ABSTRACT
BACKGROUND: In developing countries, tuberculosis (TB) is a major public health problem and the leading cause of death among patients with HIV (Human Immunodeficiency Virus). Until 2001, the tuberculin skin test (TST) was the only available tool for the diagnosis of latent tuberculosis infection (LTBI), but false-negative TST results are frequently reported. Recently, the interferon-γ (IFN-γ) release assay (IGRA) has gained ground because it can detect the IFN-γ secreted by circulating lymphocytes T cells when stimulated by specific TB antigens. However, the role of IGRA in the diagnosis of LTBI in HIV-infected patients has not been well established. METHODS: This cross-sectional study compared the accuracy of TST (performed by the Mantoux method) and IGRA (QuantiFERON-TB Gold In-Tube, Cellestis, Carnegie, Australia) on the diagnosis of LTBI among patients with HIV. LTBI is defined by LTBI risk and at least one positive test (TST or IGRA), without clinical evidence of active TB. We also assessed the accuracy of TST and IGRA among HIV patients with high and low risk for LTBI. RESULTS: Among 90 HIV patients, 80 met the study criteria for LTBI, fifty-nine (73.7%) patients were TST positive, 21 (26.2%) were negative, whereas 75 patients (93.7%) were IGRA positive, and five (6.2%) were negative. TST showed poor agreement with the diagnosis of LTBI (Kappa: 0.384), while IGRA demonstrated good agreement (Kappa: 0.769). Among 69 patients with high risk and 21 with low risk for LTBI, TST was positive in 48 (69.5%) and 11 (52.4%), while IGRA was positive in 68 (98.5%) and 7 (33.3%) patients, respectively. There were no association between TST and the level of risk (P = 0,191). Conversely, we observed a strong association between the IGRA and risk for LTBI (p < 0.001). CONCLUSIONS: Compared to TST, IGRA positivity is consistent with the risk of TB infection and seems to be a better diagnostic tool for LTBI in HIV-infected patients.
