Evaluation of a Whole‐Blood Interferon‐γ Release Assay for the Detection ofMycobacterium tuberculosisInfection in 2 Study Populations

Abstract

A whole-blood interferon-γ release assay (IGRA) is being evaluated for its potential to replace the tuberculin skin test (TST) for detecting Mycobacterium tuberculosis infection. To test the assay in a population in which tuberculosis is highly endemic and in another population that is representative of an urban United States population, 253 volunteers from Ethiopia and 175 volunteers from Baltimore were studied for responsiveness on IGRA compared with a simultaneously performed TST. The agreement between the 2 tests, beyond that due to chance, was 68% among subjects from Baltimore and only 35% among those from Ethiopia. IGRA had a sensitivity of 71%, compared with 95% sensitivity for the TST, among 21 subjects who had undergone treatment for culture-confirmed tuberculosis. The specificity was 85% for IGRA and 96% for TST among 52 subjects with no known history of exposure to tuberculosis. In its current form, with purified protein derivative used as the stimulation antigen, the IGRA was found to perform poorly in comparison to the TST in diagnosing M. tuberculosis infection. (See the editorial commentary by Nadal on pages 1457–9.)