GE-02 * CLINICAL TARGETED EXOME-BASED SEQUENCING FOR CHILDREN DIAGNOSED WITH BRAIN TUMORS: INITIAL RESULTS OF AN INSTITUTE-WIDE PROJECT ON PERSONALIZED MEDICINE

Abstract

BACKGROUND: PROFILE (DFCI#11-104) is a cancer center-wide project that aims to assess feasibility and utility of identifying genomic and potential targetable alterations in pediatric brain tumor patients. METHODS: Eligible patients/families are offered enrollment. Tumor tissues (FFPE and frozen) are collected, DNA extracted and a CLIA-certified platform is applied. First-generation platform, OncoMap, a multiplex mass spectrometry-based platform, detected mutations in 471 loci from 41 cancer genes. This has been replaced by a second-generation platform, OncoPanel, a multiplexed targeted exome-sequencing platform, which surveys 275 cancer genes to identify mutations and copy number variations, in addition to 90 introns from 30 genes to detect rearrangements. Alterations are tiered based on clinical significance and results are communicated to the treating physician. RESULTS: Between February 2013 and June 2014, 79 patients were enrolled. 25 had insufficient tumor tissue. Of the remaining 54, 33 have been profiled thus far (Oncomap = 4; Oncopanel = 29). Diagnoses include low-grade glioma (11), medulloblastoma (6), ependymoma (6), GBM (3), PNET/pineoblastoma (3), ATRT (1), AA (1), meningioma (1) and pituitary adenoma (1). Estimated mean tumor purity was 76%. We detected 401 mutations, including two GBM patients with high mutation rates (which accounted for 314 of 401 mutations detected). Of these, one had deletion of the mismatch-repair gene MSH2. Excluding these two patients, we identified 87 mutations, of which 20 have evidence for potential clinical utility (including investigational or preclinical evidence). Alterations with clinical trial relevance included gains/duplications of BRAF observed in two glioma patients; four medulloblastoma patients were noted to harbor subtype-related alterations of relevance for trial enrollment. CONCLUSIONS: We show that it is possible to perform comprehensive genomic analyses of known cancer genes in the setting of a research protocol. Significant challenges remain in determining the clinical impact of utilizing personalized genomic data and integrating these profiles into the therapeutic decisions for children with brain tumors.