Abstract

The number of patients suffering from tuberculosis (TB) globally is increasing. Due to the HIV epidemic, most patients suffering from TB reside in sub-Saharan Africa. In order to improve TB diagnostics, new tests - interferon-gamma release assays (IGRAs) - have been developed over the last decade. In this paper we evaluate the possible use of these tests in diagnosing or excluding active TB in high HIV-burden, resource-limited settings. The inability to differentiate between active and latent TB, limited data on IGRA performance in HIV-infected patients, observed false-negative results, high costs, and logistic problems limit the potential benefit of IGRAs. We also present two theoretical study designs in order to further assess IGRAs. Setting up a study on this subject is complicated by the frequent unavailability of mycobacterial cultures, the difficulty in acquiring prospective data, and the impossibility of denying treatment to a patient suspected of having active TB. We feel that current evidence does not support the implementing of IGRAs in clinical practice in settings with high endemic latent TB infection (LTBI) and high HIV prevalence. As these settings are the ones that suffer the most from the TB epidemic, we believe that the role of IGRAs in global TB control is questionable.