Identification Of Prognostic Genes Associated With The Tumor Ecosystem In Pancreatic Ductal Adenocarcinoma

Background The tumor microenvironment is crucial for pancreatic ductal adenocarcinoma (PDAC) progression, yet its influence remains incompletely understood. This study aimed to identify tumor ecosystem-related prognostic genes in PDAC. Methods We analyzed public PDAC datasets. Differentially expressed genes (DEGs) were identified and overlapped with tumor ecosystem-related genes (TERGs). Prognostic genes were selected via regression analyses to construct a risk model, stratifying patients into high-risk (HRG) and low-risk (LRG) groups. The model was evaluated and validated. Independent prognostic, immune infiltration, drug sensitivity, molecular docking, and regulatory network analyses were performed. Results Eight prognostic genes were identified: KIF23, RACGAP1, CDK1, MET, TNFSF10, LAMA3, CXCL10, and IGFBP2. The risk model showed robust predictive performance, with HRG patients having poorer survival. Age and risk score were independent prognostic factors. Immune analysis revealed a significant negative correlation between monocytes and CDK1, and a positive correlation between MDSCs and CXCL10. Drug sensitivity analysis (e.g., axitinib) showed significant IC50 differences between HRG and LRG. Molecular docking confirmed positive binding of prognostic genes to key drugs. Notably, GATA2 targeted multiple genes including CDK1, IGFBP2, and MET. Conclusion Identified eight tumor ecosystem-related prognostic genes in PDAC. The study provides a theoretical foundation for developing PDAC treatments.