Abstract 220: Early methylome-level alterations in colorectal adenocarcinoma

Abstract Introduction: Methylation-based biomarkers represent a promising tool for early tumor detection in colorectal cancer. Genome-wide methylation analysis of large cohorts is an effective way of for the discovery of new biomarker candidates. For such purposes, Illumina arrays are today the most popular sources of methylation data. Purpose: We aimed to analyze the methylation pattern differences between different groups, especially those involved in normal-adenoma and adenoma-adenocarcinoma transition. In order to enable straightforward exploration of the database we also aimed to establish an interactive web-based platform. Methods: Raw files of 2, 356 samples, out of which 1, 273 were from normal mucosa, 815 from adenocarcinoma, and 258 from precancerous adenoma tissues were processed in R using the minfi and the watermelon libraries. The dataset contains data from both Illumina HumanMethylation450K and EPIC microarray platforms, which were analyzed separately. Based on a matrix of β-values, methylated regions and CpG loci were identified using Kruskal-Wallis-test complemented with ROC analysis. For data visualization we created www.epigenplot.com, a web platform designed for the comprehensive analysis of our database. Results: When compared to normal tissues, the average methylation of both adenoma and adenocarcinoma samples was significantly lower in both HM450K and EPIC arrays (Kruskal-Wallis, p<0.001). The majority of the significant (Δβ >= 0.2, p < 0.001) methylation changes discovered in adenocarcinomas were already detectable in adenoma tissues. This occurred more often for hyper- than hypomethylation (HM450K: 80.5% and 78.58%; EPIC: 79% and 41%, respectively). Out of 75 genes previously proposed in the literature as CRC biomarker candidates, 49 and 45 (HM450K and EPIC) were significantly differentially methylated in adenocarcinoma tissues. Notably, for all these genes, the change was already present in adenoma tissues. When the difference between normal and adenocarcinoma tissues was analyzed, the methylation of the first exon of the C1orf70 gene reached the highest AUC 0.98 at both platforms. For this region, the AUC for normal and adenoma tissues was 0.97 in HM450K and 0.99 in the EPIC dataset. Conclusions: Alternations of the methylome are early events of tumorigenesis. The changes detected by methylation-based biomarkers are already present in precancerous lesions. Our web tool can be utilized for the discovery and validation of biomarkers suitable the early detection of CRC using genome-wide data. Citation Format: Dalma Muller, Balazs Gyorffy. Early methylome-level alterations in colorectal adenocarcinoma [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 220.