Abstract
BACKGROUND: CT enterography and magnetic resonance enterography have emerged as first-line imaging technologies for the evaluation of the gastrointestinal tract in Crohn's disease.
OBJECTIVE: The purpose of this work was to evaluate the accuracy of these imaging modalities to identify Crohn's disease lesions preoperatively.
DESIGN: This was a retrospective chart review.
SETTINGS: The study was conducted at a single institution.
PATIENTS: Seventy-six patients with Crohn's disease with preoperative CT enterography and/or magnetic resonance enterography were included in the study.
MAIN OUTCOME MEASURES: The number of stenoses, fistulas, and abscesses on CT enterography and/or magnetic resonance enterography before surgery were compared with operative findings.
RESULTS: Forty patients (53%) were women, 46 (60%) underwent surgery for recurrent Crohn's disease, and 46 (57%) had previous abdominal surgery. Thirty-six (47%) had a preoperative CT enterography and 43 (57%) had a preoperative magnetic resonance enterography. CT enterography sensitivity was 75% for stenosis and 50% for fistula. MRE sensitivity was 68% for stenosis and 60% for fistula. The negative predictive values of CT enterography and magnetic resonance enterography for stenosis were very low (54% and 65%) and were 85% and 81% for fistula. CT enterography had 76% accuracy for stenosis and 79% for fistula; magnetic resonance enterography had 78% accuracy for stenosis and 85% for fistula. Both were accurate for abscess. False-negative rates for CT enterography were 50% for fistula and 25% for stenosis. False-negative rates for magnetic resonance enterography were 40% for fistula and 32% for stenosis. Unexpected intraoperative findings led to modification of the planned surgical procedure in 20 patients (26%).
LIMITATIONS: This study was limited by its small sample size, its retrospective nature, and that some studies were performed at outside institutions.
CONCLUSIONS: CT enterography and magnetic resonance enterography in patients with Crohn's disease were accurate for the identification of abscesses but not for fistulas or stenoses. Surgeons should search for additional lesions intraoperatively. Patients should be appropriately counseled regarding the need for unexpected interventions (see Video, Supplemental Digital Content 1, http://links.lww.com/DCR/A162).