Assessing the Variability Between Endoscopic Scoring Indices for Evaluation of Crohnʼs Disease Activity

Abstract

Introduction: Crohn's disease endoscopic index of severity (CDEIS) and simple endoscopic score for Crohn's disease (SES-CD) are considered as gold standards to assess the state of mucosal disease in CD patients and determine the outcome of clinical trials that utilize mucosal healing as an endpoint. Although a linear offset between CDEIS and SES-CD is widely accepted, a closer look at the accuracy and impact of using the two scoring indices has not been adequately explored.Figure: Contingency tables pre- and post-normalization of CDEIS and SES-CD scores.Methods: Both CDEIS and SES-CD scores were collected at the same time in a centrally read, prospectively collected, longitudinal cohort of 118 CD patients in the TAILORIX clinical trial. Up to 3 endoscopic scores were available from each patient over a period of 1 year. Standard disease severity definitions were applied to both CDEIS and SES-CD scores. CDEIS scores were classified as remission < 3, mild 3-8, moderate 9-12, and severe >12. For SES-CD the same groups were defined as < 3, 3-6, 7-15, and >15, respectively. The two indices were normalized using linear regression and contingency tables were created for both pre- and post-normalization for the categorical outcomes of those endpoints. Results: Using the raw CDEIS and SES-CD scores, a contingency table (Table 1; Non-adjusted Agreement) showed that the overall agreement in disease severity states (Remission, Mild, Moderate and Severe) was only 59% (241/411). 33% (58/175) of the scores deemed as disease in remission by CDEIS were classified as active disease by SES-CD. Likewise, 81/146 (˜56%) of mild disease classifications by CDEIS were suggested to have moderate disease with SES-CD. After applying the observed linear regression equation (CDEIS = 0.25 + 0.69*SESCD; r = 0.92) to normalize the two scores the overall agreement improved to 80% (Table 1; Normalized Agreement). Conclusion: Although CDEIS and SES-CD scores correlate well and are often considered as endoscopic gold standard endpoints, our data demonstrates that the two are not equivalent as their ‘endoscopic categorical calls' are in agreement only 59% of the time. The data further shows that even after accounting for their known offset, their agreement is still only 80%. These results highlight the difference in the two currently accepted gold standards and elucidate the importance and clinical unmet need for establishing a single objective score to assess the mucosal state in patients with CD.