serosorting (random partner-selection proportional to availability by HIV-status (sero-proportionate)), and reproduced stable HIV epidemics (2013-2018) with HIV-prevalence 10.3%-24.8%, undiagnosed fraction 4.9%-15.8%, and treatment coverage 82.5%-88.4%. We simulated PrEP-intervention reaching stable coverage by year-1 and compared absolute difference in relative HIV-incidence reduction 10-year post-intervention (PrEP-impact) between: models with serosorting vs. sero-proportionate mixing; and scenarios in which PrEP users stopped vs. continued serosorting. We examined sensitivity of results to PrEP-effectiveness (44%-99%) and coverage (10%-50%).FindingsModels with serosorting predicted a larger PrEP-impact compared with models with sero-proportionate mixing under all PrEP-effectiveness and coverage assumptions (median (inter-quartile-range): 8.1%(5.5%-11.6%)). PrEP users” stopping serosorting reduced PrEP-impact compared with when PrEP users continued serosorting: reductions in PrEP-impact were minimal (2.1%(1.4%-3.4%)) under high PrEP-effectiveness (86%-99%); however, could be considerable (10.9%(8.2%-14.1%)) under low PrEP effectiveness (44%) and high coverage (30%-50%).InterpretationModels assuming sero-proportionate mixing may underestimate population-level HIV-incidence reductions due to PrEP. PrEP-mediated changes in serosorting could lead to programmatically-important reductions in PrEP-impact under low PrEP-effectiveness (e.g. poor adherence/retention). Our findings suggest the need to monitor sexual mixing patterns to inform PrEP implementation and evaluation.FundingCanadian Institutes of Health ResearchRESEARCH IN CONTEXTEvidence before this studyWe searched PubMed for full-text journal articles published between Jan 1, 2010, and Dec 31, 2017, using the MeSH terms “pre-exposure prophylaxis (PrEP)” and “homosexuality, male” and using key words (“pre-exposure prophylaxis” or “preexposure prophylaxis” or “PrEP”) and (“men who have sex with men” or “MSM”) in titles and abstracts. Search results (520 records) were reviewed to identify publications which examined the population-level HIV transmission impact or population-level cost-effectiveness of PrEP in high-income settings. We identified a total of 18 modelling studies of PrEP impact among men who have sex with men (MSM) and four studies were based on the same model with minor variations (thus only the most recent one was included). Among the 15 unique models of PrEP impact, three included serosorting. A total of nine models have assessed the individual-level behaviour change among those on PrEP and its influence on the transmission impact of PrEP. Specifically, the models examined increases in number of partners and reductions in condom use. Most models predicted that realistic increases in partner number or decreases in condom use would not fully offset, but could weaken, PrEP”s impact on reducing HIV transmission. We did not identify any study that examined the influence of serosorting patterns on the estimated transmission impact of PrEP at the population-level, or what could happen to HIV incidence if the use of PrEP changes serosorting patterns.Added value of this studyWe used a mathematical model of HIV transmission to estimate the influence of serosorting and PrEP-mediated changes in serosorting on the transmission impact of PrEP at the population-level among MSM. We found the impact of PrEP was higher under epidemics with serosorting, compared with comparable epidemics simulated assuming sero-proportionate mixing. Under epidemics with serosorting, when PrEP users stopped serosorting (while other men continue to serosort among themselves) we found a reduced PrEP impact compared with scenarios when PrEP users continued to serosort. The magnitude of reduction in PrEP impact was minimal if PrEP-effectiveness was high; however, could be programmatically-meaningful in the context of low PrEP-effectiveness (e.g., poor adherence or retention) and high PrEP coverage. To our knowledge, our study is the first to directly examine the influence of serosorting and PrEP-mediated changes in serosorting on the transmission impact of PrEP and its underlying mechanism.Implications of all the available evidenceOur findings suggest that models which do not consider baseline patterns of serosorting among MSM could potentially underestimate PrEP impact. In addition to monitoring individual-level behavioural change such as condom use, our findings highlight the need to monitor population-level sexual mixing patterns and their changes over time among MSM in the design and evaluation of PrEP implementation.
Publisher
Cold Spring Harbor Laboratory