P659 Prognostic markers in Treat-to-Target for Crohn’s Disease- an Asian institution’s experience

Abstract

Abstract Background Endoscopic healing (EH) is a long-term treatment target in the Treat-to-Target approach for Crohn’s Disease (CD). However, not all patients with CD will achieve this treatment target due to a combination of factors. This study aims to evaluate the proportions of patients with CD who achieve EH at 24 months from diagnosis. Secondary aims include proportion of patients who achieve corticosteroid free clinical remission (CSFR), biomarker (BM) normalization and predictors of failure to achieve EH. Methods This is a retrospective study of all patients with Crohn’s disease seen at a tertiary hospital in Singapore from January 2010 to December 2021. Patients with a confirmed diagnosis of CD based on endoscopic, clinical and histological findings were identified from existing IBD registry. Those with a follow-up period of less than 12 months, no repeat endoscopy and missing data were excluded from analysis. We collected data on baseline demographics, disease characteristics and treatment received. Treatment targets were assessed in terms of CSFR, biomarkers normalization, EH and transmural healing (when available). We defined these targets as outlined in the selecting therapeutic Selecting Therapeutic Targets in Inflammatory Bowel Disease (STRIDE) II initiative. Statistical analysis was performed in Python using statsmodels 0.14.0. Results We identified a total of 86 CD patients. After excluding 29 patients, 59 patients were included in the final analysis.Mean age at diagnosis was 33 years old, and 69.5% were males. Table 1 demonstrates the baseline demographics and disease characteristics of all patients. The median follow-up period was 101 months. Out of 59 patients, 88% (n=52) and 80% (n=47) achieved CSFR and biomarker normalization respectively at 12 months. At 24 months of follow up, EH was achieved in 54% (n =32) of all patients. Of those who achieved EH, 22 had repeat imaging and transmural healing was documented in 14 patients. On multivariate logistic regression, Montreal B3 (penetrating type) disease was predictive of failure to achieve EH (OR 9.8, p= 0.036). Use of biologics also showed a trend for failing to achieve EH though this was not statistically significant (OR 3.5, p = 0.075). Conclusion A vast majority of CD patients were able to achieve corticosteroid free clinical remission and biomarker normalization. Penetrating disease was predictive of failure to achieve EH at 24 months. This study provides additional evidence in determining treatment targets for CD patients based on their disease phenotypes. These findings could be further validated in larger cohort, prospective studies.