Tuberculosis Disease in Children and Adolescents on Therapy With Antitumor Necrosis Factor-ɑ Agents: A Collaborative, Multicenter Paediatric Tuberculosis Network European Trials Group (ptbnet) Study.

ABSTRACT

BACKGROUND: In adults, anti-tumor necrosis factor-α (TNF-α) therapy is associated with progression of latent tuberculosis (TB) infection (LTBI) to TB disease, but pediatric data are limited. METHODS: Retrospective multicenter study within the Paediatric Tuberculosis Network European Trials Group, capturing patients <18 years who developed TB disease during anti-TNF-α therapy. RESULTS: Sixty-six tertiary healthcare institutions providing care for children with TB participated. Nineteen cases were identified Crohn's disease (n = 8; 42%) and juvenile idiopathic arthritis (n = 6; 32%) were the commonest underlying conditions. Immune-based TB screening (tuberculin skin test and/or interferon-γ release assay) was performed in 15 patients before commencing anti-TNF-α therapy but only identified 1 LTBI case; 13 patients were already receiving immunosuppressants at the time of screening. The median interval between starting anti-TNF-α therapy and TB diagnosis was 13.1 (IQR, 7.1-20.3) months. All cases presented with severe disease, predominantly miliary TB (n = 14; 78%). One case was diagnosed postmortem. TB was microbiologically confirmed in 15 cases (79%). The median duration of anti-TB treatment was 50 (IQR, 46-66) weeks. Five of 15 (33%) cases who had completed TB treatment had long-term sequelae. CONCLUSIONS: LTBI screening is frequently false-negative in this patient population, likely due to immunosuppressants impairing test performance. Therefore, patients with immune-mediated diseases should be screened for LTBI at the point of diagnosis, before commencing immunosuppressive medication. Children on anti-TNF-α therapy are prone to severe TB disease and significant long-term morbidity. Those observations underscore the need for robust LTBI screening programs in this high-risk patient population, even in low-TB-prevalence settings.