Predictors of indeterminate IFN-γ release assay in screening for latent TB in inflammatory bowel diseases

Abstract

IFN-γ release assays (IGRA), widely used for latent tuberculosis screening prior to anti-TNF-α treatment, are limited by indeterminate results in patients under immunomodulatory (IM) therapy. The aim of our observational study was to delineate factors associated with indeterminate IGRA results. A total of 190 patients with inflammatory bowel disease were included. IGRA was indeterminate if the result of IFN-γ concentration was < 0·35 IU mL(-1) for tuberculosis-specific antigens and < 0·5 IU mL(-1) for the positive control. Predictors for indeterminate results were delineated from multivariate logistic regression. IFN-γ release assays was indeterminate in 26/190 (13·7%) patients. Indeterminate IGRA were associated with lower serum albumin levels (odds ratio [OR] 0·88, 95% confidence interval [CI] 0·79-0·96), lower absolute lymphocyte count (OR 0·39, 95% CI 0·18-0·75) and double IM therapy (OR 2·98, 95% CI 0·95-8·90). Sub-analysis of IM therapy revealed an association of steroid therapy with indeterminate IGRA (OR 3·19, 95% CI 1·35-7·70). Hypoalbuminaemia increased the risk of indeterminate IGRA by (OR 2·97, 95% CI 1·03-8·61) and lymphopaenia by (OR 3·28, 95% CI 1·41-7·65). After a mean of 18·5 ± 14·4 days, retesting of IGRA in 18 patients with indeterminate results yielded 9 negative vs. 9 indeterminate results. Our results reveal associations of indeterminate IGRA with low serum albumin levels and absolute lymphocyte count and double IM therapy. IGRA testing appears best to be performed prior to initiation of IM therapy in patients with inflammatory bowel disease.