Higher T-SPOT.TB threshold may aid in diagnosing active tuberculosis?: A real-world clinical practice in a general hospital

Background: This study aimed to determine whether increased cut-off of the T-SPOT.TB could aid in diagnosing active tuberculosis (ATB).

Methods: Patients suspected of having TB were enrolled to derive a T-SPOT.TB threshold value to help diagnose ATB, which was subsequently validated in real-world clinical practice.

Results: In total, 701 adult patients suspected of having tuberculosis who had undergone the T-SPOT.TB assay were included in the derivation cohort. The numbers of ESAT-6 (U = 43583, P = 0.0002) and CFP-10 (U = 41753, P < 0.0001) spot-forming cells (SFCs) significantly increased in the ATB group compared with the Latent tuberculosis infection (LTBI) group. According to receiver operating characteristic analysis, when a cut-off of 37.5 SFCs/2.5 × 10⁵ cells was used to discriminate between ATB and LTBI, the sensitivity was 57.5% (95% confidence interval [CI] 50.7%-64.2%) and the specificity was 59.8% (95% CI 55.2%-64.2%). A threshold value of 173.5 SFCs/2.5 × 10⁵ could be used to obtain a specificity of <90% to discriminate between ATB and LTBI. The diagnostic accuracy of higher T-SPOT.TB threshold values in the validation cohort was similar to that in the derivation cohort.

Conclusions: In high-burden countries, a higher threshold value of 173.5 SFCs/2.5 × 10⁵ may aid in ATB diagnosis in suspected tuberculosis patients.