Evaluation of the accuracy of several immune experimental diagnostic tools to distinguish active tuberculosis from latent infection

Abstract

Background: There is still not reliable tests to distinguish active tuberculosis (TB) from latent TB infection (LTBI) by an immune diagnostic test. QuantiFERON TB Gold (QFT) is used for LTBI diagnosis; however this test does not distinguish between LTBI and active TB. Recently, assays based on the response to antigens of latency as HBHA or Rv2628, or evaluation on RD1-specific T cells by cytometry of the modulation of the effector memory/central memory status has been evaluated.

Aim: To evaluate in a low TB endemic country as Italy, the accuracy to discriminate active TB from LTBI using several immune diagnostic tests based on different TB antigens and on the modulation of the T cell effector memory status.

Methods: Evaluation of the accuracy to distinguish active TB from LTBI in several cohorts of QFT-IT-positive subjects either HIV-uninfected (107) or HIV-infected subjects (27) with active TB, LTBI or enrolled as cured TB. Whole blood or peripheral blood cells were stimulated with RD1 and latency antigens. Interferon (IFN)-γ response was evaluated by ELISA and mono-polyfunctional cytokine expression (IFN-γ, IL-2, TNF-α) and differentiation markers by cytometry.

Results: To predict active TB we used the combination of the different experimental-immune-based approaches by logistic regression followed by ROC analysis (AUC 0.96; 95% CI, 0.87-0.99, p<0.0001). We then identified a cut-off which led to an accuracy of 92.3% sensitivity and 95.2% specificity.

Conclusions: This study proposes that a combination of several experimental approaches may help to distinguish TB disease from LTBI in a low TB endemic country.