CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Methods: We compared urine dipsticks proteinuria with eGFR among HIV-infected pregnant women starting ART in antenatal care settings in urban Zambia. Analyzing data from a prospective cohort in 3 Lusaka sites, we estimated glomerular filtration rate (eGFR) using the MDRD equation and used the lowest 10th percentile as a proxy for renal dysfunction.. We determined the sensitivity and specificity of 2+ proteinuria and 1+ proteinuria for eGFR in the lowest 10th percentile. Results: FromMarch to August 2017, 215 HIV-infected pregnant women were enrolled and had data for urine dipstick protein and serum creatinine. Median values for age was 27 years (IQR: 24-33), gestational age was 18 weeks (IQR: 15-22), CD4 count was 305 cells/µL (IQR: 191-418), and haemoglobin was 11.0 g/ dL (IQR: 10.1-11.8). Elevated blood pressure (i.e., 140/90mmHg) was noted in 12 women (5.6%). Urine dipstick protein 1+ was seen in 16 women (7.4%) and only 5 women (2.3%) had 2+ proteinuria. Having 2+ proteinuria had 0% sensitivity (97.5%CI: 0.0, 16.1) and 97.4% specificty (95%CI: 94.1, 99.2) for detection of reduced eGFR, while having 1+ proteinuria had 9.5% sensitivity (95%CI:11.7, 30.4) and 92.8% specificity (95%CI: 88.2, 96.0). Conclusion: In contrast to non-pregnant HIV infected adults in similar settings, we found that impaired renal function was rare among HIV+ pregnant Zambian women, and urine dipsticks proteinuria had <10% sensitivity for detecting reduced eGFR making it inappropriate as a screening test for reduced eGFR. This data is the first in our setting to describe renal function in HIV infected pregnant women and has potential to inform guidelines on renal monitoring in women prescribed tenofovir. 809 CONCENTRATIONS OF TFV-DP DURING PREGNANCY AMONG WOMEN USING PrEP Maria Pyra 1 , Peter L. Anderson 2 , Kenneth K. Mugwanya 1 , Jessica E. Haberer 3 , Renee Heffron 1 , Stephen Asiimwe 4 , Elly T. Katabira 5 , Nelly R. Mugo 1 , Elizabeth A. Bukusi 1 , Connie L. Celum 1 , Jared Baeten 1 1 University of Washington, Seattle, WA, USA, 2 University of Colorado Anschutz Medical Campus, Aurora, CO, USA, 3 Massachusetts General Hospital, Boston, MA, USA, 4 Kabwohe Clinical Research Center, Kabwohe, Uganda, 5 Makerere University, Kampala, Uganda Methods: Concentrations of TFV-DP were analyzed from dried blood spot samples collected from 31 pregnant and 32 non-pregnant women using PrEP in an open-label study in East Africa. The lower limit of quantification (LLQ) was 31.3 fmol/punch, using an LC-MS/MS assay. PrEP adherence was assessed by electronic monitoring (MEMS caps) as the number of days opened in the prior month. The primary analysis was a comparison of TFV-DP concentrations between pregnant and non-pregnant women; a GEE model compared each trimester, adjusted for adherence, baseline age, and BMI. Additional analyses compared concentrations in a subset of 12 women with samples before and during pregnancy; generalized linear mixed effects models were used, adjusted for adherence. Results: TFV-DP concentrations were lower during pregnancy than during non-pregnant periods, after controlling for adherence: overall, average TFV-DP concentration 637 fmol/punch in non-pregnant women versus 450 fmol/ punch in pregnant women (n=102 samples). Adjusting for adherence and baseline characteristics, differences were largest in the 2nd and 3rd trimesters: -52 fmol/punch in 1st trimester (p=0.59), -187 fmol/punch in 2nd trimester (p=0.04), and -179 fmol/punch in 3rd trimester (p=0.07). Results were larger in magnitude and significant when restricted to 83 samples with detectable TFV-DP: -104 fmol/punch in 1st trimester (p=0.27), -278 fmol/punch in 2nd trimester (p=0.004), and -260 fmol/punch in 3rd trimester (p=0.01). Among 12 women with samples before and during pregnancy, TFV-DP concentrations were 289 fmol/punch (95% CI -439 to -139, p=0.005) lower on average during compared to before pregnancy, adjusted for adherence. Conclusion: Similar to studies of plasma tenofovir concentrations among HIV-infected women on ART, we found evidence that TFV-DP levels during pregnancy are ~70% of non-pregnant levels, even after adjusting for adherence. The difference could be due to changes in clearance, volume of distribution, or adherence effects not fully accounted for by MEMS Background: The perinatal period is a time of high HIV risk for some women. PrEP may offer HIV protection, but pregnancy could alter PrEP pharmacokinetics. We examined differences in tenofovir-diphosphate (TFV-DP) concentrations during pregnancy among HIV-uninfected women on daily oral PrEP.

measurement. Additional studies are needed to understand these mechanisms and whether lower TFV-DP concentrations during pregnancy affect PrEP efficacy. 810 EX VIVO COTYLEDON PERFUSION FAIRLY PREDICTS IN VIVO ARV HUMAN PLACENTAL TRANSFER Vincent Madelain 1 , Minh Le 2 , Agnes Bourgeois Moine 2 , Sophie Matheron 2 , Diane Descamps 2 , Laurent Mandelbrot 2 , Gilles Peytavin 2 1 INSERM, Paris, France, 2 AP–HP, Paris, France Background: Assessing ARV placental transfer in HIV infected pregnant women is important for prevention of fetal infection, as well as potential toxicities. Clinical data are classically obtained by measuring cord blood (F) to mother (M) plasma concentrations (Cpl) ratio at delivery. However, only scarce studies were reported for most ARV. Methods such as the ex vivo dual perfusion of placental cotyledon model are proposed. We aimed to characterize placental transfer patterns of commonly used ARV in a large cohort, and the predictive value of clearance index (CLI) reported in literature from ex vivo experiments. Methods: In vivo data were obtained from routine therapeutic drug monitoring in Paris hospitals during the 10 last years, where ARV Cpl at delivery were documented for paired M and F samples, collected within 3 hours of each other. Most mothers were adherent to recommended ARV combinations and few fetal transmissions occurred. M and F and amniotic fluid (AF) Cpl were determined by UPLC-MS/MS. Median and 10-90 percentiles were presented for each ARV, and correlation with raw and corrected CLI were tested using Spearman’s method. Results: A total of 540 couples with M and F Cpl were included in the analysis, allowing to calculate 1876 F/M ratios encompassing 20 ARV medications and 179 AF/M ratios. Non Reverse Transcriptase Inhibitors (NRTI) had high F/M ratios above 1 except for abacavir. The F/M ratio appeared to be dependent on the M Cpl, suggesting simple diffusion. On the contrary, Protease Inhibitors (PI) showed far lower F/M ratios than NRTI, below 0.5 for all of them and only 0.04 for saquinavir. Considering NNRTI, the F/M ratio for etravirine was near to those for PI’s whereas nevirapine had a similar ratio to NRTI. Raltegravir also had a high ratio of 1.23, also dependent on M Cpl. The AF/M ratios suggested an accumulation of NRTI (except tenofovir), raltegravir and raltegravir glucuronide in AF. Raw CLI were significantly, but moderately, correlated with F/M ratios, (rho=0.59; p= 0.029). Correcting CLI with the Hutson formula (CPT 2011; 90:67–76) markedly improved the correlation (rho=0.81; p= 0.001). Conclusion: This study confirmed the heterogeneity of drug diffusion to fetus among the different classes of ARV, with high in utero exposure to NRTI and raltegravir, and lower exposure to PI. Although it was not completely predictive, the ex vivo perfusion model effectively explored maternal-to-fetal transfer prior to human administration.

Poster Abstracts

811 CABERGOLINE FOR LACTATION SUPPRESSION AMONG HIV+ AND HIV- WOMEN Sarah Humphrey, Maeve Baechler, Melissa Schiff, Jane Hitti University of Washington, Seattle, WA, USA Background: HIV+ women who are counseled not to breastfeed may benefit from pharmacologic lactation suppression. Single-dose cabergoline is highly effective yet infrequently prescribed due to concern for potential effects on maternal blood pressure (BP). We evaluated the effects of cabergoline on maternal BP and pulse immediately postpartum among women with HIV or obstetric indications for lactation suppression such as fetal/neonatal demise. Methods: We conducted a retrospective cohort study of 224 post-partum women with HIV or obstetric indications for lactation suppression who delivered

CROI 2018 305