CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

1 Kheth’Impilo, Cape Town, South Africa, 2 UNICEF, New York, NY, USA, 3 Univ of Cape Town, Cape Town, South Africa Background: Young Southern African women have the highest HIV incidence globally. Pregnancy doubles the risk of HIV acquisition, and maternal HIV acquisition contributes significantly to the Sub-Saharan paediatric HIV burden. HIV pre-exposure prophylaxis is currently contraindicated during pregnancy and breastfeeding in South Africa. Little evidence of combination HIV prevention interventions during pregnancy and lactation are available globally. We measured HIV incidence amongst pregnant and postpartum women receiving a community-based combination HIV prevention intervention in a high HIV incidence setting in South Africa. Methods: A cohort study including HIV-uninfected pregnant women was performed. Lay community-based health workers provided sexual health counselling and performed three-monthly home and clinic-based individual and couples HIV testing and counselling (HTC) until 18 months postpartum. Male partners were referred for medical circumcision, sexually transmitted infections and HIV treatment as appropriate. Kaplan-Meier analyses and Cox’s regression were used to estimate HIV incidence and factors associated with HIV acquisition. Results: A total of 1356 women were included with median age of 22.5 years (IQR: 19.4-27.0) and median gestational age at presentation of 16 weeks (IQR: 12-16). Included women received 5289 HIV tests. Eleven new HIV infections were detected over 828.3 person-years (PY) of follow up, with HIV incidence being 1.33 infections/100 PY (95% CI: 0.74-2.40). The antenatal HIV incidence rate was 1.49 infections/100 PY (95% CI: 0.64 to 2.93) and the postnatal HIV incidence rate was 1.03 infections/100 PY (95% CI: 0.33-3.19). Women within known serodiscordant couples, adjusted hazard ratio (aHR)=32.7 (95% CI: 3.8-282.2), and women with newly diagnosed HIV-infected partners, aHR=126.4 (95% CI: 33.8-472.2) had substantially increased HIV acquisition. Women with circumcised partners had a reduced risk of incident HIV infection, aHR=0.22 (95% CI: 0.03-1.86). Conclusion: Previous regional studies have measured maternal HIV incidence between 4.8 infections/100 PY to 16.8 infections/100 PY, thus HIV incidence in this study was 73%-86% lower than previous studies. Community-based combination HIV prevention interventions show promise in reducing high maternal HIV incidence. Expanded roll-out of home-based couples HTC and initiating pre-exposure prophylaxis for pregnant women within serodiscordant couples should be considered in Southern Africa. 764 HIV INCIDENCE, CASCADE, AND TESTING AMONG MOTHERS IN WESTERN KENYA Alliance Nikuze 1 , Stephen Wanjala 2 , Jihane Ben-Farhat 1 , Willis Omwoyo 3 , Laura Oyiengo 4 , Elisabeth Szumilin 5 , Leon Salumu 5 , Alexandra Vandenbulcke 5 , Valarie S. Opollo 6 , David Maman 7 1 Epicentre, Paris, France, 2 MSF, Nairobi, Kenya, 3 Ministry of Hlth, Homa Bay, Kenya, 4 Natl AIDS/STI Control Prog, Nairobi, Kenya, 5 MSF France, Paris, France, 6 Kenya Med Rsr Inst, Kisumu, Kenya, 7 Epicentre, Cape Town, South Africa Background: There are limited evaluations of prevention of mother to child HIV transmission (PMTCT) programs, particularly testing coverage in pregnant women, since the implementation of option B+ in sub-Saharan Africa. The Impact of Expanded Screening Strategies (IESS) study aimed at reconstructing the PMTCT cascade in Ndhiwa sub-county Methods: Through a cross-sectional facility-based survey at expanded programs of immunization (EPI) and maternity, mother-infant pairs were enrolled in the study from February 2016 July 2016. The questionnaires collected information on participants’ background, reproduction, pregnancy, HIV/AIDS, ART coverage and survival. From thirty three health facilities providing ART services in the sub-county, 26 were selected due to their geographical coverage. All HIV-positive mothers and children had their VL measured, regardless of their ART status. Population viral load suppression is defined the proportion of HIV-positive women with a VL<1,000 cp/mL Results: A total of 3,585 women were enrolled: 1925 at EPI 6 weeks, 1116 from EPI 9 months and 544 frommaternity. Overall median age was 23 [IQR 19-29]. At ANC, testing coverage was high with 96.5% (95%CI 95.9-97.1) women testing during their last pregnancy but systematic retesting was very low with 38.1% (95%CI 36.4-40.0) of all women tested at least twice during ANC. Among 882 HIV-positive women (HIV prevalence: 24.7%; 95%CI 23.3-26.1), 513 were already diagnosed prior to the beginning of their last pregnancy, 336 were diagnosed during their last pregnancy and 33 (higher in EPI 6 weeks than 9months 23 vs. 6) were undiagnosed at the time of the survey. Overall HIV Incidence was 4.1 new cases per100PY (95%CI 2.9-5.7). HIV incidence varies with age from 3.1 among those <19 years to 4.7 and 5.9 new cases per 100py at 20-24 and 25-29 years. It then decreased at 3.3 new cases for those age 30 years or more. The different steps of the cascade were as follow: 96.3%were diagnosed, (95%CI 94.8-97.3), 92.4% (95%CI 90.5-94.9 ) were on ART and 78.1% (95%CI 75.0-80.8) were suppressed (VL<1,000 cp/mL) Conclusion: High level of coverage and viral suppression were achieved among mothers following the implementation of option B+. Nevertheless, systematic retesting throughout pregnancy and breastfeeding needs to be reinforced as HIV incidence remained extremely high 765 DETECTION OF HIV IN BREAST MILK AMONG PREGNANT/POSTPARTUMWOMEN WITH RECENT HIV Alison L. Drake 1 , John Kinuthia 2 , Daniel Matemo 2 , Barbra A. Richardson 1 , Sandy Emery 3 , Vrasha Chohan 1 , Julie Overbaugh 3 , Grace John-Stewart 1 1 Univ of Washington, Seattle, WA, USA, 2 Kenyatta Natl Hosp, Nairobi, Kenya, 3 Fred Hutchinson Cancer Rsr Cntr, Seattle, WA, USA Background: Incident HIV infection during pregnancy/postpartum substantially increases mother-to-child HIV transmission (MTCT) risk due to high maternal HIV viral load (VL). Breast milk VL among women with recent infection has not been well characterized, and could contribute to postnatal MTCT. We prospectively measured breast milk and plasma VL among pregnant/postpartumwomen with recent HIV infection. Methods: Pregnant/postpartumwomen in Western Kenya with recent HIV infection (documented negative ≤3 months prior) were identified and prospectively followed. Women who started antiretroviral therapy (ART) <40 days after HIV diagnosis with ≥1 breast milk sample were included in the analysis. VL lower limit of detection was 1.87 and 2.17 log10 copies/mL for breast milk and plasma samples collected serially, respectively. Generalized estimating equations with a logit link were used to determine the relationship between plasma VL and duration of ART on detection of HIV in breast milk. Results: Among 25 women with recent infection, 14 were diagnosed in pregnancy and 11 postpartum. Median age was 21 years (interquartile range [IQR]:19-26) and time to ART initiation was 13 days (IQR 8-19). A total of 133 breast milk samples were tested for HIV, with a median of 6 (IQR 3-8) assays/woman. Median baseline plasma VL was 5.56 log10 copies/mL (IQR: 4.98-5.70). HIV was detected in the first breast milk sample collected from 5 (36%) women diagnosed during pregnancy and 6 (55%) diagnosed postpartum; breast milk VL ranges in both groups were similar (2.30-4.80 and 2.36-4.44 log10 copies/mL, respectively). Most (88%) women had HIV detected in ≥1 breast milk sample; individual-level breast milk VL over time are shown in Figure 1. Plasma VL and months since ART were not associated with detection of HIV in breast milk (Odds Ratio [OR] 1.15 per 1 log increase, 95% Confidence Interval [CI]:0.81-1.63) and OR 0.98, 95% CI:0.88-1.09; respectively). Two infants acquired HIV, at 2 weeks (mother diagnosed in pregnancy) and 6 months (mother/infant concurrently diagnosed). Breast milk VL for the mother who transmitted at 6 months was 4.4 log10 copies/mL at the time of diagnosis. Conclusion: Among recently infected pregnant/postpartumwomen, HIV was detected in breast milk soon after infection and often concurrent with diagnosis. Breast milk HIV was detected among all women, even among women who initiated ART in pregnancy. Early identification and treatment of incident maternal HIV infections is critical to prevent MTCT.

Poster and Themed Discussion Abstracts

CROI 2017 331