Effect of adjusted cut-offs of interferon-γ release assays on diagnosis of tuberculosis in patients with fever of unknown origin

Abstract

BackgroundTuberculosis (TB) is a leading cause of fever of unknown origin (FUO). In recent years, interferon-γ release assays (IGRAs) have been widely utilized and the cut-off values given by the manufacturers are set in countries where rates of TB are not as high. MethodsA prospective cohort study was conducted in a Chinese general hospital to evaluate the diagnostic performance of T-SPOT.TB (T-SPOT) and QuantiFERON-TB Gold (QFT) in detecting active TB (ATB) in a high TB endemic area. Test results were compared with the culture and clinically confirmed diagnosis. Further, we explored an alternative method of interpreting IGRAs by increasing the cut-off values. ResultsThe sensitivity and specificity of T-SPOT in detecting ATB were 85.3% (95% CI 81.6–94.0%) and 71.8% (95% CI 67.3–76.0%), respectively. The sensitivity and specificity of QFT were 72.3% (95% CI 62.8–80.1%) and 77.0% (95% CI 72.7–80.8%), respectively. Receiver operating characteristic analysis was used for evaluation of different cut-off values. When the cut-off values were adjusted as 125 spot-forming cells (SFCs)/ 2.5*105 cells for T-SPOT and 4.0 IU/ml for QFT, the specificity could be improved to > 90.0% (90.3% and 94.1%, respectively), and the sensitivity were 43.1% and 41.6%, respectively. The new adjusted cut-off values were validated in another independent validation cohort. ConclusionThe adjusted cut-off values of the two assays considerably improved the diagnostic value when applied to FUO patients in clinical settings.