RESUMEN
In the present study, we made an attempt to compare polymerase chain reaction-single strand conformation polymorphism (PCR-SSCP) with PCR-direct sequence analysis for their accuracy and sensitivity in detecting resistance to rifampicin (RMP). A total of 32 clinical isolates of Mycobacterium tuberculosis including 22 resistant and 10 sensitive isolates, whose drug susceptibility have been tested by conventional proportion method, were analyzed by using PCR-SSCP and PCR-sequence analysis. Among 22 RMP resistant isolates, 16 isolates showed SSCP profiles different from that of a RMP sensitive control strain, M. tuberculosis H37Rv indicating the possible existence of a sequence alteration in this region of the rpoB gene, while 6 resistant isolates displayed SSCP profiles indistinguishable from the sensitive control strain. On the other hand, all of 10 RMP sensitive isolates showed SSCP profiles similar to that of the sensitive control strain. Therefore, overall agreement rste between conventional proportion method and PCR-SSCP reached 81%. Subsequently, all of 32 clinical isolates were subjected to sequence analysis. The results from the sequence analysis revealed that all of 22 resistant isolates indeed contain mutations in the stretch of 81 bp region of rpoB gene, while none of 10 sensitive isolates contain any sequence alterations. Therefore, this study suggests that PCR-sequence analysis works more efficiently and accurately than PCR-SSCP analysis for rapid screening of RMP-resistant M. tuberculosis clinical isolates.