Accuracy of the interferon-γ release assay for the diagnosis of active tuberculosis among HIV-seropositive individuals: a systematic review and meta-analysis

BACKGROUND: Although the interferon-γ release assay (IGRA) has become a widely accepted means for the diagnosis of latent tuberculosis infection (LTBI), the role of the IGRA in diagnosing active tuberculosis (ATB) among human immunodeficiency virus (HIV)-seropositive individuals remains controversial. Previous analyses did not set up rational inclusive criteria for screening articles with strict control groups and a gold standard for ATB diagnosis. Therefore, we conducted a systematic review of the latest evidence to evaluate the accuracy of IGRA for HIV-seropositive patients. METHODS: Initially, we searched the EMBASE, Cochrane and MEDLINE databases to find research articles published from January 2000 to October 2015 that used the QuantiFERON-TB Gold In-Tube assay (QFT-IT) or the T-SPOT.TB assay (T-SPOT) to diagnose ATB among HIV-seropositive individuals. We separately calculated the pooled sensitivity, specificity, and proportion of indeterminate events and then summarized the results using forest plots to estimate the accuracy of the QFT-IT and T-SPOT assays. RESULTS: A total of 1,743 studies were discovered after searching; 11 studies met our selection standards and were included for meta-analysis. The pooled sensitivity and specificity of the QFT-IT assay were 69 % (95 % CI, 50-84 %, I(2) = 85.22 %) and 76 % (95 % CI, 53-90 %, I(2) = 98.16 %), respectively, and the optimum area under the curve (AUC) was 0.78 (95 % CI, 0.74-0.82). The pooled sensitivity and specificity of the T-SPOT assay were 89 % (95 % CI, 66-97 %, I(2) = 94.48 %) and 87 % (95 % CI, 38-99 %, I(2) = 97.92 %), respectively, and the AUC was 0.93 (95 % CI, 0.90-0.95). The pooled ratios of the indeterminate results of the QFT-IT and T-SPOT assays were 0.07 (95 % CI, 0.06-0.09, I(2) = 74.8 %) and 0.19 (95 % CI, 0.15-0.24, I(2) = 88.3 %), respectively, calculated using the fixed effect model, and 0.08 (95 % CI, 0.06-0.12, I(2) = 74.8 %) and 0.10 (95 % CI, 0.03-0.25, I(2) = 88.3 %), respectively, calculated using the random effects model. CONCLUSIONS: The IGRA does not appear to be optimal for the clinical confirmation of ATB cases in HIV-seropositive patients; however, the T-SPOT assay may have greater accuracy in distinguishing ATB cases among HIV-infected individuals than the QFT-IT assay, while the QFT-IT assay appears to reduce the occurrence of indeterminate results. Furthermore, modification and additional trial designs are required to improve diagnostic effectiveness.