CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

scenario, which achieved a 23.5% (95% simulation interval: 10.9, 34.3) reduction in new infections at 10% population coverage over 10 years. Conclusion: Under most of the scenarios tested, PrEP coverage of 15%was sufficient to reduce cumulative incidence by 25% over 10 years. Focused implementation efforts on MSM who have larger numbers of partners may maximize PrEP’s impact and possibly its cost-effectiveness at lower levels of population coverage.

benefit of LA-PrEP will depend on the efficacy observed in ongoing phase III trials, as well as the extent and duration of protection among persons who drop out of LA-PrEP care.

1152 QUANTIFYING INDIRECT BENEFITS OF PrEP: MODELING ANALYSIS OF ORAL PrEP IN ZIMBABWE Zindoga Mukandavire 1 , Daniel J. Klein 2 , Graham F. Medley 1 , Anna Bershteyn 2 , Katharine Kripke 3 , Delivette Castor 4 , Definate Nhamo 5 , Joseph Murungu 5 , Taurai Bhatasara 6 , Lawrance Nyazema 6 , Kudakwashe Takarinda 6 , Kristine Torjesen 7 1 London School of Hygiene & Tropical Medicine, London, UK, 2 Institute for Disease Modeling, Bellevue, WA, USA, 3 Avenir Health, Washington, DC, USA, 4 United States Agency for International Development, Washington, DC, USA, 5 Pangaea Zimbabwe Aids Trust, Harare, Zimbabwe, 6 Zimbabwe Ministry of Health and Child Care, Harare, Zimbabwe, 7 FHI 360, Durham, NC, USA Background: HIV incidence remains high among adolescent girls and young women (AGYW) in Zimbabwe. Tenofovir-based oral pre-exposure prophylaxis (PrEP) can prevent individual infection, but population level impact and disaggregation of the direct (primary) and indirect (secondary/onward) preventive effects remain unexplored for this population. We assessed potential population-level benefits of oral PrEP use among female sex workers (FSW) and AGYW, and parse direct from indirect effects. Methods: An individual-based HIV network model, EMOD-HIV v2.5, was parameterized for Zimbabwe to simulate provincial-level epidemics. The model was fit to age/gender/province-specific data on HIV prevalence and treatment coverage from DHS and Zimbabwe MoHCC reports. In the simulation, oral PrEP was provided to FSW and/or AGYW (18-24 years) with multiple partners at 40% coverage beginning in 2017 with a 5-year regimen of 73% effectiveness, which represents 90% efficacy with 81% adherence. Direct and indirect infections averted by oral PrEP were estimated by simulating a randomized trial (RT) in which 40% of the target population received oral PrEP and 40% received placebo drug. Direct prevention was computed from the incidence rate ratio between the two arms. Indirect prevention was computed as the difference between the number of HIV infections in the RT simulation and the number of HIV infections in a separate PrEP-free counterfactual simulation. Results: For every infection directly averted by providing oral PrEP to FSW, 1.3 additional infections were indirectly averted in the community over 5 years and 1.7 additional infections were prevented in 20 years. The ratios were 1 and 1.7 when targeting only AGYWwith multiple partners over the same horizons. Total numbers of indirect infections averted by oral PrEP increased in all ten provinces when providing PrEP to both FSW and AGYW, compared to the scenario of providing oral PrEP to FSW alone. In this scenario, the ratio of indirect to direct infections averted was 0.9 over 5 years and 1.8 over 20 years. Conclusion: Results suggest that community benefit of oral PrEP can outweigh direct individual-level benefit. Indirect benefits of PrEP should be considered when prioritizing populations for PrEP service provision.

Poster Abstracts

1151 POTENTIAL EFFECTIVENESS OF LONG-ACTING INJECTABLE PrEP IN MSM: A MODELING STUDY Brandon D. Marshall 1 , William C. Goedel 1 , Maximilian King 1 , David P. Durham 2 , Philip A. Chan 1 , Jeffrey P. Townsend 2 , Alison P. Galvani 2 1 Brown University, Providence, RI, USA, 2 Yale University, New Haven, CT, USA Background: Oral pre-exposure prophylaxis (PrEP) is highly effective for HIV prevention in men who have sex with men (MSM). To address challenges associated with daily pill adherence, injectable formulations are being tested in phase III trials. Long-acting cabotegravir (LA-CAB), an integrase strand transfer inhibitor, was demonstrated to have a strong protective effect in animal models. We used an agent-based model (ABM) to estimate the effectiveness of LA-CAB compared to oral PrEP on HIV incidence among MSM. Methods: The ABM simulated HIV transmission in a dynamic sexual network of 11,000 MSM in Atlanta, Georgia over a 10-year time period (2015-2025). The model was calibrated to reproduce current HIV prevalence and incidence among MSM in Atlanta (30.3% and 3.9 per 100 person-years). We assumed agents received bimonthly LA-CAB bimonthly injections, with retention rates based on phase II trial data. The theoretical efficacy of LA-CAB was estimated from published macaque data using logistic regression, with waning protection up to 12 months following a final injection. LA-CAB concentrations over time were estimated from human pharmacokinetic data. The impact of LA-CAB on HIV incidence compared to oral PrEP was investigated across different PrEP target coverage levels. Sensitivity analyses investigated varied efficacies, retention rates, and half-lifes of LA-CAB. Results: We estimated a theoretical efficacy of >99% among agents receiving bimonthly LA-CAB injections. Over the 10-year simulation period, HIV incidence was reduced in scenarios in which agents received LA-CAB compared to scenarios in which all received oral PrEP, at every coverage level (Figure). For example, a scenario in which 35% of HIV-uninfected agents receive LA-CAB averted an additional 255 infections over the next decade compared to a scenario in which 35% receive oral PrEP. The relative benefit of LA-CAB was sensitive to the maximum efficacy of bi-monthly LA-CAB injections and the terminal half-life after a final injection. For any given coverage level, decreasing the LA-CAB retention rate increased relative the benefit of the injectable formulation, due to a greater pool of agents with partial protection. Conclusion: Long-acting PrEP may be more effective than oral PrEP for preventing HIV acquisition in MSM. However, the population impact and relative

CROI 2018 446