BACKGROUND: Polymerase chain reaction (PCR) is a recent, rapid and reliable method in the detection of causative organism. The authors tried to determine the possibility of using PCR technique as an alternative way to detect mycobacterial DNA from paraffin-embedded tissue to avoid repeated biopsy from the patient. MATERIAL AND METHOD: Paraffin-embedded tissue blocks, the corresponding histopathologic slides, and cultural results were retrospectively searched for according to the patient's records, the granuloma clinic, Department of Dermatology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand from 1994-2000. One hundred and thirty-one tissue blocks and slides were found but only 120 cultural results were retrieved Histologic sections were reviewed for AFB findings and PCR was done using 16S rRNA sequences to detect M. tuberculosis by one-tube nested technique and multiplex PCR for M. marinum and M. fortuitum complex. RESULTS: The causative organisms were identified by AFB staining in pathologic sections 31.29%, by PCR 35.87%, and by culture 30.00% of tested samples. The sensitivity of PCR when compared to AFB result was 29.26%, specificity 61.11% but when compared to cultural results, the sensitivity of PCR was 66.67% and AFB sensitivity was 41.66% with specificity 76.19% and 72.61% respectively. CONCLUSION: The low sensitivity of the PCR method may be due to formalin fixation, deparaffinization process, DNA extraction method, the use of 16S rRNA-based primers and the length of the expected product, and the tissue type that may have Taq polymerase inhibitor. Therefore, PCR should be used to augment the information of the conventional method in the diagnosis of mycobacterial infection.
BACKGROUND: Polymerase chain reaction (PCR) is a recent, rapid and reliable method in the detection of causative organism. The authors tried to determine the possibility of using PCR technique as an alternative way to detect mycobacterial DNA from paraffin-embedded tissue to avoid repeated biopsy from the patient. MATERIAL AND METHOD:Paraffin-embedded tissue blocks, the corresponding histopathologic slides, and cultural results were retrospectively searched for according to the patient's records, the granuloma clinic, Department of Dermatology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand from 1994-2000. One hundred and thirty-one tissue blocks and slides were found but only 120 cultural results were retrieved Histologic sections were reviewed for AFB findings and PCR was done using 16S rRNA sequences to detect M. tuberculosis by one-tube nested technique and multiplex PCR for M. marinum and M. fortuitum complex. RESULTS: The causative organisms were identified by AFB staining in pathologic sections 31.29%, by PCR 35.87%, and by culture 30.00% of tested samples. The sensitivity of PCR when compared to AFB result was 29.26%, specificity 61.11% but when compared to cultural results, the sensitivity of PCR was 66.67% and AFB sensitivity was 41.66% with specificity 76.19% and 72.61% respectively. CONCLUSION: The low sensitivity of the PCR method may be due to formalin fixation, deparaffinization process, DNA extraction method, the use of 16S rRNA-based primers and the length of the expected product, and the tissue type that may have Taq polymerase inhibitor. Therefore, PCR should be used to augment the information of the conventional method in the diagnosis of mycobacterial infection.