P1310 Meta-analysis of gene signatures predicting response to anti-TNFa therapy of patients with ulcerative colitis

Abstract Background TNFa inhibitors (TNFi) have improved treatment of moderate to severe ulcerative colitis (UC), yet up to 40% of patients show primary non-response to TNFi, risking disease progression and reduced efficacy of later biologics. Several gene signatures of intestinal biopsy have been described as predictive biomarker for TNFi-response in UC patients; however, none of them were translated into clinical practice. Methods Published gene signatures predictive of TNFi-response in UC patients were compared for predictive power by calculating AUC for a combined dataset of public responder/non-responder cohorts (GS12251, GSE16879-UC), gene composition, pathway enrichment analysed with Cytoscape/ReactomeFI, and single cell expression based on a public intestinal scRNA cohort (SCP259) analysed with BioTuring. Results From 13 publications, 33 gene signatures ranging from 1–502 genes (median 20) and comprising 1,087 unique genes were included in this analysis. The highest predictive power achieved by a signature was AUC = 0.92. Signatures were stratified as high (n = 3, AUC 0.9), mid (n = 17, 0.8 AUC 0.9), and low (n = 13, AUC 0.8) predictors. Gene compositional overlap between the signatures was median 0% (mean 7%) across all signatures and median 7% (mean 16%) across high and mid predictor signatures. The top 20 most represented genes were present in 5 (15%) – 11 (33%) of the signatures. OSM gene was present in 8 (24%) signatures, which were all high or mid predictor. With functional network analysis, 189 of the 574 high and mid predictor signature genes clustered into 7 pathway modules, that were dominated respectively by extracellular matrix organization, neutrophil chemotaxis, cell surface receptor signalling, antiviral defence, JAK–STAT signalling and MHCII antigen presentation pathways. Cytokine pathway enrichment highlighted IL-4, IL-13, IL-10, IL-17, TNF and Gai pathways which had the lowest false discovery rate ( 3.7×10-13) and were present across 80% of the signatures. The signatures were most expressed in non-regulatory T cells, dendritic cells, monocytes, macrophages, fibroblast and endothelial cells, where median 35, 36, 32, 31, 36 and 28% of a signature was expressed in the respective cell population. Conclusion Despite of their similar predictive power, the analysed gene signatures had high degree of heterogeneity in composition. Although common traits were noticeable, signature heterogeneity was dominant in pathway composition and distribution across cell populations as well. We conclude that combination of these signatures might result in higher predictive power for response to TNFi in UC by better accounting for disease heterogeneity. References: West NR, Hegazy AN, Owens BMJ, et al. Oncostatin M drives intestinal inflammation and predicts response to tumor necrosis factor-neutralizing therapy in patients with inflammatory bowel disease. Nat Med. 2017;23(5):579-589. Arijs I, Li K, Toedter G, et al. Mucosal gene signatures to predict response to infliximab in patients with ulcerative colitis. Gut. 2009;58(12):1612-1619. Arijs I, Quintens R, Van Lommel L, et al. Predictive value of epithelial gene expression profiles for response to infliximab in Crohn’s disease. Inflamm Bowel Dis. 2010;16(12):2090-2098. Martin JC, Chang C, Boschetti G, et al. Single-Cell Analysis of Crohn’s Disease Lesions Identifies a Pathogenic Cellular Module Associated with Resistance to Anti-TNF Therapy. Cell. 2019;178(6):1493-1508.e20. Conflict of interest: Dr. Pérez, Tamara: No conflict of interest Ahn, Jaeil: No conflict of interest Dayton, Frances: No conflict of interest Doumas, Stavros: No conflict of interest Jacobs, Miriam T: No conflict of interest Marki, Alex: No conflict of interest