Abstract 5024: Multiomics signature reveals network regulatory mechanisms in colorectal cancer progression: Going beyond the scope

Abstract Background: Colorectal cancer (CRC) progresses from normal tissue to advanced precancerous lesions (APLs) and ultimately to malignancy, driven by stepwise genetic and epigenetic alterations. A key challenge in managing CRC is the risk of overdiagnosis in APLs, which can lead to unnecessary interventions. Identifying molecular signatures that accurately distinguish between low-grade dysplasia (LGD) and high-grade dysplasia (HGD) is crucial for optimizing risk assessment and clinical decision-making. This study integrates methylation, gene expression, and mutation analyses to uncover transcription factor (TF)-driven regulatory networks, with the aim of developing signatures that could support risk-optimization scores and reduce diagnostic uncertainty. Methods: A differentially methylated signature (DMS), previously identified from HGD and LGD samples using EM-seq and bioinformatics, was applied to the TCGA CRC methylation dataset. Unsupervised hierarchical clustering based on this DMS was performed to identify methylation-driven clusters, followed by a survival and a mutational burden analysis to assess potential associations between clusters and patient outcomes. Cluster-dependent regulatory patterns were used to construct transcription factor (TF) target networks, with hub TFs identified through network topology analysis. Functional enrichment analyses, including Gene Ontology (GO) and pathway analyses, of hub TF target genes were conducted to reveal biological processes specific to each cluster. Results: The DMS effectively distinguished 4 Clusters (CLs), one with Normal samples and 3 tumoral CLs. A continuum in methylation patterns correlated with mutational burden (p = 7.2e-4) and patient survival probability, with Tumoral CL1 showing the highest impact compared to Tumoral CL3 (p = 2.9e-2). TF enrichment and networks analysis revealed by variations in network topology metrics (p = 4.2e-10), shows unique regulatory networks CLs-dependent. Functional enrichment analysis highlighted cluster-specific pathways, such as PI3K-Akt and AMPK signaling, besides cancer specific pathways in CL1; and JAK-STAT, TNF, and IL-17 signaling in Cl3. Conclusions: This multiomics analysis reveals heterogeneity in CRC by identifying stratified regulatory mechanisms and biological processes driving progression from APLs to different CRC subtypes. By integrating these insights into risk-optimization scores, the identified signatures could enhance diagnostic accuracy, mitigate overdiagnosis, and refine patient management strategies. These findings pave the way for the development of precision diagnostic tools that optimize clinical decision-making while minimizing unnecessary interventions. Further validation is required to translate these signatures into robust clinical applications. Citation Format: Juan Carlos Higareda-Almaraz, Francesco Mattia Mancuso, Pol Canal-Noguer, Kristi Kruusmaa, Arianna Bertossi. Multiomics signature reveals network regulatory mechanisms in colorectal cancer progression: Going beyond the scope [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5024.