Summary – Project 1 Project 1 will study the ∆-Ecology of tumor-immune interactions in NSCLC, with a focus on KRAS-mutant cancers. Immunotherapy has demonstrated response rates of 30 to 45%. Most responses are followed by evolution of resistance and progression. Recently we have quantified the immune ecology of these cancers from pre-treatment biopsies and found unique ecological interactions that predetermine therapy response. A focus of this proposal will be to use both pre- and on-treatment biopsies from a clinical trial to quantify the immune ∆-Ecology during therapy. We will use a computational multiplexed-image analysis pipeline that takes whole-slide images and segments the data into cells and quadrats to analyze the spatial and spatiotemporal features representing ecological changes that drive tumor progression. We will then harness this analysis as a spatiotemporal biomarker to alter treatment strategies via mathematical modeling. The computational infrastructure will be grounded in mechanistic models, spatial statistics, machine learning, and deep learning. Targeted therapies are also a key component in the treatment of this disease. In an exciting discovery, inhibitors that target KRAS-G12C mutant cancers have been developed and have demonstrated clinical efficacy. Furthermore, preclinical work has demonstrated that these inhibitors alter the immune ecology of tumors, offering promise that combination targeted therapy and immunotherapy can significantly improve responses. We will perform in vivo work with murine tumors and deep ecological analysis of human tumors to determine the mechanisms of immune ∆-Ecology fomented by KRAS inhibitors, and probe combination therapies that synergize the anti-tumor response. Our ultimate goal is the development of a predictive model of ∆-Ecology that delivers novel treatment strategies and improves outcomes, for use in future clinical trials.