Abstract
To determine the clinical utility of a PCR-based test kit(AMPLICOR^<TM> Mycobacterium tuberculosis assay, Roche Molecular Systems)for bronchoalveolar lavage fluid(BALF), we retrospectively evaluated the efficacy of the commercially developed PCR test system comparing it with the results of conventional bacteriological methods. A total of 207 specimens(tuberculosis : 17, non-tuberculous mycobacterium(NTM) : 19, lung cancer : 50, pneumonia : 18, bronchiolitis : 11, metastatic lung tumor : 9, others : 83)from 202 patients were tested. When total tuberculosis, pulmonary tuberculosis and culture results were used as the reference standard, the sensitivity rates for the AMPLICOR test were 12/17(70.6%), 12/12(100%)and 12/13(92.3%), respectively. There were no positive results of AMPLICOR among non-tuberculous mycobacterium(19), lung cancer(50)and the specimens(45)in which acid-fast bacilli were positive because an automated bronchoscope-cleaning machine and the used bronchoscopes were transiently contaminated by NTM. The specificity for the AMPLICOR test was 97.9% using the total tuberculosis as the standard. The PCR test had a false positive rate of 2.1%. Although the test results from AMPLICOR could be obtained more quickly than that from culture, the clinical utility of this PCR-based test is similar but not superior to that of culture for detecting M. tuberculosis in BALF in terms of sensitivity and specificity.
