Targeted versus universal tuberculosis chemoprophylaxis in 1968 patients with inflammatory bowel disease receiving anti‐TNF therapy in a tuberculosis endemic region

Summary Background Anti‐tumour necrosis factor (anti‐TNF) therapy increases the risk of tuberculosis (TB). Given limitations of screening techniques, it remains uncertain if patients receiving anti‐TNF in TB endemic regions should be screened for latent infection with chemoprophylaxis restricted to those with proven infection, or if all patients should receive chemoprophylaxis. Aims To compare the incidence of active TB with infliximab (IFX) following targeted and universal TB chemoprophylaxis, and to determine the rates of adverse events (AE) related to TB chemoprophylaxis Methods A multi‐centre retrospective cohort study was performed at 18 hospitals in China of 1968 adult patients with IBD receiving IFX from 2009 to 2017. TB screening prior to IFX was performed with chest X‐ray and/or computed tomography [CT] and immune reactivity testing (interferon‐γ release assay and/or tuberculin skin test). Patients were followed‐up for a minimum of 3 months after IFX discontinuation, or until last hospital visit if IFX therapy was ongoing. Targeted strategy was defined as TB chemoprophylaxis only for patients with a positive latent TB screen, with universal strategy defined as TB chemoprophylaxis for all patients. Results Mean follow‐up was 1.07 ± 0.87 years with a total follow‐up of 2102 patient‐years. There were 1433 patients in the targeted and 483 patients in the universal TB chemoprophylaxis groups, with no significant difference in the incidence rates of active TB between groups (673.3 per 100 000 population per year vs 891.5 per 100 000 population per year, P = 0.60). In the targeted group, 55/1433 patients received TB chemoprophylaxis compared with 483/483 in the universal group, with significantly fewer AEs related to TB chemoprophylaxis in the targeted compared to the universal group (0.35% (5/1433) vs 6.8% (33/483), P