CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

Conclusion: Children followed by clinical or CD4-only monitoring had significantly longer time to switch to second-line as compared to routine or targeted VL monitoring. This likely represents delayed or under recognition of treatment failure, which has significant potential to impact their long term outcomes. The scale up of VL monitoring is likely to increase detection of treatment failure and demand for second-line ART. 816 FREQUENT FAILURE AFTER SECOND-LINE ART IN HIV-POSITIVE CHILDREN FROM LATIN AMERICA Bryan E. Shepherd 1 , Cathy Jenkins 1 , Regina Succi 2 , Daisy Machado 2 , Jorge Pinto 3 , Vanessa Rouzier 4 , Marco Luque 5 , Sten Vermund 1 , Catherine McGowan 1 , for the Caribbean, Central and South America Network for HIV epidemiology (CCASAnet) 1 Vanderbilt Univ, Nashville, TN, USA, 2 Univ Fed de Sao Paulo, São Paulo, Brazil, 3 Univ Fed de Minas Gerais, Belo Horizonte, Brazil, 4 Les Cntrs GHESKIO, Port-au-Prince, Haiti, 5 Hosp Escuela Univ, Tegucigalpa, Honduras Background: Latin America’s diverse population has varying, often limited access to newer antiretrovirals. Data on pediatric second-line ART outcomes and their predictors are needed to guide design of optimal interventions. Methods: HIV-positive ART-naïve children age ≤18, starting cART (cART 1 ) from 1999-2007 at 5 sites in Argentina, Brazil, Haiti, and Honduras, and changing to second-line cART (cART 2 ). Outcomes: cART 2 changes (≥2 agent change, outside-class substitution, or discontinuation); virologic failure on cART 2 ; all-cause mortality. Cumulative incidences were modeled with death as competing risk for non-mortality outcomes. Cox regression was used to estimate hazard ratios (HR) and 95% confidence intervals (CI) adjusting for sex, mode of infection, age, CD4, AIDS, year at cART 2 , change in CD4 and time from cART 1 to cART 2 , cART 2 start reason, cART 2 class, and site. Haiti was not included in failure analyses. Results: Of 1489 children, 413 (28%) started cART 2 ; most were perinatally-infected (93%), 52%were female, and 38% had AIDS at cART 1 start. Median ages were 5.1 (interquartile range [IQR] 1.8-10.1) and 9.2 (IQR 5.4-14.2) years and CD4s were 426 (IQR 139-878) and 456 (IQR 133-916) cells/µL at cART 1 and cART 2 starts, respectively. Median year of cART 2 start was 2007 (IQR 2005-2011). Most (56%) received protease inhibitor-based cART 2 . On cART 2 , median follow-up was 4.7 years (IQR 1.9-7.8), 202 (49%) changed regimens, 53 (13%) died, and 80 (19%) were lost to follow-up. Among 251 children with viral loads (VL), cumulative incidence of failure at 3 years after cART 2 start was 0.65 (CI=0.58-0.71, Figure). Older age at cART 2 start (HR=1.06 per year; CI=1.01-1.10) and failure as reason for starting cART 2 (HR=2.14; CI=1.24-3.71) increased risk of changing cART 2 . Children with AIDS at cART 1 start had higher risk of cART 2 failure (HR=2.24; CI=1.22-4.13), though those starting cART 2 in later years had decreased risk (HR=0.84 per year; CI=0.73-0.97). Conclusion: We observed high rates of switching from cART 2 in varied Latin American settings; 65% of children with VL data experienced treatment failure by 3 years. Interventions targeting adolescents and improved access to once-daily cART regimens are needed to improve outcomes in this population.

Poster and Themed Discussion Abstracts

817LB HIV-INFECTED CHILDREN WITH SEVERE ACUTE MALNUTRITION: EARLY VS DELAYED ART INITIATION Moherndran Archary 1 , Benn Sartorius 1 , Philip La Russa 2 , thobekile sibaya 1 , Micheal Healy 2 , Raziya Bobat 1 1 Univ of KwaZulu-Natal, Durban, South Africa, 2 Columbia Univ, New York, NY, USA

Background: Delays in prompt HIV diagnosis and ART initiation in children from low and middle-income countries, frequently results in malnutrition at initial presentation. Despite ART initiation, HIV positive children with malnutrition have a higher mortality and delayed immune recovery. The optimal timing of ART initiation in children with malnutrition has not been established Methods: Eighty-two HIV infected children with severe acute malnutrition (SAM) admitted to King Edward VIII Hospital between July 2012 and December 2015 were enrolled. Patients were randomized to initiate ART within 14 days from admission (Early arm) or delay ART initiation until nutritional recovery and more than 14 days from admission (Delayed arm). All patients received a standardized treatment and feeding protocol and were evaluated at 4, 8, 12, 24 and 48 weeks. Results: The average age of the patients at baseline was 23·3 months (SD 27·9, range 1.6–129 months). The mean time from admission to ART initiation was 5·6 days (SD 4·4) in the early arm and 23 days (SD 5·8) in the delayed arm (p<0·001). There was no significant difference in mortality (p=0·621), virologic response (p=0·527) and anthropometric response (p= 0·566) between the two groups at 48 weeks. However the rates of change in CD4, viral load (figure 1), WAZ and HAZ scores occurred earlier and favored the delayed arm. Conclusion: HIV–infected children admitted with SAM and initiated on ART demonstrated significant improvements in CD4 counts and anthropometric parameters, together with significant viral load reduction compared to baseline. In this randomized controlled trial comparing early versus delayed ART initiation in HIV infected children admitted with SAM, although the differences in CD4 count, viral suppression and anthropometric response at 48 weeks was not significant, the rates of change in CD4, viral load, WAZ and HAZ

CROI 2017 353