CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

Conclusion: Among 234 patients with WHO-defined TB symptoms, 88% did not have TB but experienced an unnecessary delay in ART initiation. Requiring TB test results for all symptomatic patients prior to ART initiation, without consideration of symptom number or severity, should be reconsidered.

722 ALCOHOL USE IS ASSOCIATED WITH INCIDENT TB INFECTION IN HIV+ AND HIV– UGANDAN ADULTS Sarah B. Puryear 1 , Atukunda Mucunguzi 2 , Laura B. Balzer 3 , Joel Kironde 2 , Judith A. Hahn 1 , Florence Mwangwa 2 , Dalsone Kwarisiima 2 , Moses R. Kamya 2 , Maya L.Petersen 4 , Edwin D. Charlebois 1 , Diane V. Havlir 1 , Gabriel Chamie1, Carina Marquez 1 1 University of California San Francisco, San Francisco, CA, USA, 2 Infectious Diseases Research Collaboration, Kampala, Uganda, 3 University of Massachusetts Amherst, Amherst, MA, USA, 4 University of California Berkeley, Berkeley, CA, USA Background: Globally, an estimated 10% of TB disease is attributable to alcohol use. Alcohol users may be at increased risk of TB disease as a result of spending more time in social venues (e.g. bars) where TB transmission is high, however the relationship between alcohol use and incident TB infection is unknown. Methods: We assessed this association in a longitudinal cohort of tuberculin skin test (TST) negative adults nested in the SEARCH study (NCT:01864603) in Eastern Uganda. Baseline (2015-16) TSTs were placed in 2,940 adults participating in a household survey, enriched for persons living with HIV. Participants with no TST induration were eligible for TST testing a year later. Our primary outcome, incident TB infection, was defined as a change in TST induration from 0mm to >5mm if HIV+ or >10mm if HIV- at reassessment. Alcohol use was assessed using the Alcohol Use Disorders Test-C (AUDIT-C). Exposure variables were: (1) any alcohol use (AUDIT-C>0) vs. no use and (2) hazardous use (AUDIT-C >3 for women, >4 for men) vs. non-drinking/ non-hazardous use. We calculated odds ratios using generalized estimating equations and used inverse probability weighting to account for incomplete measures. All models were adjusted for age, gender, household wealth, HIV status, and household TB contact. Results: One-year follow-up TSTs were completed in 1,047 (58%) of the 1,814 adults with a negative TST at baseline. Among those who completed TSTs, 84 (8%) reported alcohol use, 36 (3%) reported hazardous alcohol use, 269 (26%) were living with HIV, and 21 (2%) reported a household TB contact in the year prior. At follow-up, 177 (17%) met our definition of incident TB infection. Incident TB infection was more common in persons who reported any alcohol use compared to no use (27% vs. 16%) and alcohol use was positively associated with incident TB infection (aOR 2.0, 95% CI: 1.0-3.8, p=0.04). Hazardous alcohol use was associated with incident TB infection (aOR 2.8, 95%CI: 1.1-7.1, p=0.03) compared non-drinkers/non-hazardous drinkers. There was no association between incident TB infection and HIV status or having a household TB contact. Conclusion: In this longitudinal cohort of adults in Uganda, incident TB infection was high, positively associated with alcohol use at any and hazardous levels, and not associated with HIV-status or a known household TB contact. TB prevention efforts that focus on reducing transmission in venues shared by drinkers may decrease the latent TB reservoir in this TB risk group. 723 TUBERCULOSIS EVALUATION AMONG HIV-POSITIVE PATIENTS ON ANTIRETROVIRAL THERAPY Meaghan L. Peterson 1 , Catherine Nichols 2 , Rena Fukunaga 1 , Joseph S. Cavanaugh 1 , Patricia Hall 1 , Erin Rottinghaus 1 , N.Sarita Shah 1 , AdamMacNeil 1 1 CDC, Atlanta, GA, USA, 2 United States Agency for International Development, Washington, DC, USA Background: The World Health Organization estimates nearly 500,000 cases of tuberculosis (TB) among people living with HIV (PLHIV) go unreported each year. Among PLHIV, four-symptom TB screening (cough, fever, weight loss, and night sweats) is recommended at every clinical encounter, followed by sputum testing with Xpert MTB/RIF for positive screens. We assessed TB screening programs in countries supported by the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR).

721 HOUSEHOLD AIR POLLUTION INCREASES RISK FOR PULMONARY TB IN HIV-INFECTED ADULTS Patrick Katoto 1 , Bihehe Masemo 2 , Amanda S. Brand 3 , Aline Kusinza 4 , Brian Allwood 3 , Richard Van Zyl-Smit 5 , Nadia A. Sam-Agudu 6 , David Dowdy 7 , John Z. Metcalfe 8 , Grant Theron 3 , Kevin Mortimer 9 , Jeroein Vanoirbeek 1 , Tim Nawrot 10 , Benoit Nemery 1 , Jean B. Nachega 11 1 Katholieke University Leuven, Leuven, Belgium, 2 Université Évangélique en Afrique, Bukavu, Congo, 3 Stellenbosch University, Cape Town, South Africa, 4 Université Evangelique en Afrique, Bukavu, Congo, 5 University of Cape Town, Cape Town, South Africa, 6 University of Maryland, Baltimore, MD, USA, 7 Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA, 8University of California San Francisco, San Francisco, CA, USA, 9Liverpool School of Tropical Medicine, Liverpool, UK, 10University of Hasselt, Hasselt, Belgium, 11University of Pittsburgh, Pittsburgh, PA, USA Background: Millions of deaths related to household air pollution (HAP), pulmonary tuberculosis (PTB), and HIV occur annually in low income countries. However, little is known about the influence of HAP on PTB risk among people living with HIV (PLHIV). Methods: We conducted a case-control study among PLHIV at four clinics in eastern Democratic Republic of Congo (DRC) fromMarch 2018 to February 2019. Cases were ≥18 years old, with recent (≤5 years) or current PTB. Controls were age- and sex-matched PLHIV with neither recent nor current PTB. During home visits, HAP exposure was assessed using a validated International Multidisciplinary Programme to Address Lung Health and TB in Africa (IMPALA) questionnaire. Personal carbon monoxide (CO) exposure was assessed using the EasyLog USB CO Lascar Monitor, and volatile organic compound exposure using Radiello® passive-diffusive sampler. Urinary 1 hydroxypyrene and S-phenyl- mercapturic acid were measured. Bacteriologic confirmation of PTB (sputum smear or Xpert MTB/RIF positive), CD4 count, and antiretroviral treatment (ART) history were extracted frommedical records. Conditional multivariate logistic regression was performed to assess independent associations between HAP and PTB. Results: We recruited 435 cases and 842 controls. Median age (IQR) 41 years (33-50), 76% female. Overall median 24h-personal average CO was 5.3 [2.3-10.6] parts per million (ppm). After adjusting for sociodemographic covariates, tobacco smoking, median CD4 count, and duration on ART, each 1 ppm increase in average 24h CO exposure was positively associated with PTB (adjusted odds ratio, aOR; 95% confidence interval, CI: 1.5; 1.01-2.23). Average 24h CO level stratification by quintiles yielded a concentration dependent increase in the odds of PTB from the lowest [0.1-1.9 ppm], to highest quintile [12.3-76.2 ppm] (aOR 4.64; 95%CI: 1.04-20.65) (Fig. 1). Furthermore, for women, each additional hour spent cooking over wood fire was associated with increased odds of PTB (aOR 2.76; 95% CI: 1.02-7.47). Conclusion: Personal CO exposure and time spent cooking over wood fire (among women) were independently associated with increased odds of PTB among PLHIV in eastern DRC. Longitudinal studies are needed to confirm our findings and inform comprehensive strategies to reduce the triple burden of HAP-TB-HIV.

Poster Abstracts

CROI 2020 265