OP11: ESTABLISHING A DIAGNOSTIC PIPELINE FOR METHYLOME ANALYSIS OF PAEDIATRIC AND ADULT BRAIN TUMOURS IN THE UK USING THE HEIDELBERG CLASSIFIER

INTRODUCTION: Characteristic patterns of the DNA methylation are key features in various tumours. A promising tool for DNA methylation studies is the Illumina Infinium 450k array that assesses the DNA methylation status of 450,000 individual CpG sites. An algorithm, developed in Heidelberg, is based on more than 2000 reference brain tumours and is able to distinguish brain tumour entities based on their methylation profile. The aim is to improve the diagnostic accuracy, important for paediatric tumours, and useful for entities with unusual morphology or clinical course. METHOD: We have set up a diagnostic pipeline and have performed a trial run on a set of morphologically well-characterized paediatric and adult brain tumours and of a set of complex, unresolved tumour cases. Unprocessed idat-files of Illumina Infinium Human Methylation 450 BeadChips are uploaded to the Heidelberg v9 classifier. This tool provides a molecular classification, based on the comparison to the reference set as well as a low resolution copy number plot from the array data for assessment of deletions, including 1p19q codeletion and amplifications, and provides reliable information on the MGMT promoter methylation status. RESULTS: We present the feasibility of setting up a 450k array platform in a diagnostic setting including validation steps and interpretation of the classifier results of rare entities and approaches to refining or revising histological diagnosis. CONCLUSION: The molecular classification by Heidelberg classifier is as yet a research tool, but it does show a great potential for improving diagnostic accuracy, risk stratification and a step closer to improved patient-tailored therapy