Background: Clinical data of serial interferon-gamma release assay (IGRA) testing in routine laboratory practice are limited. IFN-gamma response to mitogen is used as a positive control in IGRA. We assessed the association between IFN-gamma response to nil, mitogen, tuberculosis (TB) mycobacterial antigens, and the variations from the results of the serial testing.Method: A total of 299 patients with serial QuantiFERON-TB Gold In-Tube (QFT-GIT) were enrolled. The medical records of patients were reviewed for demographic information, status of Mycobacterium tuberculosis infection, treatment of tuberculosis, and the quantitative response to nil, mitogen, and TB antigen.Results: The initial QFT-GIT result was positive in 142 patients (47.5%), negative in 139 (46.5%), and indeterminate in 18 (6.0%). Of total, 79.6% showed concordant results in serial testing. The discordance in serial tests was significantly high in patients with a low mitogen response (<= 3.93 IU/ml) (p < 0.0001). Quantitative TB responses around the cut-off point in serial QFT-GIT were associated with an increased conversion and reversion rates (p = 0.01, p = 0.0005), respectively.Conclusion: Because IGRAs are dynamic assays, integrated interpretation of quantitative TB response with mitogen and nil response would be helpful in serial QFT-GIT. Recommendations for the interpretation of results of serial testing for active TB will be required. (C) 2013 Published by Elsevier B.V.