PROJECT SUMMARY/ABSTRACT Metabolic reprogramming is a hallmark of malignancy and potential source of therapeutic targets. Recent work indicates that metabolic liabilities change as cancer progresses, meaning that the pathways most relevant to advanced cancers may not be apparent in locally-invasive, treatment-naïve tumors at the site of origin. Recognizing the dearth of direct information about human cancer metabolism, we developed an approach to probe the metabolic network of intact human tumors by infusing patients with stable isotope-labeled nutrients (e.g. 13C-glucose) during tumor resection or biopsy. By measuring isotope labeling in metabolites extracted from tumor samples and following the outcomes of patients who underwent this procedure, we identified metabolic properties associated with poor survival. Of hundreds of metabolic features, 13C labeling in tricarboxylic acid (TCA) cycle metabolites was the most predictive of cancer progression and early death. In non-small cell lung cancer (NSCLC), patients whose tumors have high labeling of these metabolites succumb much earlier than patients with low labeling, and blocking this pathway in mouse models of NSCLC suppresses metastasis. In clear cell renal cell carcinoma (ccRCC), TCA cycle labeling is low when tumors are localized to the kidney but much higher in metastatic tumors, and activating the TCA cycle promotes metastasis in mice. Therefore, in both kinds of cancer, data from patients lead us to conclude that oxidative mitochondrial metabolism, particularly the TCA cycle, electron transport chain (ETC) and oxidative phosphorylation (OxPhos), promote cancer progression. The success of these experiments prompts us to further study the metabolic basis of human cancer progression in the hopes of developing new insights and therapies. We propose three general directions. First, using a combination of approaches in humans and mice, we will thoroughly examine how mitochondrial metabolism stimulates metastasis to identify discrete metabolic dependencies that could be safely targeted in patients. Second, we will develop approaches to discover new metabolic liabilities in human tumors. Strategies include a pipeline to probe viable tumor explants with a series of isotope-labeled nutrients under physiological conditions to choose the most informative tracers for isotope infusions in patients; and dynamic imaging methods to observe and quantify informative aspects of metabolic flux in tumors in real time. Third, we will use the orthogonal approach of studying human inborn errors of metabolism (IEMs) to discover why some metabolic anomalies prime cells to become malignant. This approach capitalizes on a clinical cohort of over 1,000 subjects, including patients with IEMs associated with highly penetrant cancers, and will provide unique insights into cancer initiation and progression. Altogether these efforts will build on our long-standing productivity in human cancer metabolism by uncovering new me...