P123 Characteristics of small bowel involvement and association with clinical outcome in patients with Crohn’s disease
H Yoon,J Park,H Y Kim,Y J Lee,C M Shin,Y S Park,N Kim,D H Lee
DOI: https://doi.org/10.1093/ecco-jcc/jjab232.251
2022-01-01
Journal of Crohn's and Colitis
Abstract:Abstract Background The location and number of involved sub-segments in small bowel Crohn’s disease (CD) vary on each patient. This study aimed to evaluate the distribution and characteristics of involved small intestine lesions through a detailed image review at the time of diagnosis and to evaluate the relationship between them and clinical outcomes. Methods CD patients who underwent computed tomography (CT) at the initial diagnosis were retrospectively enrolled. Two abdominal radiologists reviewed the CT images of all patients. They independently rated the presence of “bowel wall thickening”, “stricture”, and “fistula or abscess” in the jejunum, distal jejunum/proximal ileum, ileum, and terminal ileum, respectively. Score 1 was assigned if both radiologists concurred regarding the presence of a lesion. Based on the total score summed across different locations in the small bowel (e.g., imaging score), we divided the patients into three groups as 0, 1–2, and 3–4. Major clinical outcomes were compared among these groups. Results A total of 152 patients were enrolled. The proportion of the patients involved in the jejunum, distal jejunum/proximal ileum, ileum, and terminal ileum was 2.0%, 30.3%, 82.2%, and 71.7%, respectively; 32.3% of the patients had Montreal classification L4 (involvement of jejunum or distal jejunum/proximal ileum). Agreement on fistula or abscess was almost perfect (kappa value = 1.0 for most segments); the degree of agreement on bowel wall thickness and stricture varied according to involved segments (kappa value = 0.690–0.936). There were 15 (9.9%), 100 (65.8%), and 37 (24.3%) patients in groups 0, 1–2, and 3–4, respectively. In Kaplan-Meier survival analysis, there was no difference in the major clinical outcomes among three groups based on the imaging score (emergency room visit p = 0.093; operation p = 0.080; new use of corticosteroid p = 0.139; and new use of biologics p = 0.126). In the multivariable analysis, disease behavior was the only risk factor associated with all clinical outcomes after adjusting other factor (emergency room visit, HR 2.127 [1.356–3.337], p = 0.001; operation, HR 8.216 [2.629–25.683], p < 0.001; new use of corticosteroid, HR 1.816 [1.249–2.642], p =0.002; and new use of biologics, HR 2.352 [1.492–3.708], p < 0.001). Conclusion As a result of a detailed CT review, about one-third and one-fourth of patients newly diagnosed as small bowel CD had L4 lesions and multiple lesions (> 2 segments), respectively. However, the location and number of involved sub-segments in small bowel CD did not affect the clinical outcomes. Initial disease behavior is a more critical factor associated with clinical outcomes than characteristics of disease location.