Direct detection of Mycobacterium tuberculosis in respiratory specimens in a clinical laboratory by polymerase chain reaction

Abstract

The emergence of epidemic multiple-drug-resistant (MDR) strains of Mycobacterium tuberculosis in conjunction with an increase in the number of reported cases of tuberculosis (TB) represents a major public health problem. In light of a recent outbreak of MDR M. tuberculosis at our center, we began the development of a polymerase chain reaction (PCR) assay for the rapid diagnosis of pulmonary TB using two sets of primers, one based on the IS6110 repeated sequence of M. tuberculosis and the other based on the protein antigen b (PAB). Reaction conditions were first optimized as to the appropriate extraction protocol and the concentrations of primer pairs, nucleotides, and MgCl2. Following a preliminary evaluation of the assay with clinical specimens, extraction and amplification procedures were further modified. PAB and IS6110 primers detected between 2 and 23 and 0.023 and 0.23 CFU of M. tuberculosis, respectively, in pooled, M. tuberculosis-negative sputa by our optimized PCR assay. After routine processing for mycobacteria, 734 specimens were subsequently amplified. DNA for amplification was obtained by boiling and beating the sediments with Tween 20. For each reaction, DNA (10 microliters) was added to an amplification mixture containing 12 pmol of IS6110 primers, 20 pmol of PAB primers, 2 mM MgCl2, 200 microM nucleotides, and 2.5 U of Taq polymerase and the mixture was then amplified for 40 cycles. The sensitivity and specificity of our PCR assay were 87.2 and 97.7%, respectively. We were unable to interpret the results for seven specimens (1%). In our experience, PCR proved to be a useful rapid diagnostic test for TB in a clinical setting and a valuable epidemiological tool for determining exposure groups in the hospital setting. Our findings also underscore the need for the systematic optimization of PCR assay conditions.