Role of Flexible Bronchoscopy in the Diagnosis of Pulmonary Tuberculosis in HIV-Infected Patients

Abstract

We attempted to determine the diagnostic yield of expectorated sputum, bronchoalveolar lavage (BAL) fluid, and bronchoscopic biopsy (BB) obtained during flexible bronchoscopy (FFB) for Mycobacterium tuberculosis (M.tb) in human immunodeficiency virus (HIV)-infected and non-HIV-infected patients. The role of FFB in making a presumptive and final diagnosis of M.tb infection in the two groups was evaluated. Medical records and chest radiographs of 227 patients who had M.tb cultured from sputum or FFB specimen from July 1990 to June 1992 were reviewed. Patients were divided into HIV-infected and non-HIV-infected groups. Chest radiographs were interpreted as typical (upper lobe, miliary or cavitary infiltrates, pleural effusion, or adenopathy) and nontypical (normal chest x-ray, focal infiltrates not in the upper lobes, interstitial or diffuse, patchy infiltrates) for pulmonary tuberculosis (TB). There were no differences between HIV-infected and non-HIV-infected patients when the overall yield of sputum. BAL fluid, and BB for M.tb were compared. In patients who had both sputum and BAL fluid available for culture, spulum was as sensitive as BAL fluid for final M.tb recovery in both HIV-infected and non-HIV-infected patients. FFB was useful in making a rapid presumptive diagnosis of mycobacterial disease [presence of acid-fast bacilli (AFB) in BAL fluid or biopsy specimen or granulomas on biopsy] in 20% of HIV-infected and 40% of non-HIV-infected patients (p = ns). There were no differences between HIV-infected and non-HIV-infected groups in the presence of typical and nontypical chest radiographs for pulmonary TB. In the HIV-infected group, sputum smears were positive for AFB in 69% of patients with chest radiographs typical for pulmonary TB, compared to 41.7% in patients with non-typical chest radiographs (p = 0.01). This difference was not significant in the non-HIV-infected group. In conclusion, sputum specimens have a high yield for making a final diagnosis of pulmonary TB in both HIV-infected and non-HIV-infected patients. If patients cannot produce sputum. FFB should be performed. If sputum smears are negative in patients in whom a rapid presumptive diagnosis of pulmonary TB is required. BB in addition to BAL should be done. HIV-infected patients do not differ from non-HIV-infected patients in the presence of chest radio-graphic features typical for pulmonary TB.