Abstract. Gliomas represent approximately 30% of central nervous system tumors and 80% of all malignant brain tumors. Clinically, gliomas are classified and graded according to histological criteria (astrocytoma, glioblastoma, oligoastrocytoma and oligodendroglioma; grade I to IV). Glioblastoma, a grade IV astrocytoma, is generally fatal within eighteen months of diagnosis, where survivals associated with grade II and III neoplasms range from three to fifteen years. Although histopathologic classification is well established and is the basis of the WHO classification of CNS tumors, it suffers from high intra- and inter-observer variability, particularly amongst grade II-III tumors. Therefore classification based on molecular criteria presents a more coherent portrait than histological classes.
An unbiased analysis across glioma grades and histologies that integrates all the possible molecular and genetic information has never been attempted. Therefore, the characterization of the molecular features that mark each of the specific Low Grade Glioma (LGG) and Glioblastoma (GBM) subgroup remains elusive. Most importantly, current analyses have not yet clarified the relationships between LGGs and highly malignant GBMs that share common genetic hallmarks such as IDH mutation or TERT promoter mutation status. Understanding these relationships is of critical importance in clinical management of gliomas and will be necessary to evolve to an objective genome-based clinical classification of these tumors. To address the above questions, we assembled a dataset comprising all TCGA newly diagnosed glioma consisting of 1,122 patients (516 LGG and 606 GBM), which have been analyzed using multiple molecular platforms including mRNA sequencing and expression arrays, DNA methylation arrays, exome sequencing, DNA copy number profiling arrays and targeted proteomics using reverse phase protein arrays. We address crucial technical challenges in analyzing this comprehensive dataset, including the integration of multiple available platforms and sources of data (e.g. multiple methylation and gene expression platforms) and have extended our analysis to pediatric gliomas and pilocytic astrocytoma to span the broader spectrum of glioma. We identified new glioma subgroups with distinct molecular and clinical features and may shed light on the mechanisms driving progression of LGG (WHO grades II and III) into full-blown GBM (WHO grade IV).