Spatiotemporal metabolic networks in a preclinical model of pancreatic cancer and associated cachexia.

Cachexia or body wasting is a significant contributing factor to the morbidity and mortality of pancreatic cancer. Despite an unmet need, its metabolic underpinnings remain poorly understood. Here, we characterize a mouse model that recaptures the longitudinal progression of human cachexia during pancreatic cancer development. We profile the metabolomes of pre-, early- and late-stage cancer-associated cachexia in male and female pancreas, its interstitial fluid (IF), plasma, liver, adipose tissue and skeletal muscle. We find that each tissue has a unique metabolome and trajectory across stages and reveal a marked lipid enrichment in the tumor IF. Using mathematical modeling, we identify metabolites participating in cross-tissue networks and in vivo validate a lactate-glucose conversion. We find systemic metabolic changes prior to weight loss and use feature selection algorithms to identify early detection markers. Our study provides a resource for system-wide metabolic evolution of pancreatic cancer and associated cachexia, highlighting the potential for timely diagnostics and interventions.