Objective: To determine whether the detection of tuberculostearic acid (TBSA) in bronchial aspirate and bronchoalveolar lavage specimens is useful for the rapid diagnosis of active pulmonary tuberculosis in patients suspected of having the disease.
Setting: A pulmonary clinic in a teaching hospital.
Patients: Forty patients suspected of active pulmonary tuberculosis but who failed to produce sputum or whose sputum smears were negative for acid-fast bacilli on at least 3 occasions, 29 of whom were subsequently confirmed to have tuberculosis. A group of 13 patients who were having fiberoptic bronchoscopy for other reasons served as controls.
Intervention: All patients had fiberoptic bronchoscopy; bronchial aspirate, bronchoalveolar lavage, and sputum specimens were obtained when possible.
Measurements and Main Results: All specimens were examined microscopically for acid-fast bacilli, cultured for mycobacteria, and assayed for TBSA by gas chromatography and mass spectrometry with selected ion monitoring. Only 4 of the 29 patients with tuberculosis were diagnosed by direct microscopy compared with 26 by TBSA assay. In 2 patients who required surgical biopsy for conventional diagnosis, the TBSA test was positive. There were no false-positive TBSA results in the 13 controls, but 2 of 5 sputum specimens from the 11 test patients in whom tuberculosis was excluded were falsely positive, probably because of contamination with mouth flora. Because sputum can rarely be obtained from these patients and may give false-positive results, it is not a good specimen for TBSA assay. Sensitivities and specificities of the test for the other specimens were as follows: aspirate, 0. 52 (CI, 0.32 to 0.71) and 1.00 (CI, 0.75 to 1.00); lavage, 0.68 (CI, 0.46 to 0.85) and 1.00 (CI, 0.84 to 1.00); aspirate and lavage combined, 0.79 (CI, 0.60 to 0.92) and 1.00 (CI, 0.86 to 1.00).
Conclusions: The TBSA assay of bronchial aspirate and bronchoalveolar lavage fluid is useful for rapidly diagnosing "smear-negative" pulmonary tuberculosis. In these specimens it is highly specific and more sensitive than microscopy. This assay could be used to diagnose other mycobacterial infections; however, it cannot distinguish among species.