TST, QuantiFERON-TB Gold test and T-SPOT.TB test for detecting latent tuberculosis infection in patients with rheumatic disease prior to anti-TNF therapy.

Abstract

Before starting tumour necrosis factor (TNF)-α blocking agents, standard tests should be used for the diagnosis of tuberculosis infection. The specificity of traditional tuberculin skin test (TST) is low in immunosuppressed patients due to prior Bacille Calmette Guérin (BCG) vaccination, non-tuberculous mycobacteria infections, false positive and negative results. In this study, we aimed to compare TST and Interferon-Gamma Release Assay (IGRA) tests for detecting latent tuberculosis infection in patients with rheumatic disease planned to receive TNF-α blocking agents. One hundred and nine patients (45 male, 64 female) with the diagnosis of rheumatoid arthritis (RA) (n= 70) and ankylosing spondylitis (AS) (n= 39) were included in the study. Age, sex, number of BCG scar, results of TST (using the Mantoux method), QuantiFERON-TB Gold test and T-SPOT.TB test were recorded for all patients. Correlation between the tests was assessed by Pearson correlation coefficient. The mean age of RA and AS patients were 50 ± 13 (19-78 years). The prevalence of latent tuberculosis was 43.1% for TST, 39.4% for QuantiFERON-TB Gold test and 13.8% for T-SPOT.TB test, compared with the evaluation using the composite criteria such as close contact with active tuberculosis infection and/or suspicious fibrotic/calcific lesions on chest X-Ray without active tuberculosis infection. There was a moderate correlation between BCG scar number and TST (p< 0.001, r= 0.495), T-SPOT.TB test and QuantiFERON-TB Gold test (p= 0.007, r= 0.406), T-SPOT.TB test and composite criteria (p= 0.024, r= 0.343). The specificity of QuantiFERON-TB Gold test was 85.7%, and sensitivity was 73.9% for all patients with rheumatic disease. It was 73.5% and 66.7% for T-SPOT.TB test, respectively. The specificity of TST was 60.3% and sensitivity was 47.8% for TST. IGRA tests are not affected prior vaccination and useful for detecting latent tuberculosis infection in patients treated with corticosteroid due to lack of correlation between test negativity and corticosteroid therapy. Also, they are useful tests for diagnosis of latent tuberculosis infection as an alternative to TST due to their specificity and sensitivity.