Diagnostic accuracy of non-contrast magnetic resonance enterography in detecting active bowel inflammation in pediatric patients with diagnosed or suspected inflammatory bowel disease to determine necessity of gadolinium-based contrast agents

Background Pediatric patients with inflammatory bowel disease (IBD) are at increased risk of gadolinium deposition given the potential need for multiple contrast-enhanced magnetic resonance enterography (MRE) exams over their lifetime. Objective To determine whether gadolinium-based contrast agents are necessary in assessing active bowel inflammation on MRE in pediatric patients with known or suspected IBD. Materials and methods We conducted a retrospective study of 77 patients (7–18 years; 68.8% male) with known ( n =58) or suspected ( n =19) IBD and endoscopy with biopsy performed within 30 days of MRE without and with contrast evaluated bowel and non-bowel findings. During three visual analysis sessions, two radiologists reviewed pre-, post-, and pre-/post-contrast MRE images. A third radiologist independently reviewed 27 studies to assess inter-reader reliability. We used Cohen kappa (κ), Fleiss kappa, (κ F ), McNemar test, and sensitivity and specificity to compare MRE readings to combined endoscopic/histopathological findings (the reference standard). Results The pre- and pre-/post-contrast-enhanced MRE vs. combined endoscopic/histopathological results had moderate agreement (85.7%; κ 0.713, P