Use of PCR and culture for detection of Mycobacterium tuberculosis in specimens from patients with normal and slow responses to chemotherapy.

In order to examine the utility of PCR as a tool for monitoring the response to short-course chemotherapy, clinical data from 162 tuberculosis patients with drug-susceptible cultures and variables associated with a slow response to antibiotic treatment were retrospectively evaluated. Initial samples were positive by smear in 106 patients and by PCR in 156 patients. A rapid response to therapy was found in 132 patients who had both smear and culture-negative results after 2 months of treatment. After 6 months of therapy, 5 other patients in this group were considered clinically cured but 20 patients had positive cultures. For these 20 patients with a slow response to therapy, treatment was continued for an additional 2-4 months and 15 patients were subsequently cured. However, of all the cured patients 7 had PCR-positive samples at the end of treatment. When these cases were correlated with therapeutic outcome, it was found that 3 of the PCR-positive but none of the PCR-negative patients had a clinical relapse during subsequent follow-up. We conclude that culture and PCR tests are highly informative when used to control the rate of response to therapy. A post-treatment positive result on PCR may be associated with a poor clinical outcome or may predict the likelihood of clinical relapse. Culture and PCR tests, as opposed to smear results, may be the key parameters for individually guiding the duration of treatment in TB patients with a poor response to standard therapy.