PD-0013 Methylation Based Biomarkers for Non-Invasive Early Detection of Colorectal Cancer

Abstract

ABSTRACT Introduction Colorectal cancer is a common disease with a world-wide annual incidence of 1.2 million. The mortality rate is high, especially for patients diagnosed with an advanced tumor stage. We recently identified DNA methylation based biomarkers for early detection of colorectal cancer (and unpublished). In the beginning of 2012, these markers were licensed to the British biotech company Oxford Gene Technology for development into a non-invasive test suitable for colorectal cancer screening (http://www.ogt.co.uk/news_events/news/659_ogt_signs_licensing_deal_for_colorectal_ cancer_biomarkers). Two tests based on cancer specific aberrant DNA methylation are currently commercially available, including analyses of VIM and SEPT9 in stool and blood samples, respectively. The aim of the present study was to compare the performance of the in house biomarker panel with previously published epigenetic biomarkers using a large series of tissue samples. Methods Quantitative methylation-specific polymerase chain reaction was used to analyze the promoter methylation status of ITGA4, NTRK2, OSMR, TUBG2, SEPT9, and VIM in test and validation series comprising altogether 169 primary colorectal cancers, 105 matching normal mucosa samples, 104 adenomas, and 107 normal samples from colorectal cancer-free individuals. Receiver operating characteristic (ROC) curves were used to evaluate the overall performance of the biomarkers in tissue samples and the results were compared with the performance of the in house biomarker panel. Results Across the test and validation series, ITGA4, SEPT9, OSMR, VIM, NTRK2, and TUBG2 were methylated in 91%, 82%, 79%, 67%, 7%, and 1% of the colorectal carcinomas, and in 79%, 88%, 78%, 54%, 3%, and 0% of the adenomas. The promoter methylation was highly cancer-specific and only 1%, 3%, 2%, 3%, 0%, and 0% of the normal mucosa samples were methylated for the same genes. Resulting areas under the ROC curves for separating colorectal carcinomas from normal mucosa samples were 0.95, 0.93, 0.92, and 0.80 for ITGA4, SEPT9, OSMR, and VIM, respectively. In comparison, the in house biomarker panel achieved an area under the ROC curve of 0.97. Conclusion In a direct comparison performed by a single lab, using the same protocol and the same series of tissue samples, the performance of our recently identified in house biomarker panel was shown to be better than previously identified DNA methylation based biomarkers, including VIM and SEPT9. This underscores the value of combining a manageable number of individually promising biomarkers into a panel, which can achieve more robustness in addition to a high sensitivity and specificity. From this comparison we conclude that the in house biomarker panel is highly suitable for developing into a non-invasive test for early detection of colorectal cancer.