DNA Methylation Microarray Identifies Novel Methylation Biomarkers for Colorectal Cancer Detection

Abstract

Background & Objective: In the United States, colorectal cancer (CRC) is the third most prevalent and deadly malignancy. Development of novel CRC-specific abnormal DNA biomarkers may advance non-invasive, cost-efficient population-based CRC screening and ultimately reduces CRC death. DNA hypermethylation is a common epigenetic abnormality in CRCs and represents a promising class of cancer biomarkers. The objective of the current study was to develop optimal DNA hypermethylation-based biomarkers for use in fecesor serum-based average-risk CRC screening. Design: We first applied DNAmethylationmicroarray analysis in order to identify novel loci demonstrating neoplasia-specific methylation in the colon. This array analysis allowed us to investigate 55% of CpG islands within the genome and was applied to 17 primary CRCs relative to 8 non-neoplastic colonic tissues (NCs) from neoplasia-free subjects. The detected CRC-associated hypermethylation events were then individually measured in 113 colonic tissues comprising 51 CRCs, 9 adenomas, 19 NCs from CRC patients (CRC-NCs), and 34 NCs from neoplasia-free subjects (control NCs) using highly sensitive and quantitative real-time quantitative methylation-specific PCR (qMSP) assays. Receiver-operator characteristics (ROC) curve analysis was applied to the qMSP data in order to assess each individual event as well as combination of events for their ability to discriminate neoplastic from non-neoplastic cases. Results: Microarray-based global methylation profiles discriminated CRCs from NCs. A bioinformatic filtering of the microarray data identified 169 candidate CRC-associated hypermethylation events, including one previously validated hypermethylation marker for fecal DNA-based CRC detection, SFRP2. Fourteen of these 169 loci were evaluated using qMSP assays. Ten of these 14 methylation events significantly distinguished CRCs from control NCs (p<.01). Of these ten events, methylation of VSX2 achieved the highest discriminative accuracy (83.3% sensitivity and 92.3% specificity; Area under ROC curve, or AUC, 0.93, p<1E-6). Similarly, CRC-NCs were significantly discriminated from control NCs by methylation of ALX3 (AUC 0.78, p<1E-4). The discrimination of CRC-NCs from control NCs was improved by a multi-locus methylation panel (AUC 0.83, p<1E-6) relative to ALX3. Although the sample numbers were small, two methylation events significantly distinguished adenomas from control NCs (p<.01). Conclusions: Systematic methylome analysis has identified 11 novel methylation events in neoplastic and non-neoplastic colonic mucosa from CRC patients that accurately discriminate CRC patients from controls. These markers merit further evaluation as candidate biomarkers for stool and circulating DNA-based CRC detection.