Factors associated with indeterminable results of interferon gamma release assay (IGRA) in routine screening for latent tuberculosis (LTB) in patients with inflammatory bowel diseases (IBD)

Introduction: Screening for LTB including tuberculin skin test (TST), chest X-ray and facultatively, a whole blood interferon gamma release assay (IGRA) is mandatory in patients with IBD before starting therapy with Tumor Necrosis Factor (TNF) alpha inhibitors. However, being the more specific test method, IGRA might be limited in subjects under IM therapy due to a substantial number of indeterminable results. Aims and Methods: We aimed to evaluate factors such as laboratory findings and disease specific data which are associated with indeterminable IGRA (QuantiFERON®) results from logistic regressions on a cohort of 190 consecutive IBD patients. QuantiFERON® was indeterminable if the result of IFN γ concentration was <0.35 IU/ml for tuberculosis-specific antigens and <0.5 IU/ml for the positive control according to manufacturer's instructions. Results: IGRA failed on samples from 26/190 (13.7%) patients, resulting in 164/190 (86.3%) patients with determinable results, 14 (7.4%) of them positive, 150 (78.9%) negative. In 15/26 (58%) subjects with indeterminable results IGRA and laboratory findings were repeated. In 6 cases, 4 under double (3 with combination of AZA and steroids) and 2 under monotherapy with IM (steroids n=1, AZA n=1), IGRA remained indeterminable after a mean of 17 days and absolute lymphocyte count remained low and unchanged 0.65 0.45 vs. 0.71 0.37 G/l (p=0.6). In contrast, in 9 cases in whom after a mean of 28 days IGRA became negative, there were only 2 subjects under double IM (both AZA and steroids), 6 under IM monotherapy (steroids in tapering n=3, AZA n=2, anti-TNF n=1) and 1 subject without IM therapy. Lymphocyte count increased in those patients from 1.1 0.8 to 1.5 1.1, (p=0.08). Multivariate analysis revealed clear association of indeterminable results of IGRA with lower albumin odds ratio (OR) of [0.87 (0.79–0.96), p=0.008], lower lymphocyte count [OR 0.38 (0.18–0.79) p=0.009] and a trend for double IM therapy [2.88 (0.95–8.8), p=0.063]. Conclusion: Our results reveal clear associations of IGRA with lower serum albumin levels and lower absolute peripheral blood lymphocyte counts. We speculate that repeated IGRA testing in case of indeterminable results might be best performed after recovery of low lymphocyte counts, e.g. induced by a forced steroid reduction.