Illuminating molecular targetable pathways in HNSCC

? DESCRIPTION (provided by applicant): We propose to better employ existing drugs to define new agents and combinations of agents to treat HNSCC, a disease with unchanged survival rates for four decades in need of new approaches, tools and perspectives. To do so we will combine the advantages of the large dataset in The Cancer Genome Atlas (TCGA) containing genomics, epigenomics and basic outcomes data but including little functional information to support causality in the disease or its treatment efficacy, with a well annotated clinical dataset that uniquely includes functional information on the sensitivity of HNSCC patient tumor cells to a panel of drugs approved or under development for human diseases but not yet applied to HNSCC. Using cutting-edge computational methods, we will mine the TCGA dataset in terms of aberrational gene pathway networks, prioritizing their relevance to HNSCC. We are taking advantage of a high throughput inhibitor assay and computational tools originally showing success in leukemia to design and employ HNSCC-specific inhibitor panels that capture the diversity of aberrational pathways in TCGA to test viable cells from patients' HNSCC tumors to predict effective targeted therapeutic agents and identify the reasons for differential response or
resistance to certain drugs among patients. Our dual-PI expertise encompasses creating and using tools of cross-platform data analysis and precision medicine (McWeeney) and repository and specialized cell culture design, biochemical and molecular analysis and orthotopic xenograft models (Kulesz-Martin). Our preliminary results show convergence of pathways and targets from functional analysis of our OHSU dataset with those highly significant from TCGA data, in which 73 pathways are represented on the initial ~120 drug panel, and another 121 pathways, which we call dark, are not represented. The dark pathways offer a source of innovative targets for creation of a HNSCC-specific inhibitor panel. Our preliminary data show 15 drugs that reach thresholds of efficacy in 7 cases tested. Our preliminary analyses of TCGA with our HNSCC patient datasets annotated by RNASeq and functionally by inhibitor assay identified differential responses to EGFR inhibitors, PI3K inhibitors and other targeted agents. Cases where single agents were ineffective were sensitive to drug combinations with EGFR inhibitors. We propose to pursue more effective therapies for HNSCC as follows: 1) Perform complementary analysis of TCGA data with -omics and functional data on OHSU HNSCC patients' cells to develop HNSCC-specific inhibitor panels and more effectively apply available drugs to HNSCC treatment, and to illuminate dark pathways as a source of new targets; 2) Prioritize targets computationally, taking advantage of state-of-the-art bioinformatics tools for cross-platform analysis of TCGA and OHSU data, and validate these in vitro as responsible for patient tumor cell-specific drug (in)sensitivities; and 3) Test predictions in situ in original patent tumors and in vivo using orthotopic xenograft models, relating results to clinical outcomes to define subsets of HNSCC for future molecular predictive tests and tailoring treatment to patient tumor characteristics.