CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

765 EARLY POSTPARTUM VIREMIA PREDICTS LONG-TERM NONSUPPRESSION AND INFANT TRANSMISSION Megan Landes 1 , Monique Van Lettow 2 , Joep J. Van Oosterhout 2 , Erik Schouten 3 , Thokozani Kalua 4 , Andreas Jahn 4 , Beth A.Tippett Barr 5 1 University of Toronto, Toronto, ON, Canada, 2 Dignitas International, Zomba, Malawi, 3 Management Sciences for Health, Lilongwe, Malawi, 4 Government of Malawi Ministry of Health, Lilongwe, Malawi, 5 US CDC Kisumu, Kisumu, Kenya Background: Long-term viral load (VL) suppression among HIV-positive reproductive-aged women on antiretroviral therapy (ART) is key to eliminating infant HIV transmission. We report trends in postpartum (PP) VL for Malawian women on ART, factors associated with detectable VL, and associations with cumulative infant HIV transmission and/or death. Methods: From 2014-2016, 4-26 week PP HIV-positive mothers were screened and enrolled with their infants in Malawi clinics. At enrollment, 12 and 24 months PP, socio-demographic and prevention of mother to child transmission of HIV (PMTCT) indicators were collected and infants had HIV-1 DNA testing. Venous samples determined maternal plasma VL (<40 copies/ml = ‘undetectable’); standard national VL monitoring guidelines were in early implementation across Malawi. Results: 585 women were retained to 24 months (N=1281,45.7%), and were more likely to be >30 years (51.6 vs 41.4%, p<0.01), parity =>4 (41.0 vs 33.5%, p=0.02) and have undetectable VL at enrollment (79.7 vs 70.8%, p<0.01) than those lost to follow-up (LTFU). Of 573 women on ART (median 29.7 mos.(IQR 26.8-61.3)), 424 (74.0%) with VL data at all 3 visits were included in analysis. Table 1 shows 341 (80.4%) women had durable undetectable VL, 83 (19.5%) had >= 1 detectable VL and 32 (7.5%) had persistent detectable VL. Cumulative incidence of detectable VLs over 24 months was higher among women with detectable VL at enrollment than with undetectable VL (74 detectable VL measures/67 women vs 19/357; p<0.01). In multivariable analysis, adjusted odds ratios for detectable VL at 24 months were 10.1 among women with 1 prior detectable VL (95%CI 3.7-28.1, p<0.001) and 240.4 for women with 2 prior detectable VLs (95%CI 68.0-849.8, p<0.001) (adjusted for age, parity, education, partner disclosure, time on ART and adherence). Women with durable undetectable VL (N=341) had no infant transmissions and 2 infant deaths (0.6% combined outcome). Women with persistent detectable VL (N=32) had 4 infant transmissions and 1 infant death (15.6%; p<0.01). Conclusion: Detectable VL in early PP signals a risk of ongoing PP viremia with major implications for infant HIV transmission, in a setting with limited high VL management. These findings suggest differentiated VL monitoring and targeted adherence support may be required during pregnancy and breastfeeding.

764 DETECTABLE HIV RNA IN LATE PREGNANCY ASSOCIATED WITH LOW TFV HAIR LEVELS AT BIRTH Jillian Pintye 1 , Yanling Huo 2 , Deborah Kacanek 2 , Karen Kuncze 3 , Kevin Zhang 3 , Hideaki Okochi 3 , Monica Gandhi 3 1 University of Washington, Seattle, WA, USA, 2 Harvard T.H. Chan School of Public Health, Boston, MA, USA, 3 University of California San Francisco, San Francisco, CA, USA Background: Adherence to antiretroviral therapy (ART) is vital to prevention of mother-to-child transmission of HIV (PMTCT) and maternal health, although peripartum life events can disrupt adherence. Hair levels measure cumulative ART exposure and are associated with viral suppression in nonpregnant and postnatal populations. We evaluated correlates of peripartum tenofovir (TFV) exposure via hair measures among women living with HIV (WLHIV) in the United States. Methods: Hair samples were collected at or shortly after childbirth among WLHIV enrolled in the Surveillance Monitoring for ART Toxicities Study of the Pediatric HIV/AIDS Cohort Study between 6/2014-7/2016. Among WLHIV on TFV-based regimens during pregnancy, TFV hair levels were analyzed using validated liquid chromatography/tandemmass spectrometry methods. Weight- normalized TFV hair concentrations were log transformed. Correlates of TFV hair concentrations were identified using multivariable linear regression. Covariates with p<0.25 in univariable models were included in multivariable models. Results: Among 370 WLHIV with TFV-based ART use during pregnancy, hair collection acceptability was high (only 65/370 [18%] of all WLHIV using TFV declined); 111 women had TFV hair levels and were included in the final analysis. Median age at delivery among the 111 WLHIV was 31 years (IQR 26-36); 70% self-identified as non-Hispanic black, 71% had achieved high school graduation, 13% reported recreational drug use during pregnancy, and 9% had unsuppressed viral loads (VL) in late pregnancy, defined as HIV-RNA ≥400 copies/mL. The median time from birth to hair collection was 4 days (IQR 1-14) and 31% of samples had TFV hair levels ≥0.038 ng/mg (equivalent to 7 doses/ week). In multivariable models (Table 1), an unsuppressed VL in late pregnancy was most strongly associated with lower peripartum TFV hair levels. Attainment of high school education and not using TFV-based ART after the 1st trimester were also independently associated with lower peripartum TFV levels. Conclusion: Unsuppressed VL among WLHIV in the U.S. during late pregnancy, a critical period for PMTCT, was strongly associated with lowmaternal TFV hair levels at birth. Over two-thirds of WLHIV had hair levels suggestive of imperfect adherence although viremia in late pregnancy was rare (9%). Efforts to improve PMTCT outcomes could incorporate drug exposure monitoring using hair or other metrics and include adherence promotion strategies that address issues unique to the peripartum period.

Poster Abstracts

CROI 2020 282