A simple blood test for detecting active tuberculosis (TB) could be key to this epidemic containment, given that a large proportion of patients are unable to produce sputum for testing. Currently available interferon-γ release assays (IGRAs) are inadequate to diagnose active TB, with reported pooled sensitivity and specificity both under 81%.

To explore whether cytokines/chemokines other than interferon-γ in response to long-term cell stimulation could improve the ability to distinguish between different TB infection status.

We prospectively enrolled subjects with newly diagnosed pulmonary TB and their household contacts in Santiago. All contacts were tested with IGRA. Peripheral blood mononuclear cells were obtained and antigen-specific stimulated for 72 h before collecting their culture supernatants.

Subjects with active TB displayed markedly low cytokines/chemokines secretion upon PBMC stimulation, with lower GM-CSF being the best differentiator from IGRA(+) contacts, with 71% (95% CI 53–85) sensitivity, 86% (95% CI 65–97) specificity and AUC = 0.79 (p = 0.0003). On the other hand, when compared to the uninfected IGRA(−) contacts, higher level of IL-2 secretion was the best indicator of active TB, with 73.5% (95% CI 56–87) sensitivity, 85% (95% CI 66–96) specificity and AUC = 0.79 (p = 0.0001). No single cytokine/chemokine released upon stimulation could accurately differentiate between active TB and all TB contacts grouped together.

GM-CSF and IL-2 provided the best yield to differentiate active TB from latent TB and from TB uninfected, respectively, with higher specificities than that reported for IGRAs. However, none of both resulted sensitive enough to be used as a stand-alone biomarker for active TB.