The usefulness of routine chest radiograph examinations in patients treated with TNF inhibitors for inflammatory arthritis in South Korea.

Abstract

We aimed to investigate the usefulness of routine chest radiograph (CXR) examinations for patients with inflammatory arthritis treated with a tumor necrosis factor (TNF) inhibitor in terms of (i) the role of CXR in baseline latent tuberculosis infection (LTBI) screening and (ii) detecting asymptomatic active tuberculosis after TNF inhibitor initiation. From January 2011 to June 2017, 469 patients with inflammatory arthritis were enrolled in the study at a tertiary referral center in South Korea. At our institution, CXR was performed for all patients undergoing a tuberculin skin test (TST) and/or an interferon-gamma release assay (IGRA) at the LTBI screening visit. LTBI treatment was determined by (i) positive TST or IGRA or (ii) CXR findings suggestive of spontaneously healed tuberculosis. After TNF inhibitor initiation, patients were recommended to undergo CXR at a specified interval. Of 469 patients, 187 were treated for LTBI. Among them, 181 patients were treated for LTBI because of a positive TST or IGRA result. TST was considered positive if induration size was ≥10 mm. The remaining six patients were considered positive on the basis of CXR findings compatible with spontaneously healed tuberculosis, such as noncalcified nodules with distinct margins and fibrotic linear opacity, despite demonstrating negative results for TST and IGRA. Thus, CXR had a diagnostic value as a baseline LTBI test in 6 (1.3%) patients. After TNF inhibitor initiation, 2 patients who had respiratory symptoms were diagnosed with active tuberculosis. For asymptomatic patients, routine CXR follow-up could not detect any case of active pulmonary tuberculosis within 1 year (n = 219) or after 1 year (n = 217). CXR should be performed as one of the LTBI screening tests for patients with inflammatory arthritis in a tuberculosis-prevalent country. However, after TNF inhibitor treatment, routine CXR follow-up was not advantageous.