CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

controllers and women with an undetectable HIV load on day 14 (+/-3 days) were excluded. Time to undetectable was inferred from the mid-point between the last detected and first undetectable HIV loads. Data analysis was performed in SPSS. Results: 160 women with median age 30 (range 17-42) years; 83% Black African/Caribbean, 12%white; HIV was acquired heterosexually in 96% and by IDU in 2%. Four had hepatitis virus co-infection. Baseline CD4 count and viral load were 313 (30–1136) cells/mL and 17290 (548–588020) cp/ml respectively. cART was commenced at 22 (5–39) weeks gestational age (GA). In the first 14 days of therapy, HIV half-life (T/2) was 2.5 (1.4–5.3) days and 105 (66%) achieved an undetectable HIV load at 36 weeks GA, 130 (81%) by delivery. Data were analysed by third agent: PI (n=114; 71%), NNRTI (n= 31; 18%) or Integrase Inhibitors (II) (n=7; 4%) giving HIV T/2 of 2.6; 2.3 and 1.5 days, whilst time to undetectable were 42, 41 and 27 days respectively. There were significant differences in T/2 by class (p = 0.001) and by individual agent (p = 0.02) but not for time to undetectable. T/2 was significantly slower for Atazanavir (ATZ) than for Nevirapine, Saquinavir and Lopinavir (Table). In logistic regression analysis ATZ (p 0.02) and GA at starting cART (p 0.03) were associated with HIV load at delivery. GA being 4 weeks later in those with a detectable HIV load at delivery whilst ATZ was associated with high suppression rate. Conclusions: A range of cART combinations were effective in achieving undetectable HIV load at 36 weeks GA and at delivery. Correlation between initial decay and viral load at delivery was poor: ATZ had both a slower initial viral decay and a high rate of viral suppression. The limited data on II suggest they performwell in pregnancy and may be useful for high HIV load, especially if treatment is initiated after 20 weeks GA.

795 Viraemic Episodes Occur Frequently in Postpartum South AfricanWomen on ART Landon Myer 1 ; Lorn Dunning 1 ; Nei-Yuan Hsiao 1 ;Tamsin Phillips 1 ; Allison Zerbe 2 ; James McIntyre 3 ; Elaine J. Abrams 4 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: With growing numbers of pregnant women starting antiretroviral (ART) there are significant concerns regarding adherence, especially during the postpartum (PP) period. Viral load (VL) monitoring for pregnant and PP women is promoted in international guidelines to identify nonadherence and treatment failure, but little is known about the frequency and magnitude of elevated VL in this group. Methods: From April 2013-May 2014, we recruited a cohort of pregnant women initiating ART from a public sector primary care antenatal clinic in Cape Town, South Africa. Separate from routine HIV care, up to 9 VLs were conducted at regular intervals from pre-initiation through 12 months PP. In analysis, VL taken after reaching viral suppression (VS, <50 copies/mL [c/mL]) were divided into major (>1000c/mL) and minor (>50 and ≤1000c/mL) viraemic episodes (VE). Mixed-effects Poisson models were used to examine the incidence rates (IR) and rate ratios (IRR) of VE over time by maternal characteristics. Results: Among 607 women (median age, 28 years; median gestation at initiation, 21 weeks; median CD4, 345 cells/uL, median pre-ART VL, 3.99 log 10 c/mL, 4% previous defaulting on ART), 86% (n=523) achieved VS and are included. After VS, 2636 VL tests were conducted over a total of 5092 woman-months (wm) of observation (707 wm antenatally, 4385 wm PP); 56% and 39% of women were breastfeeding at 6 and 12 months PP. Overall, 117 major and 56 minor VE were observed (IR, 2.3 and 1.1 per 100wm, respectively; p<0.001); 60% of major VE involved sustained viraemia>1000c/mL (≥2 consecutive measures). Peak viraemia post-VS (median, 3.79 log 10 c/mL) was linearly correlated with pre-ART VL (p<0.001). From initial VS to 12m PP, 70% of women maintained VS consistently while 22% experienced at least 1 major VE and the vast majority of VE (98%) occurred PP (Figure). Major VE after VS was independently associated with younger age (IRR, 0.90 per year; p<0.001), previous defaulting on ART (IRR, 3.28 vs no previous ARVs; p=0.030), and PP follow-up (IRR, 7.85 vs antenatal sampling; p<0.001). Among PP women, VE frequency increased with increasing duration of ART use (IRR, 1.16 per month after VS; p<0.001). Conclusions: Viremia occurs frequently after VS in women initiating ART in pregnancy. VL monitoring postpartum is likely to have substantial yield within ART programs; however the relative contributions of nonadherence versus viral resistance, and in turn the most appropriate intervention strategies, require urgent attention.

Poster Abstracts

332

CROI 2016