RESUMO
OBJECTIVES: Currently available interferon (IFN)-γ-release assays (IGRA) cannot discriminate active tuberculosis (TB) from latent TB infection (LTBI), and so have limited clinical utility for diagnosing active TB. Since numbers of tumour necrosis factor (TNF)-α-producing T cells are highly correlated with active TB, we hypothesized that detecting IFN-γ- and/or TNF-α-producing T cells would overcome this limitation of IGRA. This study evaluated the diagnostic performances of the IFN-γ and TNF-α dual release fluorospot assay for active TB. METHODS: Adult patients with suspected TB including recent TB exposers were prospectively enrolled over a 28-month period. In addition to the conventional IGRA test (i.e. QuantiFERON-In-Tube), a fluorospot assay for detecting IFN-γ- and TNF-α-producing T cells was performed. The final diagnoses were classified by clinical category. Patients with confirmed or probable TB were regarded as active TB, and patients with not active TB were further classified as having not active TB with and without LTBI, based on the QuantiFERON-In-Tube results. RESULTS: A total of 153 patients including 45 with active TB and 108 with not active TB (38 LTBI vs. 70 not LTBI) were finally analysed. The sensitivity and specificity of the QuantiFERON-In-Tube assay for active TB were 84% (95% confidence interval (CI), 70-93) and 70% (95% CI 61-79), respectively. The IFN-γ/TNF-α dual release assay by fluorospot had substantially higher diagnostic specificity (94%) for diagnosing active TB than the IFN-γ single release assay (72%, p < 0.001), without compromising sensitivity (84% vs. 89%, p 0.79). CONCLUSIONS: The fluorospot-based IFN-γ/TNF-α dual release assay appears to be a simple and useful test for diagnosing active TB.