Abstract 1703: Discovery and validation of 20 novel breast cancer methylation biomarkers: A new workflow for methylation biomarker development

Abstract

Abstract Methylation is a process of regulation of gene expression. Methylation pattern changes are implicated in the pathology of cancer and other disorders but at the same time they can be used as biomarkers for clinical disease management. There are four primary fields of use for biomarkers in clinical disease management: diagnosis, prognosis/prediction, prevention, pharmacoepigenomics. Methylation biomarkers have already been shown to fulfill the requirements of each of the above categories. Therefore, a lot of research in the field is currently focused on discovery and validation of methylation biomarkers for clinical use. The process of biomarker development can be subdivided into five steps workflow: discovery of the biomarker, initial clinical biomarker validation, retrospective biomarker validation, prospective validation and the use of the marker in clinical practice. Following the above workflow, we have combined the Roche/NimbleGen Methylation arrays and Methylation Sensitive High Resolution Melting (MS-HRM) (ref: 1, 2), into one platform for methylation biomarker development and validation. Our approach allowed us, firstly to discover 20 novel loci undergoing aberrant methylation during breast cancer development. Subsequently the biomarkers were subjected to initial clinical validation. This procedure aimed to evaluate to what extent the biomarkers can distinguish 274 breast tumor samples (disease) from 78 breast normal tissue samples (healthy). Our results show that the specificity of almost all of the novel biomarkers except for one was above 80%, with four of these showing hypermethylation only in tumor tissue, which corresponds to 100% specificity. The odds ratio (OR, interpreted as the probability with which our new biomarker can distinguish a tumor sample from a non-malignant sample) for our novel biomarkers was in the extraordinary range from 4(2; 7) to 563(166; 1902), Overall our results clearly illustrate the exceptional accuracy of the chosen approach for methylation biomarker discovery and development. All the biomarkers that successfully completed initial clinical validation are now undergoing retrospective validation (see above) based on almost 1200 archival breast cancer samples with detailed clinical follow up. This procedure will allow investigating the clinical management field of use for each of the novel biomarkers.