Background: Interferon-gamma release assays (IGRAs) have provided a new method for the diagnosis of Mycobacterium tuberculosis infection. However, the role of IGRAs for the diagnosis of active tuberculosis (TB), especially in HIV-infected patients remains unclear.
Methods: We searched PubMed, EMBASE and Cochrane databases to identify studies published in January 2001-July 2011 that evaluated the evidence of using QuantiFERON-TB Gold in-tube (QFT-GIT) and T-SPOT.TB (T-SPOT) on blood for the diagnosis of active TB in HIV-infected patients.
Results: The search identified 16 eligible studies that included 2801 HIV-infected individuals (637 culture confirmed TB cases). The pooled sensitivity for the diagnosis of active TB was 76.7% (95%CI, 71.6-80.5%) and 77.4% (95%CI, 71.4-82.6%) for QFT-GIT and T-SPOT, respectively, while the specificity was 76.1% (95%CI, 74.0-78.0%) and 63.1% (95%CI, 57.6-68.3%) after excluding the indeterminate results. Studies conducted in low/middle income countries showed slightly lower sensitivity and specificity when compared to that in high-income countries. The proportion of indeterminate results was as high as 10% (95%CI, 8.8-11.3%) and 13.2% (95%CI, 10.6-16.0%) for QFT-GIT and T-SPOT, respectively.
Conclusion: IGRAs in their current formulations have limited accuracy in diagnosing active TB in HIV-infected patients, and should not be used alone to rule out or rule in active TB cases in HIV-infected patients. Further modification is needed to improve their accuracy.