BackgroundIdentifying latent tuberculosis infection (LTBI) in people migrating from TB endemic regions to low incidence countries is an important control measure. However, no prospective longitudinal comparisons between diagnostic tests used in such migrant populations are available.ObjectivesTo compare commercial interferon (IFN)-gamma release assays (IGRAs) and the tuberculin skin test (TST) for diagnosing LTBI in a migrant population, and the influence of antecedent TST and LTBI treatment on IGRA performance.Materials and MethodsThis cohort study, performed from February to September 2012, assessed longitudinal IGRA and TST responses in Nepalese military recruits recently arrived in the UK. Concomitant T-SPOT.TB, QFT-GIT and TST were performed on day 0, with IGRAs repeated 7 and 200 days later, following treatment for LTBI if necessary.Results166 Nepalese recruits were prospectively assessed. At entry, 21 individuals were positive by T-SPOT.TB and 8 individuals by QFT-GIT. There was substantial agreement between TST and T-SPOT.TB positives at baseline (71.4% agreement; κ = 0.62; 95% CI:0.44–0.79), but only moderate concordance between positive IGRAs (38.1% agreement; κ = 0.46; 95% CI:0.25–0.67). When reassessed 7 days following TST, numbers of IGRA-positive individuals changed from 8 to 23 for QFT-GIT (p = 0.0074) and from 21 to 23 for T-SPOT.TB (p = 0.87). This resulted in an increase in IGRA concordance to substantial (64.3% agreement; κ = 0.73; 95% CI:0.58-0.88). Thus, in total on day 0 and day 7 after testing, 29 out of 166 participants (17.5%) provided a positive IGRA and of these 13 were TST negative. Two hundred days after the study commenced and three months after treatment for LTBI was completed by those who were given chemoprophylaxis, 23 and 21 participants were positive by T-SPOT.TB or QFT-GIT respectively. When individual responses were examined longitudinally within this population 35% of the day 7 QFT-GIT-positive, and 19% T-SPOT.TB-positive individuals, were negative by IGRA. When the change in the levels of secreted IFN-γ was examined after chemoprophylaxis the median levels were found to have fallen dramatically by 77.3% from a pre-treatment median concentration of IFN-γ 2.73 IU/ml to a post-treatment median concentration IFN-γ 0.62 (p = 0.0002).ConclusionsThis study suggests differences in the capacity of commercially available IGRAs to identify LTBI in the absence of antecedent TST and that IGRAs, in the time periods examined, may not be the optimal tests to determine the success of chemoprophylaxis for LTBI.