CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

Mia Ekström 3 , Susanne Strömdahl 3 , Larry W. Chang 4 , Ronald H. Gray 5 , Steven J. Reynolds 4 , Maria Wawer 5 , David Serwadda 1 1 Makerere University College of Health Sciences, Kampala, Uganda, 2 Rakai Health Sciences Program, Kalisizo, Uganda, 3 Karolinska Institute, Stockholm, Sweden, 4 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 5 Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA Background: Pre-Exposure prophylaxis (PrEP) has been recommended for key and priority populations most-at-risk of HIV. In 2017, one of the first larger attempts to scale up PrEP using tenofovir and lamivudine at population- level in Uganda, was initiated by the Rakai Health Sciences Program in HIV hyper-endemic trading centers and fishing communities on Lake Victoria with CDC-Uganda under PEPFAR support. We report on acceptability and retention of clients on the program at 9 months of follow-up. Methods: Program data from implementing clinics were used for the evaluation. Acceptability of PrEP was defined as having been initiated on PrEP after satisfying the eligibility criteria of being at high HIV risk. Retention on PrEP was measured at months 1, 3, 6, and 9 following PrEP enrolment. Multivariable modified Poisson regression was used to estimate prevalence ratios and 95% confidence intervals for the association between covariates, acceptability and retention on PrEP. Results: A total of 2637 individual were screened for PrEP of whom 2439 (93%) were eligible; 2285 (94%) of the eligible clients enrolled on PrEP. Enrolled clients included sex workers (54.0%), fisher folk (20.3%), truck drivers (11.2 %), Adolescent girls and young women (4.9%), HIV-negative individuals in discordant relationships (7.9 %) and others (1.7%). Acceptance of PrEP did not differ significantly by age, gender and risk categories, except for lower acceptance among fisher folk (PR=0.87, 95% CI=0.84,0.91) compared to individuals in discordant couples as well as a slightly higher uptake among those divorced/separated compared to married individuals (PR=1.03, 95% CI=1.0- 1.06). Retention, as measured by returning to the clinic for refills, was 47.6% at month 1, 31.3% at month 3, 16.3 % at month 6 and 4.8 % at month 9. Retention was lowest among adolescent girls and young women who did not identify as sex workers (PR=0.38, 95% CI=0.23-0.64) and among fisher folk (PR=0.32, 95% CI=0.24-0.42) compared to individuals in discordant relationships. Retention was higher among individuals aged 25-34 ( PR=1.21, 95% CI=1.04- 1.42) and 35+ (PR=1.38, 95% CI=1.15-1.65) compared to ages 15-24. Retention did not differ by sex and marital status. Conclusion: Acceptability of PrEP was high in this population; however, clients, especially younger women and fisher folk who are highly mobile, rapidly dropped out of the program. Research on reasons for discontinuation and interventions to optimize retention on PrEP are critical to program success. 998 IMPACT OF A CONTRACEPTIVE RING ON VAGINAL BACTERIA ASSOCIATED WITH HIV ACQUISITION Jeanne M. Marrazzo 1 , Sujatha Srinivasan 2 , Elizabeth M. Irungu 3 , Katherine Thomas 4 , Nelly R. Mugo 3 , Peterson Mwaniki 3 , Catherine Kiptinness 5 , Kenneth Ngure 6 , Kate Heller 4 , Meighan Krows 4 , Kristina Garcia 2 , Steven Gakuo 3 , Tina L. Fiedler 2 , David N. Fredricks 2 1 University of Alabama at Birmingham, Birmingham, AL, USA, 2 Fred Hutchinson Cancer Research Center, Seattle, WA, USA, 3 Kenya Medical Research Institute, Nairobi, Kenya, 4 University of Washington, Seattle, WA, USA, 5 Kenyatta National Hospital, Nairobi, Kenya, 6 Kenyatta University, Nairobi, Kenya Background: Specific vaginal bacteria have been associated with increased risk for HIV acquisition in sub-Saharan African women, including several linked to bacterial vaginosis (BV). Limited data support a favorable effect of a contraceptive vaginal ring (CVR) containing estrogen and progesterone (NuvaRing) on vaginal bacteria. Pregnancy is an independent risk for HIV acquisition and transmission; thus, contraception may comprise biomedical prevention for women with or at risk for HIV, and hormonal modulation of key vaginal bacteria, which are also impacted by menses, might also be advantageous. We randomized women treated for BV in Thika, Kenya to continuous (menstrual suppression) vs. cyclic (regular menses) use of NuvaRing, and assessed effects on key vaginal bacteria associated with HIV acquisition. Methods: Women aged 18-40 years were enrolled and treated for BV with oral metronidazole. One month later, they were randomized and seen monthly for 7 months, when vaginal swabs were collected. Concentrations of bacterial taxa previously shown to be associated with increased HIV risk, and one associated with protection (L. crispatus), were measured using quantitative PCR. We used linear mixed models stratified by randomization arm and HIV status at

enrollment to compare mean differences in log10 bacterial DNA concentrations at the visit prior to CVR initiation relative to 2-3 months post-initiation as an early marker of CVR impact. Results: Between April 2016 to November 2017, 151 women (median age 27 y) were enrolled and 122 (81.9%) initiated CVR use, and 98 had qPCR data available (22 of whomwere HIV-infected) at a total of 277 visits (98 pre-CVR and 179 post-CVR insertion). Women in the continuous use CVR group had significantly reduced concentrations of all high-risk bacteria measured at 2-3 months post- insertion (Table). Similarly significant results were seen in women with HIV, with the exception of no change in P. bivia. Conclusion: Continuous CVR use with menstrual suppression over 2-3 months reduced quantities of bacteria previously associated with increased HIV acquisition risk in women. Vaginal rings are a promising strategy that should be evaluated for delivery of multipurpose prevention in Kenyan women.

Poster Abstracts

999 DOES BV MODIFY THE EFFECT OF HORMONAL CONTRACEPTION ON HIV ACQUISITION? Michelle C. Sabo 1 , Barbra A. Richardson 1 , Walter Jaoko 2 , Kishorchandra Mandaliya 2 , Jared Baeten 1 , Ludo Lavreys 3 , Hal Martin 4 , Julie M. Overbaugh 1 , R. Scott McClelland 1 1 University of Washington, Seattle, WA, USA, 2 University of Nairobi, Nairobi, Kenya, 3 Maisha Consulting BVBA, Vilvoorde, Belgium, 4 Gilead Sciences, Inc, Foster City, CA, USA Background: Multiple studies have established an association between depot-medroxyprogesterone acetate (DMPA) and HIV acquisition in women. A recent study of serodiscordant couples in Zambia was the first to find that DMPA and oral contraceptive pills (OCPs) were associated with increased risk of HIV acquisition when BV was present, but not when BV was absent. The purpose of this study was to test the hypothesis that BV is an effect modifier of the association between hormonal contraception and HIV acquisition in the Mombasa Cohort, a long-term open cohort study of female sex workers in Kenya. Methods: Visits contributed by HIV-negative women participating in the cohort between February 1993 and April 2017 were included. Women provided behavioral data (including contraceptive use) and underwent a physical examination and lab testing for HIV and sexually transmitted infections (STIs) at monthly visits. Cox proportional hazards models were used to assess the relationship between HIV seroconversion and use of DMPA, OCPs, or contraceptive implants (each evaluated separately), compared to no hormonal contraception. BV, defined as a Nugent score ≥7 was assessed for effect modification of the relationship between contraceptive use and HIV seroconversion. Results: A total of 1,985 women contributed 7,107 person-years of follow- up, and 307 women seroconverted for antibodies to HIV (4.32/100 person- years). At baseline, 1178 (59.3%) women reported use of condoms only or no contraception, 410 (20.7%) reported using DMPA, 227 (11.4%) reported using oral contraceptives, 69 (3.5%) reported using implants, and 645 (32.5%) had BV. Both DMPA (aHR 1.72, 95% CI 1.34, 2.20; p<0.001) and OCPs (aHR 1.48, 95% CI 1.05, 2.09; p=0.02) were associated with increased risk of HIV acquisition in analyses adjusted for demographic factors, risk behaviors, and the presence of STIs. Contraceptive implants were not associated with HIV acquisition (aHR 0.99, 95% CI 0.40, 2.45; p=0.9). In this cohort, BV was not a significant modifier of the effects of DMPA (p=0.4), OCPs (p=0.9), or implants (p=0.5) on HIV acquisition.

CROI 2019 391