IS6110 restriction fragment length polymorphism typing of clinical isolates of Mycobacterium tuberculosis from patients with pulmonary tuberculosis in Madras, south India.

Madras, India. To explore the utility of a standardized IS6110/PvuII deoxyribonucleic acid (DNA) fingerprinting restriction fragment length polymorphism (RFLP) typing method for distinguishing between isolates of Mycobacterium tuberculosis, and to assess the potential for distinguishing between relapse versus reinfection rates. To assess RFLP heterogeneity in the population, initial isolates, obtained from the sputum of tuberculous 98 patients in diagnosis and follow-up during short-course chemotherapy, were stored and compared. To assess the frequency of disparity between the RFLP type of the initial isolate and one obtained after successful completion of chemotherapy, either during relapse or as an isolated positive culture, 124 isolates comprising 62 such pairs were coded and compared both blind and after decoding. Although a wide variety of DNA band patterns (fingerprints) was present, the isolates from 39 (40%) of the patients showed a single copy of IS6110. Only 15 pairs of coded initial and follow-up isolates could be identified as having the same band pattern when isolates with zero or single bands were excluded. Nevertheless, after decoding, in a retrospective analysis that included all isolates, those isolates that bacteriologically defined a patient's relapse more often showed RFLP type identity with the initial isolate (19 of 30 comparisons) than did isolates that were obtained as isolated positive cultures (3 of 32 comparisons) (X2 P < 0.001). Tests of sensitivity to chemotherapeutic drugs, catalase activity and resistance to thiophene-2-carboxylic acid hydrazide were of minimal value in discriminating between isolates. Despite the high frequency of single- and zero-band isolates in this population, the discriminatory power of RFLP typing with IS6110 is sufficiently high to be useful for clinical and epidemiological studies.