CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

Conclusion: HEU children with fetal EFV-exposure may be at higher risk for delays in some neurodevelopmental and social-emotional domains than HEU children with fetal exposure to non-EFV-based ARVs.

787 HIV-EXPOSED UNINFECTED INFANT GUT MICROBIOME EVOLUTION IN THE FIRST YEAR OF LIFE Kathleen M. Powis 1 , Jeremy E. Wilkinson 1 , Galeb Abu-Ali 1 , Samuel W. Kgole 2 , Gosego Masasa 2 , David B. Gootenberg 3 , Deborah Kacanek 1 , Francis Banda 4 , Sikhulile Moyo 2 , Mompati O. Mmalane 2 , Joseph Makhema 2 , Roger L. Shapiro 1 , Douglas Kwon 3 , Curtis Huttenhower 1 1 Harvard University, Boston, MA, USA, 2 Botswana Harvard AIDS Institute Partnership, Gabarone, Botswana, 3 Ragon Institute of MGH, MIT and Harvard, Cambridge, MA, USA, 4 University of Botswana, Gaborone, Botswana Background: HIV-exposed uninfected (HEU) infants experience higher infectious morbidity than HIV-unexposed uninfected (HUU) infants. Infant commensal gut microbiome influences the developing infant immune system, with dysbiosis associated with immune activation. We compared gut microbiome evolution in the first year of life and hospitalizations between HEU and HUU infants Botswana. Methods: Women living with HIV (WLHIV) and HIV-uninfected (HIV-U) women were enrolled in the Infant Gut Microbiome Study between 36 weeks gestation and 3 days post-delivery. Study eligibility required vaginal delivery of a full- term (≥ 37 weeks gestation), singleton infant, with commitment to exclusively breastfeed (EBF) for six-months. WLHIV had to be on first line efavirenz- containing antiretroviral treatment regimen ≥ six weeks prior to delivery. Infant rectal swabs were obtained at delivery, 1, 3, 6, 9, and 12 month study visits. DNA extracted from rectal swabs was used to perform 16S rRNA gene sequencing, amplicon data processing, taxonomic profiling, and downstream biostatistical analysis. Results: Longitudinal gut microbiome was analyzed from 315 rectal swabs contributed by 58 infants, 34 (59%) of whomwere HEU. Median EBF duration did not differ between HEU and HUU infants, (5.65 vs 5.70 months (mos); p=0.36). Total breastfeeding duration was shorter among HEU infants (6.0 vs 9.0 mos; p=0.02). Significant differences were observed across time from birth to 12 months in both HEU and HUU subsets (filtered terminal taxa relative abundance Bray-Curtis dissimilarity PERMANOVA; p < 0.05). Terminal taxa differences can be seen among time points (Figure 1A), whereas HUU and HEU compositions showed no significant differences averaged across all time points (Figure 1B). Only 4 infants hospitalizations occurred, 3 among HEU infants without differences in microbiome between hospitalized and non-hospitalized infants. Conclusion: Significant changes in the gut microbiome of both HEU and HUU infants in the first year of life were noted, as would be expected. However, we did not observe differences in the 30 most prevalent taxa between HEU and HUU infants, or differences between hospitalized and non-hospitalized infants. Given the small number of hospitalizations, we were underpowered to detect such a difference. Further studies are needed to better understand how differences in breastfeeding duration influence gut microbiome and immune system phenotype and function of HEU infants.

Poster Abstracts

788 DIFFERENCES IN GUT MICROBIOME IN HIV-INFECTED VERSUS HIV- EXPOSED, UNINFECTED INFANTS Wei Li A. Koay 1 , Hyunwook Koh 2 , Mutsa Bwakura-Dangarembizi 3 , Myron Levin 4 , Adriana Weinberg 4 , Ni Zhao 2 , Deborah Persaud 2 1 Children’s Research Institute, Children’s National Health System, Washington, DC, USA, 2 Johns Hopkins University, Baltimore, MD, USA, 3 University of Zimbabwe, Harare, Zimbabwe, 4 University of Colorado Denver, Denver, CO, USA Background: Gut dysbiosis is observed in adults with HIV compared with uninfected adults and implicated in persistent inflammation (IF) and immune activation (IA). Little is known about the gut microbiome in HIV-infected (HIV+) infants, who also have persistent IF and IA compared with HIV-exposed uninfected (HEU) infants. The gut microbiome in breastfed (BF) and non- breastfed (NBF) HIV+ and HEU infants was assessed. Methods: 40 (20 HIV+ and 20 HEU) infants on co-trimoxazole prophylaxis were selected from a clinical trial (IMPAACT P1072/NWCS 612) of rotavirus vaccine (RotaTeqTM) based on stool availability, having age-matched (3 mos.) and breastfeeding-matched HEU control infants. 16S rRNA (V3V4) sequences from stool DNA were assigned organizational taxonomic units with QIIME. α- (Chao1, abundance coverage estimator (ACE), Shannon, Simpson) and β- (Bray-Curtis, Jaccard, unweighted and weighted UniFrac) diversity, and differentially abundant taxa (linear discriminant analysis effect size (LEfSe)) were analyzed. Multivariate analysis was performed, adjusting for HIV status, breastfeeding and gender. Microbiome regression-based kernel association test was used for multivariate analysis of β-diversity. Results: An HEU NBF infant was excluded for low read count. There were more females in the HIV+ than HEU (80% vs. 47.4%; p=0.048) group; HIV+ infants had lower mean CD4% (32.6 vs 39.9; p=0.032) and WHO weight-for-age Z score (-1.1 vs -0.5; p=0.042) than HEU. HIV+ infants had higher α-diversity (Chao1 p=0.004; ACE p=0.003; Fig. A, B) and differed significantly by β-diversity (Jaccard p<0.001, unweighted UniFrac p=0.004; Fig. C, D) compared with HEU; even after adjusting for breastfeeding and gender. LEfSe showed taxa differences between HIV+ and HEU from phylum to genus level, with enrichment of Veillonella and Klebsiella genera in HIV+, and Actinomyces, Alloiococcus, Akkermansia, Weeksellaceae genera in HEU. BF infants had significantly lower α-diversity and differed by β-diversity (all measures p<0.05 after adjusting for HIV infection status and gender) compared with NBF infants.

CROI 2019 305