TY - JOUR T1 - Influence of serosorting and intervention-mediated changes in serosorting on the population-level HIV transmission impact of pre-exposure prophylaxis among men who have sex with men: a mathematical modelling study JF - medRxiv DO - 10.1101/2020.02.26.20025700 SP - 2020.02.26.20025700 AU - Linwei Wang AU - Nasheed Moqueet AU - Anna Simkin AU - Jesse Knight AU - Huiting Ma AU - Nathan J. Lachowsky AU - Heather L. Armstrong AU - Darrell H. S. Tan AU - Ann N. Burchell AU - Trevor A. Hart AU - David M. Moore AU - Barry D. Adam AU - Derek R. MacFadden AU - Stefan Baral AU - Sharmistha Mishra Y1 - 2020/01/01 UR - http://medrxiv.org/content/early/2020/02/26/2020.02.26.20025700.abstract N2 - Background HIV 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.Methods We 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%).Findings Models 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%).Interpretation Models 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.Funding Canadian Institutes of Health ResearchEvidence before this study We 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 study We 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 evidence Our 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.Competing Interest StatementThe authors have declared no competing interest.Funding StatementThis study was funded by the Canadian Institutes of Health Research (CIHR) foundation grant FN-13455. SM and DHST are supported by a CIHR and the Ontario HIV Treatment Network (OHTN) New Investigator Award. TAH is supported by an OHTN Applied HIV Research Chair Award. DMM and NJL are supported by Scholar Awards from the Michael Smith Foundation for Health Research (#5209, #16863). NM was supported by the CIHR-funded Canadian HIV Trials Network Postdoctoral Fellowship. We would like to thank Kristy Yiu for supporting submission and project coordination, and Steven Tingley for helpful discussions surrounding model structure. Some of the model parameters in the current modeling paper drew on estimates published in Wang et al. 2019 (https://doi.org/10.1093/aje/kwz231).We acknowledge the Engage study and its funders (Canadian Institutes of Health Research (CIHR) Team Grant [TE2-138299]; CIHR Canadian HIV Trials Network [CTN 300]; Canadian Foundation for AIDS Research [Engage]; Canadian Blood Services [MSM2017LP-OD]; Ontario HIV Treatment Network (OHTN) [1051]; Ryerson University [no related grant number]; and Public Health Agency of Canada [4500370314]) which supported the independently published results in Wang et al. 2019 (https://doi.org/10.1093/aje/kwz231).Author DeclarationsAll relevant ethical guidelines have been followed; any necessary IRB and/or ethics committee approvals have been obtained and details of the IRB/oversight body are included in the manuscript.YesAll necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.YesI understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).YesI have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.YesAll model code (MATLAB) and parameter values can be found on our GitHub project repository. Data include: fixed parameter values, calibrated parameters: prior ranges and accepted posterior parameter sets, and calibration targets: acceptable ranges and accepted posterior predicted values. Values of model parameters (inputs) and calibration targets (outputs) were estimated from the data sources described in Table S2.1. https://github.com/mishra-lab/prep-serosort ER -