CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

770 SELF-REPORT AND DRY BLOOD SPOTS AS MARKERS OF ANTIRETROVIAL ADHERENCE IN PREGNANCY Maria L. Alcaide 1 , Shandir Ramlagan 2 , Violeta Rodriguez 1 , Ryan Cook 3 , Karl Peltzer 2 , Stephen Weiss 1 , Deborah Jones 1 1 Univ of Miami, Miami, USA, 2 HIV/AIDS, STIs, and TB Rsr Prog, Pretoria, South Africa, 3 Univ of California Los Angeles, Los Angeles, CA, USA

Background: Adherence to antiretroviral (ARV) therapy is essential for Prevention of Mother to Child Transmission (PMTCT). In South Africa (SA), PMTCT first line antiretroviral (ARV) regimens are TDF + 3TC (or FTC) + EFV. While self-report is widely used to assess adherence to ARVs, it may be over-reported. This study compared two self-report adherence scales with detection of ARV in dried blood spots (DBS) among HIV infected (HIV+) pregnant women in SA. Methods: N = 392 HIV+ pregnant women receiving ARVs completed two self-reported adherence measures [Visual Analog Scale (VAS), AIDS Clinical Trials Group Adherence] and underwent a blood collection for DBS ARV testing at week 32 of pregnancy. Self-report adherence was coded as adherent if no missed doses were reported. DBS adherence was defined as 3 drugs detected (TDF + 3TC + EFV) or TDF + EFV detected. An area under the receiver operating characteristic curve (AUROC) analysis was conducted to examine the performance of the VAS and the ACTG scales in identifying participants as adherent, using DBS as the gold standard. Kappa statistics (κ), accuracy, sensitivity, specificity, positive predictive values (PPV) and negative predictive values (NPV) were calculated. Results: DBS ARV detection was as follows: adherent = 74% ; Non-adherent = 26%. The proportion of participants identified as adherent by self-report were 86% (VAS) and 80% (ACTG). VAS relative to DBS: AUROC = 0.543 (VAS performed poorly in predicting adherence by DBS); κ = 0.101, (slight intermeasure agreement between the VAS and the DBS); accuracy = 0.719 [95% CI 0.67, 0.76]; sensitivity = 0.907 [95% CI 0.87, 0.94]; specificity = 0.178 [95% CI 0.11, 0.27]; PPV = 0.761 [95% CI 0.71, 0.80], and NPV = 0.400 [95% CI 0.26, 0.56]. ACTG relative to DBS: AUROC = 0.538, (ACTG performed poorly in predicting adherence by DBS); κ = 0.081 (poor intermeasure agreement between the ACTG and the DBS); accuracy = 0.673 [95% CI 0.63, 0.72]; sensitivity = 0.818 [95% CI 0.77, 0.86]; specificity = 0.257 [95% CI 0.18, 0.35]; PPV = 0.760 [95% CI 0.71, 0.81]; and NPV 0.329 [95% CI 0.23, 0.44]. Conclusion: Detectable levels of ARV were suboptimal in this population, indicating high risk of perinatal HIV infection and ARV resistance. Programs to strengthen ARV adherence among HIV+ pregnant women in rural SA are needed. Validation of self-reported ARV adherence among pregnant HIV+ women in SA are warranted to support PMTCT goals. Funded by NIH (R01HD078187 and P30AI073961). 771 UNINTENDED PREGNANCY PREDICTS SUBSEQUENT RAISED VIRAL LOAD IN THE POSTNATAL PERIOD Kirsty Brittain 1 , Tamsin Phillips 1 , Allison Zerbe 2 , Agnes Ronan 1 , Greg Petro 1 , Nei-Yuan M. Hsiao 3 , James A. McIntyre 4 , Elaine J. Abrams 5 , Landon Myer 1 , for the Maternal Child Health - AntiretroviralTherapy (MCH-ART) Study 1 Univ of Cape Town, Cape Town, South Africa, 2 Columbia Univ, New York, NY, USA, 3 NHLS, Cape Town, South Africa, 4 Anova Hlth Inst, Johannesburg, South Africa, 5 ICAP at Columbia Univ, New York, NY, USA Background: Unintended pregnancies are common in HIV-infected women in resource-limited settings and are a source of ongoing MTCT. Meanwhile non-adherence and raised viral load (VL) on ART occur frequently during the postpartum period, but the determinants of these are poorly understood. We investigated whether raised VL on ART is more common postpartum in women who had an unplanned pregnancy. Methods: Working in a public sector clinic in Cape Town, South Africa, we followed consecutive HIV-infected women initiating ART from their 1st antenatal care (ANC) visit up to 18 months postpartum. Pregnancy intentions were measured at the 1st ANC visit with the London Measure of Unplanned Pregnancy (α=0.87); analyses categorised pregnancies as planned, ambivalent or unplanned. Viral load (VL) testing (Abbott RealTime HIV-1) was conducted at 3-6 monthly intervals postpartum along with standardised assessments of depression, alcohol use and intimate partner violence (IPV). In analysis, Poisson models examined the associations between pregnancy planning and elevated VL≥1000 copies/mL. Results: A total of 358 women (mean age, 29 years) were followed for a median of 18 months postpartum. At 1st ANC visit, planned, ambivalent and unplanned pregnancy was reported by 20%, 21% and 59%, respectively. Overall, 115 women (32%) experienced one or more VL≥1000 copies/mL postpartum, with elevated VL occurring more frequently at each time point in the postpartum period in women reporting ambivalence or an unplanned pregnancy (Figure). Compared to women reporting a planned pregnancy, those reporting ambivalence (risk ratio [RR]: 1.90; 95% CI: 1.05-3.42) or unplanned pregnancy (RR: 2.05; 95% CI: 1.21-3.46) were more likely to experience elevated VL≥1000. These associations persisted after adjustment for demographic characteristics and were independent of depression, alcohol use and IPV. Neither breastfeeding duration nor MTCT varied significantly by pregnancy planning. Based on the most conservative adjusted estimates, we calculate that up to 30% of elevated VL in women on ART postpartummay be associated with unplanned pregnancy in this setting. Conclusion: These novel data suggest that elevated VL postpartum occurs more frequently in women with unplanned pregnancies. This underscores the need to incorporate pregnancy planning into routine care for HIV-infected women, and to recognise that postpartumwomen who did not intend their pregnancies may require specific attention from counselling and support interventions. 772 SUBSEQUENT PREGNANCY OUTCOMES IN WOMEN DURING FOLLOW-UP IN PROMISE 1077HS Jose H. Pilotto 1 , Sean Brummel 2 , Risa M. Hoffman 3 , Paula Britto 2 , Gaerolwe Masheto 4 , Linda Aurpibul 5 , Nahida Chakhtoura 6 , Karin Klingman 7 , Judith S. Currier 3 , for the 1077HSTeam 1 Fiocruz, Rio de Janeiro, Brazil, 2 Harvard Univ, Boston, MA, USA, 3 Univ of California Los Angeles, Los Angeles, CA, USA, 4 Harvard Univ, Gaberone, Botswana, 5 Chiang Mai Univ, Chiang Mai, Thailand, 6 NICHD, Bethesda, MD, USA, 7 DAIDS, NIAID, Bethesda, MD, USA Background: Rates of adverse pregnancy outcomes for women who conceive on antiretroviral therapy (ART) may be increased, but data are conflicting. Methods: In PROMISE 1077HS, asymptomatic HIV+, non-breastfeeding women with pre-ART CD4 cell count ≥400 cells/mm³ who started ART during pregnancy were randomized up to 42 days after delivery to continue (cART) or discontinue ART (dART). LPV/RTV with TDF/FTC or ZDV/3TC was the preferred study regimen. Sixty sites in Argentina, Botswana, Brazil, China, Haiti, Peru, Thailand and the US participated between 12/2011-11/2014. Women randomized to dART were recommended to restart if a subsequent pregnancy occurred or for clinical indications. This analysis includes outcomes for all subsequent pregnancies that occurred prior to offering all women ART in 7/2015. We compared subsequent pregnancy outcomes among women in the cART versus dART arm using Fisher’s exact test (post hoc analysis). Results: Subsequent pregnancies occurred in 277/1652 (17%) women (cART: 144/827, dART: 133/825). A pregnancy outcome was recorded for 266 women with median age 27.4 years (IQR 23.7, 31.1) at pregnancy diagnosis, and median CD4 688 cells/mm³ (IQR 529, 867) recorded at 2 months prior to pregnancy diagnosis. Two hundred (75%) live births were included, 40 (15%) spontaneous abortions (<20 weeks gestation), 18 (7%) induced abortions (<20 weeks gestation) and 8 (3%) stillbirths (≥20 weeks gestation). At 12 weeks prior to pregnancy diagnosis, 86% (120/140) in the cART group were on a boosted/non-boosted PI regimen versus 6% (8/140) NNRTI. In the dART arm, 19/126 (15%) restarted ART prior to pregnancy diagnosis: 74% (14/19) were on a PI regimen versus 26% (5/19) NNRTI. After pregnancy diagnosis (first regimen during pregnancy), there was frequent use of PIs in the cART arm (89% (124/140) PI versus 7% (10/140) NNRTI) and among those restarting ART in the dART arm (53% (67/126) PI versus 27% (34/126) NNRTI). Spontaneous abortions were more common in the cART arm (cART: 19.3% (27/140), dART: 10.3% (13/126); p=0.06), as were stillbirths (cART: 4.3% (6/140), dART: 1.6% (2/126); p=0.29). When stillbirths and spontaneous abortions were combined, there was a statistically significant higher rate in the cART arm (cART: 23.6% (33/140), dART: 11.9% (15/126); p=0.02). Conclusion: Women randomized to continue ART after their index pregnancy who subsequently conceived were more likely to have spontaneous abortion or stillbirth compared to women randomized to stop ART. 773 HIV AND RISK OF POSTPARTUM INFECTION, COMPLICATIONS, AND MORTALITY IN RURAL UGANDA Lisa Bebell 1 , Joseph Ngonzi 2 , Mark J. Siedner 3 , Winnie Muyindike 4 , Bosco M. Bwana 2 , Laura E. Riley 1 , Yap Boum 5 , David Bangsberg 6 , Ingrid V. Bassett 1

Poster and Themed Discussion Abstracts

CROI 2017 334