Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P219 Accuracy of imaging for the preoperative measurement of the ileocolonic Crohn’s disease length: results of a prospective cohort study

A. Brouquet*1, A.-S. Rangheard2, J. Ifergan2, T. Lazure3, F. Carbonnel4, C. Penna1, S. Benoist1

1Bicêtre Hospital - Université Paris Sud, Oncologic and Digestive Surgery, Le Kremlin Bicêtre, France, 2Bicêtre Hospital - Université Paris Sud, Radiology, Le Kremlin Bicêtre, France, 3Bicêtre Hospital - Université Paris Sud, Pathology, Le Kremlin Bicêtre, France, 4Bicêtre Hospital - Université Paris Sud, Gastroenterology, Le Kremlin Bicêtre, France

Background

Indication for surgery in patients with ileocolonic Crohn disease is based on several factors including the presence of complications (stricture, abscess, or internal fistula), response to medical therapy, presence of risk factors of postoperative recurrence, and the length of disease. This study aimed to evaluate the accuracy and reliability of imaging for the measurement of ileocolonic Crohn’s disease length.

Methods

Fifty-four consecutive patients who underwent resection for ileocolonic Crohn’s disease were prospectively included between 2011 and 2014. All patients had preoperative magnetic resonance or computed tomographic enterography. Two independent radiologists scored disease length on imaging. All measurements were compared with disease length assessed at pathology on non-fixed specimen.

Results

Median Crohn’s disease length on imaging was 20.5 cm (range 2–73) and 20 cm (range 3–90), as measured by 2 radiologists. Interobserver agreement was substantial (kappa = 0.69), and the level of correlation significance was 0.001 (r = 0.82). Median pathological Crohn’s disease length was 16.5 cm (range 2–75), and measurement was well correlated to radiologic measurement (r = 0.76, p < 0.001). Crohn’s disease length measurement on imaging was correct ± 5 cm, underestimated, and overestimated in 30 (55%), 6 (11.1%), and 18 (33.3%) patients, respectively. Crohn’s disease length ≤ 20 cm found on imaging in 26 patients was confirmed in 25 (96%) at pathology, whereas length > 20 cm found in 28 patients was confirmed in 18 (64%) at pathology. Imaging sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy for predicting length ≤ 20 cm were 71%, 95%, 96%, 64%, and 79%, respectively.

Conclusion

Imaging accurately and reliably identifies localised Crohn’s disease ileitis ≤ 20 cm. In patients with more extended disease, it tends to overestimate length and should be interpreted with caution.