Despite advances in systemic therapy, brain metastases remain a significant cause of mortality in non-small cell lung cancer (NSCLC) patients. Nearly 50% of patients with NSCLC will develop brain metastases during the course of their disease. Although many of the biological pathways and processes associated with solid tumor metastasis have been defined, these advances have still not led to robust biomarkers for predicting metastatic behavior in primary NSCLC. We hypothesize that subpopulations of primary NSCLC tumor cells evolve through a multistep process of genomic and epigenomic alterations that result in a metastatic cell phenotype. While the end functional consequence may be the same, the exact alterations that occur between patients and perhaps even between tumor cells in the same patient are likely to be highly varied, thus explaining why single gene biomarkers or correlative gene expression signatures are not sufficiently predictive in all cases. In this proposal, we plan to leverage 'next generation'sequencing (NGS) technologies for simultaneous exome and RNA sequencing to perform a comprehensive yet focused comparative analysis of the genomes of patient-matched primary NSCLC and brain metastatic tumor cell populations. A judiciously selected population of approximately 75 patients with adenocarcinoma will be used for a discovery set and will include paired primary tumor and brain metastasis cases as well as paired primary tumor and late stage lymph node metastasis cases for comparative purposes. We will identify both genomic and gene expression alterations that are enriched in metastatic tumor cell populations and recurrent across patients. Rather than focusing on individual genes, we will identify both synergistic and complementary alterations in key genomic regulators of pathways known to contribute to the metastatic phenotype. For those genomic and gene expression alterations identified in this first aim, we will construct a risk-based model that will be validated using targeted, 'deep'sequencing in an independent, retrospective cohort of 300 primary NSCLC adenocarcinomas from patients with and without the eventual development of brain metastases. This proposal will utilize both a relatively large and refined cohort of patients to study a specific and clinically relevant phenotype in NSCLC (brain metastasis) and will apply an innovative, biologically based risk model approach to identify a more robust set of biomarkers to predict brain metastatic behavior in NSCLC patients. The availability of such a predictor will allow for more personalized therapies for NSCLC patients and ultimately reduce the morbidity and mortality associated with treatment and disease progression.
Although brain metastasis is a significant cause of mortality in patients with non-small cell lung cancer (NSCLC), there are few biomarkers that accurately predict metastatic behavior. We will utilize exome and RNA sequencing technology, a carefully identified cohort of NSCLC patients, and an innovative analytical approach to identify genomic and transcriptome alterations that are both enriched in metastatic cell populations and highly recurrent in patients with brain metastases. The alterations identified will be used to build a ris-based model for brain metastasis using an independent validation cohort, ultimately defining a multi-marker genomic assay for more accurate prediction of metastatic behavior and improved therapeutic management of NSCLC patients.