BI-14 * GENOMIC PROFILING OF A PREDICTIVE SIGNATURE FOR MET-TARGETED THERAPY IN GLIOBLASTOMA

Abstract

The success of molecular targeted therapy against cancer depends on discovering the tumor driver genes and the molecular determinants that control the pathway activity. Glioblastoma (GBM) is one of the most devastating cancers due to its highly infiltrating nature, and MET pathway activation is a major cause of invasion in both primary and recurrent tumors. Because MET inhibitors are in clinical trials against GBM, there may be clinical utility from developing more effective patient enrollment strategies tailored to targeted therapeutics. Previously, we reported (Xie et al., PNAS 2012) that GBM tumors with high levels of hepatocyte growth factor (HGF) often display HGF-autocrine activation through its receptor MET, which is a key molecular feature in sensitivity to MET inhibitors. In this study, we sought to develop a molecular signature that can be used as a biomarker to identify GBM patients whose tumor would be vulnerable to treatment with MET inhibitors. Because GBM is a heterogeneous disease in which drug response in the individual patient can be influenced by a variety of different mechanisms, the expression of a single gene was not anticipated to be sufficient to pinpoint sensitivity to the drug; rather, a hypothesis-driven, biomarker-based molecular signature would likely be of a higher value. We analyzed genomic data from GBM patients in The Cancer Genome Atlas (TCGA) Network as well as from preclinical tumor models. We found that GBM tumors sensitive to MET inhibitors share common genomic profiles. More importantly, using patient-derived xenograft models, a 25-gene molecular signature was identified that predicted sensitivity to MET inhibitors. Our findings are a proof-of-concept for the use of genomic signatures to identify GBM patients with greater vulnerability for MET-targeted therapy.