P128 Development of the Paediatric Inflammatory Crohn's MRE Index (PICMI)- results from the ImageKids Study

Abstract

Objectives and Study: There is no validated MRE-based inflammatory index for children although Crohn's disease (CD) is more often panenteric compared to adults. The aim of the multicenter prospective ImageKids study was to develop such an index termed Pediatric Inflammatory Crohn's disease MRE Index (PICMI). Method(s): 158 children with CD (age 14+/-2.4 years, 54% males, median disease duration (2.1 (IQR 0.3-4.3) years) underwent an MRE and ileocolonoscopy along with extensive collection of clinical data including SESCD and wPCDAI and bio-sampling in 21 sites globally. All MREs were scored independently by the site radiologist and by two central radiologists; the bowel was divided into 20cm segments and sections (jejunum, ileum TI and colon). Measured variables included length of involved segments, a radiologist global assessment of inflammation on a 0-100mm visual analog score (VAS) and 9 items selected by a Delphi group of 30 international radiologists and a systematic review of the literature (reported separately): wall thickness, T2 intensity, enhancement, DWI, narrowed lumen, comb sign, ulcerations, vascular engorgement and motility as well as the adult MaRIA score. Result(s): 368 involved bowel segments were scored in 158 enrolled children. Based on univariate analyses, the following items were selected for the PICMI Index: wall thickness, T2 intensity, DWI, narrowed lumen, comb sign, ulcerations, and mesenteric T2 intensity (all P< 0.001; Figure1). In an attempt to avoid gadolinium exposure in children, mucosal and mesenteric enhancement were not selected but will be evaluated further in the validation stage. Linear regression was used to determine the weights of scoring between bowel sections. The radiologist assessment was found to be well predicted by linear regression on the index variables. The models showed satisfying linear fit; for some sections retained only some of the variables (jejunum R2 =0.80, ileum R2 =0.30, TI R2 =0.41; all p< 0.0001) Conclusion(s): The PICMI index is being developed as a MRE tool for pediatric CD to supplement endoscopic assessment. The draft weighted PICMI will be now validated on a separate cohort and compared with the MaRIA (Figure presented).