Robust Validation and Comprehensive Analysis of a Novel Signature Derived from Crucial Metabolic Pathways of Pancreatic Ductal Adenocarcinoma

SIMPLE SUMMARY: Pancreatic ductal adenocarcinoma (PDAC) has a high mortality rate and a poor prognosis. To solve the above limitations of multiomics studies of metabolism in PDAC and optimize the prognosis of PDAC clinically, we demonstrated a 16-gene prognostic signature based on the metabolic pathways called gbcxMRS. The prognostic value varied in six public datasets and our own data cohort in Shanghai Cancer Center by RT-PCR. Notably, gbcxMRS also accurately predicted poor PDAC subtypes and recurrence. It also highly associated with immune infiltration cells. Furthermore, high gbcxMRS may indicate high sensitivity to irinotecan, docetaxel, and CTLA4 inhibitor immunotherapy. ABSTRACT: Pancreatic ductal adenocarcinoma (PDAC) is a malignant tumor with a dismal prognosis. PDAC have extensively reprogrammed metabolic characteristics influenced by interactions with normal cells, the effects of the tumor microenvironment and oncogene-mediated cell-autonomous pathways. In this study, we found that among all cancer hallmarks, metabolism played an important role in PDAC. Subsequently, a 16-gene prognostic signature was established with genes derived from crucial metabolic pathways, including glycolysis, bile acid metabolism, cholesterol homeostasis and xenobiotic metabolism (gbcx). The signature was used to distinguish overall survival in multiple cohorts from public datasets as well as a validation cohort followed up by us at Shanghai Cancer Center. Notably, the gbcx-related risk score (gbcxMRS) also accurately predicted poor PDAC subtypes, such as pure-basal-like and squamous types. At the same time, it also predicted PDAC recurrence. The gbcxMRS was also associated with immune cells, especially CD8 T cells, Treg cells. Furthermore, a high gbcxMRS may indicate high drug sensitivity to irinotecan and docetaxel and CTLA4 inhibitor immunotherapy. Taken together, these results indicate a robust and reproducible metabolic-related signature based on analysis of the overall pathogenesis of pancreatic cancer, which may have excellent prognostic and therapeutic implications for PDAC.