P312 Retrospective study using ultrasonography in predicting clinical relapse of Crohn's disease

Abstract

Background Endoscopy is the standard modality for inflammatory bowel disease, but it is invasive and poorly tolerated. On the other hand, trans-abdominal ultrasonography (US) is minimally invasive, inexpensive, and has excellent characteristics not found in other examinations, such as the ability to detect minute changes in the intestinal wall. It would be clinically useful if US could predict clinical relapse. Methods Of 185 patients with CD who underwent US between April 2011 and April 2021, we retrospectively evaluated 75 patients who were in clinical remission (CDAI < 150) at the time of their first US at our hospital for clinical relapse and maintenance of remission at 1 year after the procedure. The severity of US was evaluated by US-CD (0-52 points in total), which divided the intestine into four segments(ileum, right-sided colon, transverse colon, and left-sided colon) and was calculated the following US parameters: bowel wall thickness (BWT) (0–3), loss of stratification (0–2), degree of blood flow signaling by a color Doppler study (0–3), presence of increasing echogenicity mesentery (0–2), and intestinal stenosis (0–3). US images were analyzed and interpreted in a consensus manner by two technologists with more than 10 years of experience, blinded to the patient's clinical information. Results Of the 75 cases, 58 were males and 17 were females with a median age of 30 years (range 18-58), median BMI of 20.6 kg/square meter (range 16.2-32.5), 23 cases were ileal type, 47 cases were ileocolonic type, and 5 cases were colonic type. Sixty-eight patients remained in clinical remission during the observation period, and 7 patients had clinical relapse (CDAI > 150). The median CDAI at US was 71 (range 14-143), median CDAI at relapse was 189 (range 151-288), and median observation period was 365 days (range 6-365). The median time to relapse was 193 days (range 166-288). ROC analysis was performed on the US-CD, and the highest AUC of 0.820 was obtained when the cutoff value was 11. Six patients in the relapse group and 20 patients in the remission maintenance group had US-CD of 11 or higher, and 1 patient in the relapse group and 48 patients in the remission maintenance group had a score of less than 11. The sensitivity was 85.7%, specificity was 70.6%, positive predictive value was 23.1%, and negative predictive value was 98%. Logistic regression analysis identified "US-CD 11 or higher" as a factor contributing to clinical relapse (p=0.018). Conclusion US-CD ≥11 is a prognostic predictor of clinical relapse in CD.