Evaluation of Two PCR-Hybridization Methods for the Detection of Mycobacterium tuberculosis

Kim, Jin Hee; Jang, Sook Jin; Moon, Dae Soo; Park, Young Jin

Abstract

BACKGROUND
We evaluated two homemade polymerase chain reaction-hybridization (PCRH) assays for the detection of M. tuberculosis. METHODS: Two PCRH assays, using a digoxigenin-luminescent probe (DL-PCRH) and using a radioactive probe (R-PCRH) were developed and compared. To determine the detection limit of each assay, 10-fold serial dilution samples of M. tuberculosis DNA were tested. To determine the specificity of each assay, 4 nontuberculous mycobacteria (NTM) and 13 bacteria other than mycobacteria were tested. Sputum samples from 50 patients with tuberculosis and 26 patients with nontuberculous diseases were tested by DL-PCRH, R-PCRH, culture, AFB stain and PCR assay methods, respectively. RESULTS: Both of the DL-PCRH and R-PCRH methods showed the same detection limit of 100 fg of M. tuberculosis DNA which was a log higher than that for the PCR. Sensitivity of both methods for the detection of M. tuberculosis in sputum specimens was 84%, which were slightly inferior to that of the culture (90%) and similar to that of the PCR (82%). However, both DL-PCRH and R-PCRH methods correctly identified M. tuberculosis for all six specimens that showed weakly-positive or negative PCR results. No false-positive results were obtained for patients with nontuberculous diseases (n=26), NTM (n=4) and other bacteria (n=13). CONCLUSIONS: Although the diagnostic efficiency of the two PCRH assays were similar to that of the PCR, both methods have greater objectivity in reading results than the PCR. This suggests that the PCRH is suitable for use in large number of clinical samples for the rapid detection of M. tuberculosis.