Identifying predictive signalling networks for Vedolizumab response in ulcerative colitis

Area of Science:

• Immunology
• Molecular Biology
• Bioinformatics

Background:

• Ulcerative colitis (UC) patient response to biologic therapy is unpredictable.
• A significant unmet need exists for biomarkers to guide UC therapy selection.
• Predicting anti-integrin therapy response in UC is essential for personalized treatment.

Purpose of the Study:

• To identify predictive signaling pathways for anti-integrin therapy response in UC patients.
• To develop biomarkers for stratifying UC patients based on treatment efficacy.
• To enable personalized therapy selection for ulcerative colitis.

Main Methods:

• Utilized two public gene expression datasets (GSE73661, GSE72819) of UC patients treated with anti-integrin therapies (Vedolizumab, Etrolizumab).
• Applied a diffusion-based signaling model focused on T-cell receptor signaling networks.
• Analyzed network connectivity between receptors and transcription factors to predict treatment response.

Main Results:

• The diffusion model accurately predicted Vedolizumab (VDZ) and Etrolizumab response, outperforming original gene expression analysis.
• Identified 48 receptor-TF pairs predicting VDZ response (AUC ≥ 0.76), with FFAR2-NRF1, FFAR2-RELB, FFAR2-EGR1, and FFAR2-NFKB1 as top predictors.
• Discovered 40 receptor-TF pairs for Etrolizumab, with CD40-NFKB2 as the best predictor (AUC = 0.72).
• Highlighted subnetworks in cytokine and fatty acid signaling, suggesting their role in stratifying UC patient subgroups.

Conclusions:

• Identified novel signaling pathways capable of predicting anti-integrin therapy efficacy in UC.
• Findings support the development of a clinical decision-making tool for UC patient stratification.
• Potential for advancing personalized therapy alternatives for ulcerative colitis patients.