CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

Conclusions: In HIV-positive adults receiving care at Ethiopian health centers mortality was associated with reduced performance score and malnutrition, with different distribution with regard to gender and TB co-infection at inclusion. These robust variables could be used at clinic registration to identify persons at increased risk of early mortality.

745 Empiric TB Therapy Versus IPT in HIV-Infected Persons Initiating ART (ACTG A5274 48 w) Johnstone J. Kumwenda 1 ; Amita Gupta 2 ; Xin Sung 3 ; Miyahara Sachiko 3 ; Evelyn Hogg 4 ; Lynne Jones 5 ; Andrew Zolopa 6 ; Gregory P. Bisson 7 ; Mina Hosseinipour 8

1 Coll of Med John Hopkins Proj, Blantyre, Malawi; 2 Johns Hopkins Univ Sch of Med, Baltimore, MD, USA; 3 Harvard Univ, Boston, MA, USA; 4 Social & Scientific Systems, Inc, Silver Spring, MD, USA; 5 Frontier Sci & Tech Rsr Fndn, Inc, Amherst, NY, USA; 6 Stanford Univ, Stanford, CA, USA; 7 Univ of Pennsylvania, Philadelphia, PA, USA; 8 Univ of North Carolina at Chapel Hill, Chapel Hill, NC, USA Background: Up to 26% of patients initiating ART in low and middle income countriesbdie within the first year of initiating ART, with TB being a major contributor. Strategies to reduce TB/HIV- associated mortality are urgently needed. Methods: A5274 REMEMBER study is a multi-country randomized clinical trial comparing ART + four-drug empiric TB therapy vs. ART + isoniazid preventive therapy (IPT) in HIV-infected individuals with CD4 counts <50 cells/mm 3 . Participants were screened for TB prior to entry using a symptom screen, locally available diagnostics, and GeneXpert when available. Randomization was stratified by CD4 count (<25 vs. ≥25 cells/mm 3 ) and poor prognostic factors (body mass index <18.5, hemoglobin <8 g/dl, and recent hospitalization). The primary endpoint of survival at 24 weeks post-randomization was reported elsewhere and demonstrated no effect of empiric TB treatment on mortality. To evaluate the possible effects of the intervention on longer-term outcomes, we compared the probabilities of death, death or AIDS progression, and confirmed or probable TB by week 48 between arms. Kaplan-Meier method was used to estimate the endpoint probabilities, which were compared with the z-test. Results: We screened 1368 individuals and enrolled 850 (62%) participants. Of 850 enrolled, 53%were male, 90%were black. The median (IQR) age was 36 (30-42) years. The median (IQR) baseline CD4 count was 18 (9, 32) cells/mm 3 . At week 48, there was no statistical difference in mortality between the empiric arm (7.2%; 95% CI: 5.1%, 10.1%) and the IPT arm (8.7%; 95% CI: 6.4%, 11.8%), absolute risk difference 1.6% (95% CI: -2.1%, 5.2%; p=0.41). As in the 24-week analysis, the probability of death or AIDS progression was not significantly different between arms [19.3% (95% CI: 15.8%, 23.4%) for the empiric arm vs. 15.3% (95% CI: 12.2%, 19.1%) for the IPT arm], absolute difference -4.0% (95% CI: -9.1%, 1.1%; p=0.13). At week 48, the empiric arm had more TB compared to the IPT arm (5.6% vs. 2.4%, respectively), absolute risk difference -3.2% (95% CI:-5.9%, -0.5%; p=0.02; unchanged in competing risk analysis). Conclusions: In a high TB burden population of participants with advanced HIV, there was no demonstrable benefit of empiric TB therapy on mortality at 48 weeks. These data support the implementation of enhanced screening for TB prior to ART initiation and the use of IPT, even in persons with advanced HIV residing in high TB burden regions of the world. Time to Death by Treatment Strategy

Poster Abstracts

746 TB Treatment Outcomes for HIV/TB Coinfected Children in Resource-Limited Countries

James G. Carlucci 1 ; Aaron M. Kipp 1 ; Meridith Blevins 2 ; QuyT. Du 3 ; Lorna Renner 4 ; Gary Reubenson 5 ; John Ssali 6 ; MarcelYotebieng 7 ; Mary Lou Lindegren 2 ; April C. Pettit 1 1 Vanderbilt Univ Sch of Med, Nashville, TN, USA; 2 Vanderbilt Inst for Global Hlth, Nashville, TN, USA; 3 Children’s Hosp 1, Ho Chi Minh City, Vietnam; 4 Univ of Ghana Sch of Med and Dentistry, Accra, Ghana; 5 Rahima Moosa Mother and Child Hosp, Johannesburg, South Africa; 6 Masaka Regional Referral Hosp, Masaka, Uganda; 7 Ohio State Univ, Columbus, OH, USA Background: Diagnosis and treatment of tuberculosis (TB) are challenging in HIV/TB co-infected children. The World Health Organization (WHO) recommends nucleic acid amplification tests (NAAT) for TB diagnosis, and a 4-drug regimen including ethambutol during the intensive phase of treatment (IP). However, many children are diagnosed with TB based on clinical criteria without microbiological confirmation, and ethambutol is sometimes omitted from IP due to concerns about toxicity. We assessed whether TB treatment outcomes differed by mode of diagnosis or IP regimen. Methods: We conducted an observational cohort study among HIV/TB co-infected children (<16 years) enrolled at HIV treatment sites in 5 regions of the International Epidemiologic Databases to Evaluate AIDS (IeDEA) consortium from 2012-2014. Modified Poisson regression was used to estimate the relative risk (RR) of an unfavorable TB outcome (death, treatment failure, default, or unknown). Results: Data were collected from 372 children. Excluding those missing CD4 counts, 295 were included in the main analysis. Median age was 5.7 years (Interquartile Range [IQR] 2-9.6), 46%were female, median CD4 count was 252 (IQR 60-666), 22% had unfavorable TB outcomes, 21% had at least one positive TB test (smear, culture, or NAAT), and 79%

310

CROI 2016