DOP054 Identification of a Molecular Signature of Disease Severity in Ulcerative Colitis Using a Systems Biology Approach and Clinical Impact Using VARSITY Trial Data

Abstract Background We aimed to develop and validate a transcriptomic signature that correlates with clinical disease activity and therapy response using the Inflammatory Bowel Disease (IBD) Plexus® real-world data platform (Crohn’s & Colitis Foundation), CytoReason Disease Model based on published ulcerative colitis (UC) molecular studies, and a large prospective randomized clinical trial (VARSITY1). Methods A molecular disease activity score was built by identifying shared differential expression signals in transcriptomics data from inflamed v healthy control UC samples using an elastic net regression model trained against disease activity scores.2 The resulting model assigned a continuous molecular activity score per sample based on expression of the selected genes, with interpretability supported by feature importance and downstream biological evaluation. Signature module involvement was assessed using disease model data from 1985 mucosal biopsies from 18 UC clinical studies, including pts treated with golimumab, infliximab or vedolizumab. Pathway differential expression analysis was performed using gene set variation analysis (GSVA) with validation using molecular data from VARSITY.1 Results An 85 gene signature was identified, which represents enrichment of endothelial cells, alpha-beta T cells, epithelial, fibroblasts, and plasma cells (Figure). In contrast, published molecular inflammation gene sets (400–600 genes) 3, 4 are primarily enriched for immune and epithelial cells.3, 4 Our molecular activity signature also captured tertiary lymphoid structure, vascular, connective tissue, and hematopoiesis signaling—processes implicated in severe UC—thus demonstrating broader representation of disease mechanisms. Our molecular signature correlated with Mayo total/endoscopic scores across multiple IBD cohorts, and real-world registries eg., IBD Plexus. Assessment of our gene signature in VARSITY data revealed a significant correlation with therapy response: ie.,change in endoscopic and histological activity at week 52 (Table). Conclusion We developed and validated a refined molecular signature of UC disease activity encompassing diverse cellular populations which correlated with change in endoscopic and histologic disease severity in a prospective centrally read clinical trial. Validation across multiple real-world registries and clinical trials highlights its potential as a biomarker for clinical evaluation and the gene panel size relative to prior work offers enhanced feasibility for use to define residual disease burden in pts with endoscopic and histologic improvements in disease activity. This is an important step towards precision medicine in IBD using a molecular endotype-based classification of UC. References: 1.Sands B, et al. Vedolizumab versus Adalimumab for Moderate-to-Severe Ulcerative Colitis. N Engl J Med. 2019;381(13):1215-1226. 2.Zou H, et al. Regularization and Variable Selection Via the Elastic Net. J R Stat Soc Series B Stat Methodol. 2005;67(2):301-320. 3.Argmann C, et al. Biopsy and blood-based molecular biomarker of inflammation in IBD. Gut. 2023;72(7):1271-1287. 4.Friedrich, M. et al. IL-1-driven stromal-neutrophil interactions define a subset of patients with inflammatory bowel disease that does not respond to therapies. Nat Med. 2021;27(11):1970-1981. Conflict of interest: Dulai, Parambir: Grant:Takeda, Bristol Meyer Squibb, Merck, Genentech, Geneoscopy. Personal Fees: Abbvie, Abivax, Alimentiv, Bristol Meyer Squibb, Boehringer Ingelheim, Celltrion, Cristcot, Genentech, Geneoscopy, Janssen, Lilly, Merck, Pfizer, Sanofi, Takeda. Panchal, Pratik: Employee of Takeda and holds stock/options in Takeda. Tang, Jay: Employee of Takeda and holds stock/options in Takeda. Wennbo, Haakan: Takeda employee Agboton, Christian: Employee of Takeda and holds stock/options in Takeda. Romo Bautista, Itzel: Employee of Takeda and holds stock/options in Takeda. Cogan, Adi: Employee and shareholder of CytoReason, which has received research funding from Takeda. Wyrobnik, Iris: Employee and shareholder of CytoReason, which has received research funding from Takeda. Farahmand, Saman: Employee of Takeda and holds stock/options in Takeda. Paraskevopoulou, Maria: Other: Maria Paraskevopoulou is employee and shareholder of Takeda Schuster, Ronen: Employee and shareholder of CytoReason, which has received research funding from Takeda. Sirota-Madi, Alexandra: Employee and shareholder of CytoReason, which has received research funding from Takeda. Shen-Orr, Shai: Consultant for and shareholder of CytoReason Chowers, Yehuda: Employee and shareholder of CytoReason, which has received research funding from Takeda. Arunachalam, Vinayagam: Employee of Takeda and holds stock/options in Takeda.