Radiogenomic Biomarker and Multiomic Data Integration to Predict Radiation Response in Lung Cancer
Radiation therapy remains a cornerstone treatment for patients with locally advanced lung cancer, however knowing which patients will respond and which will not respond is still poorly understood. The goal of this project is to analyze genomic and radiomic data from patients with NSCLC to understand how tumors change during therapy and create models to predict therapeutic response that will assist with clinical decision making.
Research Summary
Technical Abstract
Radiation therapy is the single most utilized anti-cancer agent; nearly 70% of all cancer patients will receive radiation at some point in their cancer journey. Radiation plays a crucial role in almost half of all cancer cures. The sequencing of the human genome, completed nearly 20 years ago, followed by the large scale cancer sequencing effort in The Cancer Genome Atlas (TCGA) have provided an unprecedented understanding of cancers in the primary and metastatic setting. In those same years, medical oncology has undergone three major phase transitions: targeted therapies have changed the way we think many diseases with specific actionable mutations; immunotherapy has revolutionized the treatment of many of those without; and antibody-drug conjugates have increased the specificity of our cytotoxics. Radiation treatment decision making, however, has not seen these same changes from biological influences, instead having relied on advances in medical physics and computer science to drive our advances. While the number of trials has ballooned in radiation oncology of late, spurred on by encouragement, and funding, from pharmaceutical companies interested in the synergy between novel (and profitable) compounds in the form of immune checkpoint inhibitors and antibody-drug-conjugates, with radiation, our understanding of the relative benefits and best choices for individual patients has not seen the same increases. In fact, we have struggled to parse out the differences between these novel combinations and standard chemoradiotherapy in phase II trials, largely because of the combinatorial nature of our trials, and the sheer number of open questions. In this project, we seek to make headway toward personalizing radiation therapy treatment choices. Leveraging our experience in using gene signatures to predict individual patient radiation benefit, together with expertise in radiomics and genomics, we will track the progress and outcomes of patients with non-small cell lung cancer treated with standard of care chemoradiation using high resolution genomic and radiomic measures. The dynamics of these genomics through time will be correlated to the high temporal density radiomics features to allow for translation and generalization to all patients treated with modern technique. Through these complimentary -omic modalities, we aim to leverage our experience in creating signatures of therapeutic response to admit personalized treatment choice in the upfront setting, and opportunities to change course using real- time information gleaned from daily imaging. This pilot project will enable the development of novel multimodal data integration methodologies to interrogate radiation treatment response in a comprehensive manner.