Abstract
The heterogeneity of pancreatic adenocarcinoma (PAAD) tumors is complex. The purpose of this work was to use immunologic signature gene sets (ISGS) activity alterations to identify subtypes of PAAD and to develop valid prognostic indicators. We discovered tumor-related differentially expressed gene sets and their associated subtypes by analyzing changes in ISGS activity via the cancer genome Atlas and the genotype-tissue expression datasets. To evaluate the relationship between these subtypes and clinical characteristics, the immunological microenvironment, and tumor immune dysfunction and exclusion, we conducted survival studies. Furthermore, we identified subtype-specific gene sets and developed a prognostic risk score (RS) model and a corresponding nomogram. The robustness and generalizability of the RS model were validated using the gene expression omnibus dataset. Additionally, differences in drug sensitivity and tumor mutation burden were compared between risk groups. On the basis of variations in ISGS activity, 2 subtypes of PAAD were distinguished. The prognosis is worse for subtype 1 patients, and there is no statistically significant difference between the clinicopathologic characteristics of the 2 subtypes. However, subtype 1 patients do not respond well to immunotherapy, and the immunological microenvironment varies significantly across subtypes. Additionally, patients were assigned to distinct risk groups by a prognostic RS model based on differentially expressed genes linked to both subtypes. Individuals belonging to distinct risk categories exhibited varying degrees of medication sensitivity; high-risk patients also had shorter survival periods and greater tumor mutation burden. For patients with pancreatic cancer, nomograms that include RS and clinicopathologic variables are may provide useful tools. Our research used the ISGS to identify subtypes of PAAD and created prognostic RS models. Through the integration of the RS and clinicopathological parameters into nomograms, prognostic prediction may be reinforced, and may provide new perspectives for future personalized treatment strategies.