Summary Accurate pathological diagnosis is crucial for optimal management of cancer patients. For the ~100 known central nervous system (CNS) tumour entities, standardization of the diagnostic process has been shown to be particularly challenging - with substantial inter-observer variability in the histopathological diagnosis of many tumour types. We herein present the development of a comprehensive approach for DNA methylation-based CNS tumour classification across all entities and age groups, and demonstrate its application in a routine diagnostic setting. We show that availability of this method may have substantial impact on diagnostic precision compared with standard methods, resulting in a change of diagnosis in up to 12% of prospective cases. For broader accessibility we have designed a free online classifier tool (www.molecularneuropathology.org) requiring no additional onsite data processing. Our results provide a blueprint for the generation of machine learning-based tumour classifiers across other cancer entities, with the potential to fundamentally transform tumour pathology.

Link to publication record in Explore Bristol Research PDF-document This is the author accepted manuscript (AAM). The final published version (version of record) is available online via Lancet at https://www.sciencedirect.com/science/article/pii/S1470204517301559?via%3Dihub . Please refer to any applicable terms of use of the publisher. University of Bristol -Explore Bristol Research General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Full terms of use are available:

ATRX and IDH1-R132H immunohistochemistry with subsequent copy number analysis and IDH sequencing as a basis for an "integrated" diagnostic approach for adult astrocytoma, oligodendroglioma and glioblastoma Following the concepts of the "ISN-Haarlem", we rediagnosed the series to obtain "integrated" diagnoses with 155 tumors being astrocytomas, 100 oligodendrogliomas and 150 glioblastomas. In a subset of 100 diffuse gliomas from the NOA-04 trial with long-term follow-up data available, the "integrated" diagnosis had a significantly greater prognostic power for overall and progression-free survival compared to WHO 2007. Based on the "integrated" diagnoses, loss of ATRX expression was close to being mutually exclusive to 1p/19q codeletion, with only 2 of 167 ATRX-negative tumors exhibiting 1p/19q codeletion. All but 4 of 141 patients with loss of ATRX expression and diffuse glioma carried either IDH1 or IDH2 mutations. Interestingly, the majority of glioblastoma patients with loss of ATRX expression but no IDH mutations exhibited an H3F3A mutation. Further, all patients with 1p/19 codeletion carried a mutation in IDH1 or IDH2. We present an algorithm based on stepwise analysis with initial immunohistochemistry for ATRX and IDH1-R132H followed by 1p/19q analysis followed by IDH sequencing which reduces the number of molecular analyses and which has a far better association with patient outcome than WHO 2007. or IDH2 mutations. Interestingly, the majority of glioblastoma patients with loss of ATRX expression but no IDH mutations exhibited an H3F3A mutation. Further, all patients with 1p/19 co-deletion carried a mutation in IDH1 or IDH2. We present an algorithm based on stepwise analysis with initial immunohistochemistry for ATRX and IDH1R132H followed by 1p/19q analysis followed by IDH sequencing which reduces the number of molecular analyses and which has a far better association with patient outcome than WHO 2007.