P495 Usefulness of a capsule endoscopy scoring system in understanding the disease state of Crohn’s disease
T. Omori, H. Kambayashi, S. Murasugi, A. Ito, M. Yonezawa, K. Tokushige
Institute of Gastroenterology, Tokyo Women’s Medical University, Tokyo, Japan
Background
Crohn’s disease (CD) is a pathological condition that develops because of inflammatory and organic changes such as stenosis. The Lewis Score (LS) and Capsule Endoscopy Crohn’s Disease Activity Index (CECDAI) are employed for scoring small bowel lesions in CD using small bowel capsule endoscopy (SBCE). However, it remains unclear whether the obtained score can contribute to the understanding of the disease state or the selection of treatment. This study aimed to examine whether the existing scoring systems (LS and CECDAI) and the new Crohn’s Disease Activity in Capsule Endoscopy (CDACE) scoring system proposed by us were useful in understanding the disease state of CD.
Methods
After obtaining the results of intestinal patency evaluations using patency capsules from June 2012 to June 2018, 196 times (108 CD patients) who underwent SBCE were included in this retrospective study. The cut-off values of LS, CECDAI, and CDACE were calculated using the receiver operating characteristic curve to determine the extent of small intestinal lesions and the presence or absence of stenotic lesions. Using the cut-off values, the sensitivity (Se), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV) of each scoring system were determined. In CDACE, the small bowel is divided into quartiles, and the degree of inflammation (range: 0–4) for each quantile (range: 0–16) (inflammation score: A), number of quartiles with inflammation (range: 0–4) (range score: B), and degree of stenosis (range: 0–3) (stenosis score: C) were scored; the CDACE score was calculated as follows: A × 100 + B × 10 + C (score range: 0–1643).
Results
For LS, the Se, Sp, PPV, and NPV were 83.7%, 73.5%, 75.9%, and 81.8%, respectively, for inflammation in multiple quantiles (2/3 quantiles or more) (cut-off value: 10) and 94.9%, 90.5%, 71.2%, and 98.6%, respectively, with regard to the presence or absence of stenosis (cut-off value: 196). For CECDAI, the Se, Sp, PPV, and NPV were 93.8%, 86.1%, 82.6%, and 93.4%, respectively, for multiple quantiles (both proximal and distal) (cut-off value: 5) and 51.3%, 87.9%, 51.3%, and 87.9%, respectively, with regard to the presence or absence of stenosis (cut-off value: 11). For CDACE, the Se, Sp, PPV, and NPV were 100%, 82.4%, 78.6%, and 100%, respectively, for multiple quantiles (multiple locations excluding two consecutive zones) (cut-off value: 320) and presence or absence of stenosis was identified in the first digit.
Conclusion
CECDAI was more likely to predict the extent of inflammation than LS, although it was difficult to determine the presence or absence of stenosis. Based on the scores, CDACE may be a useful scoring system in clinical practice where the disease state of CD needs to be determined.