OP05. Development of three practical indices for mucosal healing among patients with moderate to severe Crohn's disease
W. Sandborn1, R. Panaccione2, J.‑F. Colombel3, E. Louis4, M. Yang5, R. Thakkar5, J. Chao5, P. Mulani5
1University of California, San Diego, La Jolla, United States; 2University of Calgary, Director, Inflammatory Bowel Disease Clinic, Calgary, Canada; 3Centre Hospitalier Universitaire de Lille, Hôpital Claude Huriez, Lille, France; 4University of Liège and CHU Liège, Department of Gastroenterology, Liège, Belgium; 5Abbott Laboratories, Illinois, United States
Background: Healing of the intestinal mucosa is considered a valid and important metric of improvement in Crohn's disease (CD). However, a patient must undergo an endoscopy to verify the presence of mucosal ulceration. Regular monitoring using endoscopy is challenging owing to its expensive and intrusive nature. There are currently no easily accessible proxy measures for identifying patients who are likely to have achieved mucosal healing (MH) in CD.
Methods: The practical MH indices were developed using intent-to-treat (ITT) patient-level data from the EXTend the Safety and Efficacy of Adalimumab Through ENDoscopic Healing (EXTEND) trial (160 mg adalimumab at Week 0, 80 mg at Week 2, and randomised to 40 mg every other week or placebo at Week 4). Each index corresponds to 1 of 3 mucosal healing measures: Simple Endoscopic Score for CD (SES-CD) <5, CD Endoscopic Index of Severity (CDEIS) score ≤4.24, and observed MH based on a review committee opinion. For each measure, a logistic regression model was developed to predict MH at Week 12 using demographic, laboratory, and CD severity characteristics. Important predictors were selected using the Least Absolute Shrinkage and Selection Operator (LASSO) procedure, which identifies a parsimonious model while retaining predictive accuracy. These 3 indices, named as SES-CD Based Index, CDEIS Based Index, and Observed MH Based Index, were evaluated using the accuracy (c-statistic), sensitivity, and specificity measures.
Results: Important factors associated with MH were similar across the 3 indices (Table). The performance of each index was also evaluated and listed in the table.
|Name||Components of Each Index (important factors associated with MH)||Accuracy (c-statistic)||Sens (%)||Spec (%)|
|SES-CD Based Index||Female, platelet count, CRP, rectal bleeding, a CDAI (stool)b||0.81||82.1||69.4|
|CDEIS Based Index||Age, albumin, platelet count, CRP, CDAI (stool and general well-being)b||0.84||81.1||73.3|
|Observed MH Based Index||Age, duration of CD, platelet count, CRP, rectal bleeding, a CDAI (stool)b||0.79||44.8||89.9|
Conclusions: Platelet count, CRP, and CDAI (number of liquid or very soft stools) are the common factors associated with MH across all 3 indices. The accuracy rates suggest the practical indices may be a clinically useful guide for identifying patients who are likely or unlikely to have achieved MH.