BACKGROUND: Interferon-γ release assays (IGRAs) are increasingly used for the diagnosis of latent tuberculosis infection (LTBI). Because of the lack of a gold standard for the diagnosis of LTBI, IGRAs are compared to the tuberculin skin test (TST) and yield conflicting results. We assessed the usefulness of an IGRA test, QuantiFERON(®)-TB Gold In-Tube (QFT-G-IT), for diagnosing LTBI compared with TST in the setting of a contact screening study. METHODS: A prospective comparison between the QFT-G-IT and the TST in TB contact subjects in a low TB burden area was conducted sequentially between January 2006 and December 2012. RESULTS: A moderate concordance between the two tests (κ = 0.44 for TST cut-off of 5 mm and κ = 0.56 for TST cut-off of 15 mm) was found. A better agreement was shown in younger contacts and in non-vaccinated contacts when using a TST of 15 mm. Independent risk factors for a TST(+)/QFT-G-IT(-) discordance were history of BCG vaccination and age between 31 and 59 years. Discordance was also more frequent using a TST cut-off value of 5 mm. QFT-G-IT(+)/TST(-) was infrequent and was found in older contacts. CONCLUSIONS: Based on our data, we cannot recommend the use of QFT-G-IT as the only test to rule out LTBI, especially in older patients.
BACKGROUND: Interferon-γ release assays (IGRAs) are increasingly used for the diagnosis of latent tuberculosis infection (LTBI). Because of the lack of a gold standard for the diagnosis of LTBI, IGRAs are compared to the tuberculin skin test (TST) and yield conflicting results. We assessed the usefulness of an IGRA test, QuantiFERON(®)-TB Gold In-Tube (QFT-G-IT), for diagnosing LTBI compared with TST in the setting of a contact screening study. METHODS: A prospective comparison between the QFT-G-IT and the TST in TB contact subjects in a low TB burden area was conducted sequentially between January 2006 and December 2012. RESULTS: A moderate concordance between the two tests (κ = 0.44 for TST cut-off of 5 mm and κ = 0.56 for TST cut-off of 15 mm) was found. A better agreement was shown in younger contacts and in non-vaccinated contacts when using a TST of 15 mm. Independent risk factors for a TST(+)/QFT-G-IT(-) discordance were history of BCG vaccination and age between 31 and 59 years. Discordance was also more frequent using a TST cut-off value of 5 mm. QFT-G-IT(+)/TST(-) was infrequent and was found in older contacts. CONCLUSIONS: Based on our data, we cannot recommend the use of QFT-G-IT as the only test to rule out LTBI, especially in older patients.