Transmission of Mycobacterium tuberculosis undetected by tuberculin skin testing.

RATIONALE The development of tuberculin skin test (TST) positivity following infection by Mycobacterium tuberculosis is not invariable and may depend on bacillary as well as host factors. OBJECTIVES First, to compare the diagnostic performance of the TST and a form of in vitro IFN-gamma release assay (IFNGRA) in the circumstances of a contact investigation prompted by an unusually severe index case of infectious pulmonary tuberculosis. Second, to investigate the ability of the strain of M. tuberculosis responsible to induce cytokine secretion from monocytes in vitro. METHODS A routine TST-based tuberculosis-contact screening procedure supplemented by the use of an "in house" IFNGRA that assays the T-cell response to the M. tuberculosis-specific antigens ESAT-6, CFP-10 (presented as a fusion protein within the inactivated adenylate cyclase of Bordetella pertussis), and purified protein derivative of M. tuberculosis. Isolation and genetic typing of the strain of M. tuberculosis responsible, and investigation of its ability to induce cytokine secretion from monocytes in vitro. MEASUREMENTS AND MAIN RESULTS TST screening suggested a low rate of transmission with just 2/75 unequivocally positive responses. By contrast, the IFNGRA suggested an infection rate of 16/75 (22%). When compared with two reference strains of M. tuberculosis (H37Rv and CDC1551), the outbreak strain induced lower levels of tumor necrosis factor-alpha and interleukin-12p40 (p < 0.04), cytokines associated with the development of delayed-type hypersensitivity. CONCLUSIONS These data suggest that infection by M. tuberculosis can be undetected by TST, and that this may partially relate to strain differences in immunogenicity.