CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

849 METRICS OF MOBILITY BY SEX ARE ASSOCIATED WITH HIV INCIDENCE IN RURAL KENYA & UGANDA Carol S. Camlin 1 , Edwin D. Charlebois 1 , Maya L. Petersen 2 , Laura B. Balzer 3 , Torsten B. Neilands 1 , Dalsone Kwarisiima 4 , James Ayieko 5 , Gabriel Chamie 1 , Craig R. Cohen 1 , Elizabeth A. Bukusi 5 , Moses R. Kamya 4 , Diane V. Havlir 1 1 University of California San Francisco, San Francisco, CA, USA, 2 University of California Berkeley, Berkeley, CA, USA, 3 University of Massachusetts Amherst, Amherst, MA, USA, 4 Infectious Diseases Research Collaboration, Kampala, Uganda, 5 Kenya Medical Research Institute, Nairobi, Kenya Background: There are few longitudinal studies conducted with clear ascertainment of exposures and outcomes to understand pathways linking mobility and HIV acquisition; multiple measures and sex-stratified analyses are needed to measure the complex and dynamic nature of mobility and its influence on HIV risk in specific settings. Methods: This study examined risk of HIV acquisition over 3 years among mobile and non-mobile adults in 32 rural communities in Uganda and Kenya in SEARCH, a cluster randomized trial conducted from 2013 through 2017. We examined HIV incidence by mobility pre-baseline and over 3 years, and evaluated differences by sex. Poisson regression was used to estimate incidence rate ratios (IRRs) of HIV acquisition among mobile relative to non-mobile adults, with sex-stratified multivariate models adjusted for region, demographic characteristics, and clustering by community. Results: In a cohort of 117,114 adults who were aged ≥15 years, stable residents, and HIV-negative at baseline, 704 new HIV infections occurred by year 3; 11,337 adults (9.7%) had lived ≥1 month outside of their community in past 12 months (8% of women & 11.8% of men, p<0.001). By year 3, 8,511 (7.3%) had out-migrated (7.6% of women & 6.9% of men, p<0.001). In multivariate analyses, risk of HIV acquisition by year 3 was higher in adults who reported pre-baseline mobility (relative to no mobility): IRR=1.47 [95%CI 1.18-1.82]) for having lived ≥1 month outside community in past 12 months (similar effects for women & men), and IRR=1.42 [95%CI 1.11-1.81], for having spent some nights away in past month in men only. Temporal ordering of exposure and outcome cannot be definitively ascertained for mobility reported at year 3, but magnitude of association was highest for having lived >12 months in past 3 years outside community (IRR=3.2 [95%CI 2.21-4.64]), and having changed residence in past 12 months (IRR=2.3 [95%CI 1.8-2.95]). Significant heterogeneity was seen by sex for HIV incidence and mobility, with effects for women but not men in year 3 report of living outside community >12 months in past 3 years, IRR=3.9 [95%CI 2.63-5.78], and >6 months in past year IRR=1.84 [95%CI 1.23-2.76]). Conclusion: Mobility is significantly associated with risk of HIV acquisition in rural Kenya and Uganda; its effect on HIV incidence is influenced by both sex and type of mobility.

850 HIV INCIDENCE AND VIRAL BURDEN AT THE COMMUNITY LEVEL IN HPTN 071 (POPART) Timothy Skalland 1 , Helen Ayles 2 , Ethan A.Wilson 1 , Ayana Moore 3 , Nomtha Mandla 4 , Justin Bwalya 2 , Nkatya Kasese 2 , Rory Dunbar 4 , Estelle Piwowar- Manning 5 , Susan H. Eshleman 5 , Peter Bock 4 , Sarah Fidler 6 , Richard J. Hayes 7 , Deborah J. Donnell 1 1 Fred Hutchinson Cancer Research Center, Seattle, WA, USA, 2 Zambart, Lusaka, Zambia, 3 FHI 360, Durham, NC, USA, 4 Desmond Tutu TB Centre, Western Cape, South Africa, 5 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 6 Imperial College London, London, UK, 7 London School of Hygiene & Tropical Medicine, London, UK Background: HPTN 071 (PopART) was a 3-arm cluster-randomized trial that evaluated use of a combination HIV prevention strategy to reduce HIV incidence. The intervention package included universal HIV testing and treatment (UTT). The trial was conducted in 21 high HIV prevalence communities in Zambia and South Africa (7 matched community triplets). The study primary outcome was HIV incidence in the period 12 to 36 months after the start of the study, measured in a Population Cohort (PC) of ~2000 randomly-selected adults per community (aged 18-44). The intervention effect was greatest in the study arm that included treatment according to national guidelines (Arm B); a lesser effect was observed in the full UTT arm (Arm A), compared to standard of care (Arm C). Methods: For each community, HIV incidence was estimated in the primary analysis period (PC12-PC36), weighted by age and sex. HIV viral load was measured in all HIV-positive PC participants 2 years after the start of the study (PC24). Viral suppression was defined as a viral load <400 copies/mL. Viral burden was defined as the estimated proportion of the entire community (both HIV positive and HIV negative persons) that were not virally suppressed at PC24, weighted by age and sex. We examined associations of viral burden at PC24 with HIV incidence, and whether it mediated the PopART intervention effect on HIV incidence. Results: HIV viral burden at PC24 was strongly associated with HIV incidence (Figure 1; p<0.001). We estimated a mean difference of -1.2% in viral burden for Arm A vs C (95% CI: -2.8%, 0.4%) and a mean difference of -0.85% for Arm B vs C: (95% CI: -2.5%, 0.8%). The average causal mediation effect of viral burden on HIV incidence was not significant (Arm A vs. C p=0.50; Arm B vs. C p=0.77). Conclusion: In HPTN 071 (PopART), higher HIV incidence was associated with higher viral burden. However, the reduction in viral burden did not explain the differential reduction of HIV incidence observed in the two trial intervention arms.

Poster Abstracts

CROI 2020 317