The significance of isolating low numbers of mycobacterium tuberculosis in culture of sputum specimens.

Because fewer than five Mycobacterium tuberculosis colonies were isolated from single sputum specimens from several patients with little clinical evidence of pulmonary tuberculosis, we reviewed all of our low colony isolates (LCIs) from an 18-month period. Thirty-one patients had 35 LCIs and were separated easily into groups that either appeared to have tuberculosis by clinical criteria or appeared to be uninfected. Patients in the "tuberculosis likely" group reasons such as concurrent antituberculosis therapy or an interstitial location of their pulmonary lesion to account for their LCIs. In contrast, patients in the "tuberculosis unlikely" group were all culture-negative when reevaluated two to four months after their first culture, and they had no evidence of tuberculosis on chest x-ray film despite having received no antituberculosis therapy. Half had negative tuberculin reactions, without generalized anergy. Epidemiologic study showed that most had been cultured the same day as patients with smear-positive tuberculosis, and the possibility of cross-contamination was investigated.