Rapid detection of Mycobacterium tuberculosis in various clinical specimens by using polymerase chain reaction combined with a nonradioactive hybridization system.

A DNA amplification system using the polymerase chain reaction (PCR) combined with a nonradioactive digoxigenin-labeled probe hybridization was employed to detect Mycobacterium tuberculosis in clinical specimens. One hundred and thirty specimens were tested by several methods including routine culture method, acid-fast staining, BACTEC 460 detection system, PCR, and PCR-hybridization techniques. Sixteen out of 130 specimens were culture positive on Middlebrook 7H11 agar, 10 were positive with acid-fast staining, 18 were positive with BACTEC 460 detection system, 23 were positive with PCR technique, and 62 were positive with PCR-nonradioactive hybridization technique. When compared with culture results, PCR-nonradioactive hybridization had an overall sensitivity of 100% (16/16) and a specificity of 59.7% (68/114). However, 28 out of 46 (60.9%) PCR-nonradioactive hybridization positive specimens which were culture negative had clinical data supporting the diagnosis of tuberculosis. In addition, 4 specimens which were negative by routine culture but positive by BACTEC 460 detection system and two specimens which were negative by routine culture but positive by acid-fast staining were all positive by PCR-hybridization technique. These data suggest that routine culture method may not be sensitive enough to detect M. tuberculosis in all kinds of clinical specimens. Taking this deviation into account, the specificity of PCR-nonradioactive hybridization technique may be rectified range from 63% (68/108) to 79.1% (68/86). PCR itself is not satisfactory enough to detect M. tuberculosis in specimens (the sensitivity and specificity were 56.3% and 87.7%, respectively) in this study. However, when it combines with DNA hybridization technique, they can be a very powerful and rapid diagnostic tool to detect M. tuberculosis in clinical specimens.