CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

799 Timing of ART Initiation in Pregnancy and Birth Outcomes in South AfricanWomen Thokozile Malaba 1 ;Tamsin Phillips 1 ; Greg Petro 1 ; Kirsty Brittain 1 ; Allison Zerbe 2 ; Agnes Ronan 1 ; James McIntyre 3 ; Elaine J. Abrams 4 ; Landon Myer 1 1 Univ of Cape Town, Cape Town, South Africa; 2 ICAP at Columbia Univ, New York, NY, USA; 3 Anova Hlth Inst, Johannesburg, South Africa; 4 ICAP, Columbia Univ Mailman Sch of PH, New York, NY, USA Background: Studies have suggested that use of triple-drug antiretroviral therapy (ART) during pregnancy may be associated with prematurity, low birthweight (LBW) and/or small for gestational age (SGA) deliveries. However the nature of any association remains controversial. There are few data from Africa where most ART use in pregnancy occurs; the quality of pregnancy dating in most African studies is poor; and there are limited data on this question involving the most commonly used non-nucleoside reverse transcriptase inhibitor (NNRTI)-based regimens.

Methods: We recruited a cohort of ART-eligible HIV-infected women (n=1464) making their first antenatal visit at a primary care facility in Cape Town, South Africa, March 2013-June 2014. Pregnancy dating was based on a combination of obstetric ultrasound conducted by a research sonographer on all women, last menstrual period and clinical exam. All women were followed from their first antenatal visit through delivery with outcomes abstracted from clinic records. Analyses compared pregnancy outcomes between women on ART at conception versus those initiating at different gestations. Results: In the cohort (median age, 29y; 17% nulliparous; median CD4 cell count 374 cells/uL), 38% (n=575) were on ART prior to conception (93% NNRTI-based regimens; majority TDF+3TC/FTC+EFV/NVP; 7% PI-based regimens excluded) and the remaining 62% (n=907) initiated the public sector first-line regimen TDF+FTC+EFV. Median gestation at initiation was 21 weeks. Overall, 4% of pregnancies ended in a miscarriage or stillbirth; this did not vary by timing of ART initiation (p=0.86). In 1275 live singleton births (mean birthweight, 3048g; 22% preterm; 14% LBW; 21% SGA), prematurity, LBW and SGA deliveries did not vary systematically between women on ART at conception versus those initiating ART during pregnancy (Figure). The absence of associations between adverse birth outcomes and timing of ART initiation persisted after adjusting for maternal age, parity, height, CD4 and viral load at first antenatal visit; only decreased CD4 was associated with increased prematurity and LBW (p<0.001 for both associations) and only nulliparity was associated with increased SGA (p<0.001). Conclusions: These reassuring findings from a well-characterised routine care cohort demonstrate that timing of initiation of widely used NNRTI-based regimens before or during pregnancy does not appear to be associated with adverse pregnancy outcomes.

800 Higher Mortality in HIV-Exposed/Uninfected vs HIV-Unexposed Infants, Botswana

Gbolahan Ajibola 1 ; Gloria Mayondi 1 ; Jean Leidner 2 ; Haruna Jibril 3 ; Joseph Makhema 1 ; Mompati Mmalane 1 ; Modiegi Diseko 1 ; Roger L. Shapiro 4 ; Betsy Kammerer 5 ; Shahin Lockman 6 1 Botswana Harvard AIDS Inst Partnership, Gaborone, Botswana; 2 Goodtables Consulting, Oklahoma, OK, USA; 3 Ministry of Hlth, Gaborone, Botswana; 4 Harvard Sch of PH, Boston, MA, USA; 5 Boston Children’s Hosp, Harvard Med Sch, Boston, MA, USA; 6 Brigham and Women’s Hosp, Harvard Med Sch, Boston, MA, USA Background: It is unknown whether higher mortality previously observed in HIV-exposed uninfected (HEU) infants (compared with HIV-unexposed [HU] infants) in resource limited settings persists when mothers have access to effective antiretroviral treatment (ART). We compared rates of 2-year infant mortality in HEU vs. HU infants in Botswana to assess outcomes in an era where maternal ART is available. Methods: We enrolled HIV-infected and HIV-uninfected mothers (during pregnancy or 1 week postpartum) and their babies in the prospective observational “Tshipidi” study in 2 sites (1 city and 1 village) in Botswana fromMay 2010-July 2012. Live born infants and their mothers were followed for 24 months postpartum; data on socio-demographic factors, health, and psychosocial characteristics were collected at baseline and ~6 monthly, and infant health outcomes ascertained. Participants did not receive their primary medical care through the study. Mothers chose infant feeding method with counseling; per Botswana guidelines, HIV-infected mothers choosing replacement feeding could receive free formula. Results: 949 mothers (474 HIV-infected, 475 HIV-uninfected) and 910 live born infants (453 HIV-exposed, 457 HIV-unexposed) were enrolled. HIV-infected women were older (median 29 vs. 25 years old, p<0.001), had median CD4 410 cells/mm 3 , and 32% took ART during pregnancy. Infants born to HIV-infected mothers had significantly higher risk of death compared with HU infants, even after excluding children documented to be HIV-infected (HR 2.9, p = 0.009, 95% CI 1.3-6.6). The 24-month infant mortality rates, stratified by initial feeding method and HIV infection status, are shown in the table. Conclusions: HIV-exposed/uninfected infants were significantly more likely to die before 24 months than HIV-unexposed infants, with most deaths occurring in the first 3 months of life. The small number of breastfed HEU babies does not permit conclusions about the independent effects of HIV-exposure vs. formula-feeding on infant mortality.

Poster Abstracts

334

CROI 2016