Long-term performance of the IGRA to predict and prevent active tuberculosis development in HIV-infected patients

To estimate the long-term performance of the interferon-gamma release assay (IGRA) in predicting active tuberculosis (ATB) development among human immunodeficiency virus (HIV) infected patients in an intermediate TB burden country. A retrospective cohort study was conducted. HIV-infected patients with 1 IGRA result but no current or previous ATB who had been retained in care for 1 year were enrolled and observed for ATB development from 2006 to 2016. Sixty-two IGRA-positive and 354 IGRA-negative patients were observed for a median period of 4.03 years. ATB incidence in IGRA-positive vs. negative patients was respectively 15.24/1000 person-years (py) (95%CI 4.15-39.02) vs. 0.67/1000 py (95%CI 0.01-3.68) (P = 0.001). The sensitivity, specificity and positive and negative predictive values of the IGRA were respectively 80.0% (28.4-99.5%), 85.9% (82.1-89.1%), 6.5% (4.0-10.2%) and 99.7% (98.4-99.9%). Among the 62 IGRA-positive patients, 25 were treated for latent tuberculous infection (LTBI) and 37 were not. Patients with LTBI treatment did not develop ATB, while four cases of ATB occurred among patients without LTBI treatment. Among the 257 IGRA-negative patients who received a follow-up IGRA, 34 (12.9%) patients converted to positive. No ATB developed in IGRA converters. Screening for LTBI using IGRAs and LTBI treatment can be useful to prevent ATB in an intermediate TB burden country. .