The objective of this project is to develop effective precision medicine for treatment of KRAS-mutant non-small cell lung cancer (NSCLC) by determining responses of molecularly characterized patient-derived xenograft (PDX) models with or without KRAS mutations to targeted therapeutic agents. The priority will be given to the agents in the NCI Investigational New Drugs (NCI-IND agents) portfolio to increase the success of early-phase clinical trials in patients with KRAS-mutant NSCLCs. The central hypothesis of the proposed study is that a large set of well-characterized KRAS-mutant NSCLC PDXs that recapitulate multiple characteristics of the biological context of human KRAS-mutant cancers will provide a preclinical platform for identifying effective therapeutic strategies for KRAS-mutant NSCLC and molecular biomarkers capable of identifying treatment responders. Activating mutations of the KRAS gene are known to be among the major genomic alterations associated with lung adenocarcinoma. Thus far, treatment of KRAS-mutant NSCLC remains an unmet medical need. Moreover, development of active anticancer drugs is challenged by high failure rates caused by a lack of preclinical tumor models highly predictive of therapeutic effects in humans and of biomarkers that identify responders in clinical trials. We will overcome this challenge by determining the therapeutic activities of targeted drugs and identifying potential predictive biomarkers capable of identifying responders to those agents in clinically and molecularly annotated PDX models derived from NSCLC. Our research team has extensive expertise in establishing NSCLC PDXs, molecularly characterizing PDXs and primary tumors, developing novel therapeutic strategies for NSCLC, and translating effective preclinical therapeutic strategies for clinical trials. We have generated 190 NSCLC PDXs and completed genomic characterization of 82 PDXs (22% with oncogenic KRAS mutations) and their respective primary tumors. Together, our team is positioned to make a major contribution to the overall objectives of the NCI's PDXNet to generate data on PDX models to support early clinical trials of precision therapies. To that end, we propose three specific aims: 1) determine molecular subtypes of NSCLC PDXs with KRAS mutations; 2) determine responses of NSCLC PDX models with or without KRAS mutations to treatment with investigational new drugs approved for patient use, with priority given to the NCI-IND agents; and 3) characterize molecular biomarkers and the mechanistic relationship between molecular subtypes and treatment responses. The proposed studies are highly relevant to the following Research Areas of NCI's PDX Development and Trial Centers: 1) mechanism-based drug combinations in treatment of genetically defined tumor subgroups that explore the correlation of genetic abnormalities with tumor response; and 2) mechanisms that contribute to the sensitivity or resistance of PDX models to therapeutic agents. The success of the proposed studies will lead to effective precision therapy for KRAS-mutant lung cancers, for which current therapies are ineffective.