CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

modest but significant differences were seen as early as 4 months with increases through 12 months (Fig. 1). Estimated treatment differences during follow-up were: VAS=1.9 (95% CI 1.2-2.5); GHP=3.6 (2.8-4.5); PCS=0.8 (0.5-1.1); MCS=0.9 (0.4-1.3) (p<0.001 for each QOL measure). Conclusions: In this HIV-positive population with CD4 >500 cells/mm3 that was generally in good health, all QOL measures improved in the immediate compared to the deferred ART group. These findings provide further support to the superiority of early ART as reported for major clinical outcomes in the START study.

476 Gender and Racial Disparities in Initial Antiretroviral Treatment Outcome: ACTG A5257 Heather J. Ribaudo 1 ; Raphael J. Landovitz 2 ; Lumine H. Na 3 ; Jeffrey L. Lennox 4 ; Ighovwerha Ofotokun 4 ; Susan E. Cohn 5 ; Daniel Kuritzkes 6 ; Judith S. Currier 2 ; for the ACTG A5257 Study Team 1 Harvard Sch of PH, Boston, MA, USA; 2 David Geffen Sch of Med at Univ of California Los Angeles, Los Angeles, CA, USA; 3 Univ of Edinburgh, Edinburgh, UK; 4 Emory Univ Sch of Med, Atlanta, GA, USA; 5 Northwestern Univ, Chicago, IL, USA; 6 Harvard Med Sch, Boston, MA, USA Background: Inferior virologic outcomes of initial antiretroviral treatment (ART) for women and ethnic/racial minority groups have been reported. We examined this finding in a US-based setting with modern ART regimens and determined whether socio-demographics or non-adherence explain these differences. Methods: ACTG A5257 randomized 1809 participants to ART with emtricitabine/tenofovir disoproxil fumarate plus atazanavir/ritonavir (/r), raltegravir, or darunavir/r. Study follow-up ended in 2013. This planned secondary analysis included 1762 participants categorized as non-Hispanic white, non-Hispanic black, and Hispanic based on self report. Cox proportional hazards models examined the association of sex and race/ethnicity on the hazard of virologic failure (VF) adjusting for demographic and baseline factors including age, HIV-1 disease status and socio-demographic variables; adherence by self-report was included as a time-updated covariate. Analyses were intent to treat. Results: The study sample was 34%white, 43% black, and 22% Hispanic. Median age was 37 years; 24%were women. Probability of VF by 96 weeks [95%CI] was 11% [10, 13] in men and 16% [12, 19] in women. VF probability was 7% [5, 9] in whites, 17% [14, 20] in blacks and 13% [10, 17] in Hispanics. Differential effects of sex or race/ethnicity by treatment armwere not apparent (P>0.40). A greater VF risk for women compared to men was apparent in unadjusted analysis (P=0.005) but not after adjustment for race/ ethnicity (P=0.29). Compared to whites, blacks and Hispanics had a greater hazard of VF (unadjusted HR=2.8 [2.0, 3.8] and 2.0 [1.4, 2.8], respectively). While adjustment for socio-demographic factors appeared to account for the excess VF risk for Hispanics (1.2 [0.7, 1.8]), an excess risk remained for blacks after adjustment for socio-demographics and adherence (1.7 [1.1, 2.5]). Other factors associated with higher VF risk included non-adherence, younger age, high pre-ART viral load, low income, less education, IV drug history, and not working or in school (Figure). Conclusions: Women and ethnic/racial minority groups in the US remain at greater risk of VF of initial ART with modern ART regimens. For women and Hispanics, this excess risk appears explained by race/ethnicity and socio-demographics, respectively. In contrast, an excess risk of VF for blacks remains after adjustment for non-adherence and socio- demographics. This work helps define populations at high VF risk who may benefit from early targeted interventions.

Poster Abstracts

185

CROI 2016