Abstract 1445: Integrative genomics identifies actionable targets for therapy in medulloblastoma subgroups

The application of high-resolution genomics to the study of medulloblastoma has recently led to a significant enhancement in our understanding of its pathogenesis, implicating new genes and gene families, previously uncharacterized molecular processes, and the existence of distinct biological subgroups: WNT, SHH, Group 3, and Group 4. Despite these advances, few genomic studies have profiled sufficient cases to identify recurrent genetic events, including those restricted to a particular molecular subgroup, making it difficult to discriminate driver genes from passengers. To specifically address these issues we have performed a comprehensive genome-wide copy number analysis of 1,250 frozen primary medulloblastomas and summarized their genomes by medulloblastoma subgroup using Affymetrix SNP6 arrays and a custom nanoString expression assay. Significant regions of interest were verified and validated both within the MAGIC cohort and in non-overlapping samples using a combination of FISH and a custom nanoString copy number assay. The most prevalent oncogenic events observed in medulloblastoma included those targeting known oncogenes and tumor suppressors such as members of the MYC family (MYCN, MYC, and MYCL1), cell cycle regulators (CCND2 and CDK6), genes involved in RTK/PI3K/mTOR signaling (PDGFRA, IRS2, PTEN, and TSC1), and components of the SHH pathway (GLI2 and PTCH1). Integration of our copy number data with nanoString results allowed for subgroup-specific genomic analyses and the identification of multiple novel candidates that appear to be targeted in a subgroup-restricted manner. Significant focal copy number aberrations affecting the chemokine receptor CXCR4 and candidate breast cancer oncogenes LMO4 and BCAS3 were revealed in SHH medulloblastoma, providing insight into novel pathways that likely cooperate with canonical SHH signaling in this subgroup. Similarly, we identified novel genetic events restricted to the poor prognosis Group 3 and Group 4 medulloblastomas, including recurrent high-level amplification of members of the TGFβ pathway (ACVR2A and ACVR2B) in Group 3 and apparent deregulation of the RB pathway (RB1, CCND2, and CDK6) in Group 4. Definitive elucidation of the genetic events contributing to the initiation, maintenance, and progression of medulloblastoma will be an essential prerequisite for the future development of rationally designed targeted therapies. Our current study of >1,200 medulloblastoma genomes has shown that medullobl