The goal of this observational study is to explore the spatial multi-omics mechanism of cachexia in pancreatic cancer and construct a high-precision multimodal intelligent early warning model for its early diagnosis in pathologically confirmed pancreatic ductal adenocarcinoma patients aged ≥18 years with an expected survival of ≥1 month.
The main questions it aims to answer are: 1) Can spatial transcriptomics/metabolomics reveal the interaction mechanism between tumor microenvironment and metabolism in pancreatic cancer cachexia? 2)Can a multimodal AI early warning model be built to achieve accurate early identification of pancreatic cancer cachexia? Researchers will compare the surgical cohort and advanced cohort to see the differences in cachexia incidence and progression rules among different subgroups. Participants will complete baseline questionnaires on physical and mental state, nutrition and muscle strength, undergo imaging examinations and provide biological samples (blood, feces, tissue etc.), and receive regular follow-ups for updated questionnaires, imaging rechecks and supplementary biological sample collection until death, loss to follow-up or voluntary withdrawal.