High Rates of Clinical and Subclinical Tuberculosis among HIV-Infected Ambulatory Subjects in Tanzania

Abstract

We sought to determine the prevalence of active tuberculosis among ambulatory HIV-infected persons in Tanzania with CD4 cell counts of > or =200 cells/mm3 and a bacille Calmette-Guerin vaccination scar. Subjects who volunteered for a tuberculosis booster vaccine trial were screened for active tuberculosis by obtainment of a history, physical examination, chest radiography, sputum culture and acid fast bacillus (AFB) stain, and blood culture. All subjects underwent a tuberculin skin test (TST) and lymphocyte proliferation assays (LPAs) for detection of responses to mycobacterial antigens. Active tuberculosis was identified at baseline in 14 (15%) of the first 93 subjects who were enrolled: 10 (71%) had clinical tuberculosis (symptoms or chest radiograph findings), and 4 (29%) had subclinical tuberculosis (positive sputum AFB stain or culture results but no symptoms or chest radiograph findings). An additional 6 subjects with subclinical tuberculosis were identified subsequently. The 10 subjects with subclinical tuberculosis included 3 with positive sputum AFB stains results and 7 who were only identified by a positive sputum culture result. Compared with subjects who did not have tuberculosis, the 10 subjects with subclinical tuberculosis were more likely to have peripheral lymphadenopathy, positive TST results, and elevated LPA responses to early secreted antigenic target-6 (ESAT). Eight of 10 patients had received isoniazid because of a positive TST result before active tuberculosis was recognized. Clinical and subclinical tuberculosis are common among ambulatory HIV-infected persons, and some cases can only be identified by sputum culture. World Health Organization guidelines for screening for latent tuberculosis before treatment do not recommend sputum culture and, therefore, may fail to identify a substantial number of HIV-infected persons with subclinical, active tuberculosis.