CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

1 Massachusetts General Hosp, Boston, MA, USA, 2 Mbarara Univ of Sci and Tech, Mbarara, Uganda, 3 Epicentre, MSF, Mbarara, Uganda, 4 MGH-Ragon Inst, Boston, MA, USA Background: HIV infection may increase the risk of postpartum infection and infection-related mortality; yet there are insufficient data on epidemiology, microbiology, and risks factors in resource-limited settings. We hypothesized that postpartum infection incidence and attributable mortality in Mbarara, Uganda would be significantly higher in HIV-infected than HIV-uninfected women. Methods: We performed a prospective cohort study of 4,231 women presenting to a regional referral hospital in 2015 for delivery or postpartum care. We collected vital signs after delivery, and performed microbiologic evaluation of febrile and hypothermic women. All febrile, hypothermic, and a subset of 1,708 randomly selected normothermic women were followed with postpartum phone interviews completed at 6 weeks. The primary outcome of interest was incident in-hospital postpartum infection. Secondary outcomes included in-hospital complications (mortality, re-operation, ICU transfer, need for imaging, blood transfusion, wound infection) and 6-week all-cause mortality. We analyzed risk of mortality and complications using Chi squared analysis. We performed univariable and multivariable logistic regression analyses to identify correlates of postpartum infection, with a particular focus on HIV infection. Results: Mean age was 25.2 years and 481 participants (12%) were HIV-1 infected, with a median CD4+ T-cell count of 487 (IQR 338, 698) cells/mm3. Approximately 90% of HIV- infected women (193/215 selected for in-depth survey) were on ART. While hospitalized, 4.8% (205/4231) of the total cohort were febrile or hypothermic, of whom 174 (85%) had blood and urine samples collected. The most common causes of fever were postpartum endometritis (76/193, 39%), urinary tract infection (25/174, 14%), bloodstream infection (5/174, 3%) and malaria (5/174, 3%). Cumulative incidence of postpartum infection was 2.0% and did not differ by HIV status (AOR 1.4, P=0.49, Table 1). However, significantly more HIV-infected women than HIV-uninfected women (4.4% versus 1.2%, P=0.001) developed an in-hospital postpartum complication. Maternal mortality incidence was 0.11% (2/1768) in-hospital and rose to 0.26% (4/1526) by 6 weeks postpartum, without differences by HIV serostatus (P=0.71 and 0.24, respectively). Conclusion: For women in rural Uganda with high rates of ART coverage, HIV infection was associated with increased risk of postpartum complications, but not with in-hospital postpartum infection, in-hospital or 6-week all-cause mortality.

Poster and Themed Discussion Abstracts

774 CMV VIREMIA IN HIV-POSITIVE AND -NEGATIVE PREGNANT WOMEN IN BOTSWANA Natasha O. Moraka 1 , Maryanne R. Ibrahim 2 , Sikhulile Moyo 1 , Prisca Thami 1 , Gloria Mayondi 1 , Christiana Smith 3 , Adriana Weinberg 4 , Rosemary Musonda 1 , Simani Gaseitsiwe 1 , Shahin Lockman 5 1 Botswana Harvard AIDS Inst Partnership, Gaborone, Botswana, 2 Doris Duke Intl Clinical Rsr Fellowship, Los Angeles, CA, 3 Univ of Colorado, Denver, CO, USA, 4 Univ of Colorado, Aurora, CO, USA, 5 Harvard Univ, Boston, MA, USA Background: Despite the success of PMTCT in Botswana, HIV-exposed uninfected (HEU) infants are more than twice as likely to die than HIV-unexposed infants. In Botswana, 26% of pregnant women are HIV infected, 96% of whom are CMV sero-positive. We studied the prevalence of CMV viremia in HIV-infected vs. HIV-uninfected women. We also explored the relationship between detectable maternal CMV viremia and adverse pregnancy outcomes, as well as morbidity and mortality in both HEU and HIV-unexposed children. Methods: We enrolled (during or within 1 week of pregnancy) and followed (for 2 years postpartum) 443 HIV-infected and 451 HIV-uninfected mothers and their 453 HEU / 457 HIV-unexposed live-born infants in a prospective observational study in Botswana (“Tshipidi”). Maternal plasma samples from 295 HIV-positive and 51 HIV-negative women were randomly selected. CMV DNA was quantified from plasma samples obtained at enrollment using the Roche COBAS® AmpliPrep/COBAS® TaqMan® CMV Test (threshold of detection = 50 copies/mL). Results: Median maternal baseline age and CD4 were 27.6 years (Q1, Q3: 23, 33) and 422 (Q1, Q3: 307,570), respectively. A total of 315 (91.2%) samples were successfully amplified, with 14.0% (95%CI 10.1- 17.8) positive for CMV DNA. Only 1(0.3%) participants had a quantifiable (>150 copies/mL) CMV DNA with the Roche TaqMan assay. There was a trend towards higher CMV detection among HIV-infected vs. HIV-uninfected mothers (15.4% vs. 6.1%, respectively; p=0.080). Among HIV-infected women, the presence of detectable CMV DNA was not associated with the composite adverse pregnancy outcome (preterm delivery or small for gestational age, OR = 1.46; 95%CI 0.77 – 2.78); and was not associated with the composite outcome of child hospitalisation and/or death by 24 months (HR = 1.40; 95% CI 0.72 – 2.69). Conclusion: Very few HIV-infected and –uninfected women had detectable CMV DNA. CMV viremia was not associated with any pregnancy and infant outcomes in this small sample. 775 PREGNANCIES IN WOMEN WHO ACQUIRED HIV PERINATALLY Mira Hleyhel 1 , Catherine Dollfus 2 , Roland Tubiana 3 , Christine Rouzioux 4 , Pierre Frange 4 , Stephane Blanche 4 , Jérôme Le Chenadec 1 , Albert Faye 5 , Laurent Mandelbrot 6 , Josiane Warszawski 1 1 INSERM, Le Kremlin-Bicetre, France, 2 Trousseau Hosp, Paris, France, 3 Pitié-Salpétrière Hosp, Paris, France, 4 Necker Hosp, Paris, France, 5 Univ Paris Diderot, Paris, France, 6 Univ Paris Diderot, Colombes, France Background: Pregnancies in women perinatally infected with HIV (PHIV) are increasing in frequency. We investigated whether perinatal HIV acquisition was associated with differences in care management, a higher risk of obstetric complications, uncontrolled viral load near delivery, and adverse neonatal outcomes. Methods: Between 2006 and 2014, of the 2201 pregnant women aged 18-30 years enrolled in the National French Perinatal Cohort (ANRS-EPF-CO1), 508 were primiparous and aware of their HIV-positive status before conception: 46 PHIV and 462 non-PHIV. We compared the risk of uncontrolled viral load (>50 cp/mL), obstetric complications (pre- eclampsia, diabetes, prematurity), and adverse neonatal outcomes (stillbirth, low birthweight) between these two groups.

CROI 2017 335