Pre-operative Small Bowel Follow Through Detects more Crohnʼs Disease Strictures than CT Enterography in Patients with Multiple Strictures: Comparison with Intra-operative Findings

Abstract

Purpose: Computerized Tomography Enterography (CTE) has emerged as an imaging procedure for assessment of CD activity and anatomy. Comparison of CTE with small bowel follow through (SBFT) for detection of strictures using surgical findings as the gold standard in CD has not been performed. We sought to determine the accuracy of preoperative SBFT and CTE studies and its correlation with the intra-operative strictures or fistulas in CD pts undergoing SB surgery. Methods: This was a retrospective cohort study at a tertiary referral center. CD pts who had undergone SB surgeries between 1/2006 and 9/2008 and had a preoperative SBFT and/or CTE within 4 months prior were included. This interval was based on our protocol that optimizes medical therapy over 8-12 weeks preceding surgery. We reviewed the radiology reports and documented the number of strictures and/or fistulas seen by either modality. All surgeries were performed by a single CD surgeon. The number of intra-operative strictures was determined by the insertion of a Foley catheter to 200 cm from the resected segment followed by 2cm balloon inflation with pull through. Chi-square tests were used to compare detection rates between the different techniques. Results: Sixty-one CD pts who underwent SB surgery and had at least one preoperative imaging study within 4 months prior. Of those, 53 pts had SBFT, 19 pts had CTE and 12 pts had both. On a per patient analysis, SBFT correctly identified 48/51 (94.1%) pts who were found to have any stricture which was higher than the detection rate for CTE (12/15 pts, 80%, p=0.09). Using the most conservative estimate for test performance, SBFT correctly identified all strictures in 73.2% of pts, a rate similar to CTE (57.1%, p=0.3). On a per stricture analysis, SBFT identified 62/88 (70.5%) strictures found during surgery compared to 11/21 for CTE (52.4%, p=0.09). With pts with at least 2 strictures the sensitivity was 52.6% for SBFT and 16.7% for CTE (p=.11), however on a per stricture analysis SBFT had a significantly higher detection rate compared to CTE, 62.1 vs. 30.8 (p=0.03). In the subgroup of pts who had had both examination within 4 months prior to surgery, SBFT identified 8/11 strictures (72.8%) compared to 5/11 for CTE (45.5%, p=0.18). With regards to detection of fistulas, both CTE (4/6, 67%) and SBFT (7/12, 58.3%) performed comparably (p=0.7). Conclusion: In our CD cohort, SBFT showed a trend towards more accurate detection of strictures compared to CTE with equivalent performance for fistulas detection. In patients with multiple strictures SBFT had significantly better detection rate than CTE on a per stricture basis. Larger, head-to-head studies may be required to determine the optimal pre-operative SB imaging modality.