Ultrasound and MRI predictors of surgical bowel resection in pediatric Crohn disease.

ABSTRACT

BACKGROUND: Imaging predictors for surgery in children with Crohn disease are lacking. OBJECTIVE: To identify imaging features of the terminal ileum on short-interval bowel ultrasound (US) and MR enterography (MRE) in children with Crohn disease requiring surgical bowel resection and those managed by medical therapy alone. MATERIALS AND METHODS: This retrospective study evaluated patients 18 years and younger with Crohn disease undergoing short-interval bowel US and MRE (within 2 months of one another), as well as subsequent ileocecectomy or endoscopy within 3 months of imaging. Appearance of the terminal ileum on both modalities was compared between surgical patients and those managed with medical therapy, with the following parameters assessed bowel wall thickness, mural stratification, vascularity, fibrofatty proliferation, abscess, fistula and stricture on bowel US; bowel wall thickness, T2 ratio, enhancement pattern, mesenteric edema, fibrofatty proliferation, abscess, fistula and stricture on MRE. A two-sided t-test was used to compare means, a Mann-Whitney U analysis was used for non-parametric parameter scores, and a chi-square or two-sided Fisher exact test compared categorical variables. Imaging findings in surgical patients were correlated with location-matched histopathological scores of inflammation and fibrosis using a scoring system adapted from the Simple Endoscopic Score for Crohn Disease, and a Spearman rank correlation coefficient was used to compare inflammation and fibrosis on histopathology. RESULTS: Twenty-two surgical patients (mean age 16.5 years; male/female 13/9) and 20 nonsurgical patients (mean age 14.8; M/F 8/12) were included in the final analysis. On US, the surgical group demonstrated significantly increased mean bowel wall thickness (6.1 mm vs. 4.7 mm for the nonsurgical group; P = 0.01), loss of mural stratification (odds ratio [OR] = 6.3; 95% confidence interval [CI] 1.4-28.4; P = 0.02) and increased fibrofatty proliferation (P = 0.04). On MRE, the surgical group showed increased mean bowel wall thickness (9.1 mm vs. 7.2 mm for the nonsurgical group; P = 0.02), increased mean T2 ratio (4.6 vs. 3.6 for the nonsurgical group; P = 0.03), different enhancement patterns (P = 0.03), increased mesenteric edema (P = 0.001) and increased stricture formation (OR = 8.2; 95% CI 1.8-36.4; P = 0.005). Nineteen of 22 ileocecectomy specimens showed severe inflammation and 21/22 showed severe fibrosis, with significant correlation between inflammation and fibrosis scores (ρ = 0.55; P = 0.008); however, correlation with imaging findings was limited by the uniformity of findings on histopathology. CONCLUSION: Children with terminal ileal Crohn disease requiring surgical bowel resection demonstrate more severe manifestations of imaging features traditionally associated with both active inflammation and chronic fibrosis than those managed medically on US and MRE, findings that are corroborated by histopathology. These features may potentially serve as imaging biomarkers indicating the necessity for surgical intervention.