Establishment of a prognostic model for pancreatic cancer based on mitochondrial metabolism related genes

Abstract

Abstract Aim Pancreatic ductal adenocarcinoma (PAAD) is recognized as an exceptionally aggressive cancer that both highly lethal and unfavorable prognosis. The mitochondrial metabolism pathway is intimately involved in oncogenesis and tumor progression, however, much remains unknown in this area. In this study, the bioinformatic tools have been used to construct a prognostic model with mitochondrial metabolism-related genes (MMRGs) to evaluate the survival, immune status, mutation profile, and drug sensitivity of PAAD patients. Method Univariate Cox regression and LASSO regression were used to screen the differentially expressed genes (DEGs), and multivariate Cox regression was used to develop the risk model. Kaplan–Meier estimator was employed to identify MMRGs signatures associated with overall survival (OS). ROC curves were utilized to evaluate the model's performance. Maftools, immunedeconv and CIBERSORT R packages were applied to analyze the gene mutation profiles and immune status. The corresponding sensitivity to pharmaceutical agents was assessed using oncoPredict R packages. Results A prognostic model with five MMRGs was developed, which defined the patients as high-risk showed lower survival rates. There was good consistency among individuals categorized as high-risk, showing elevated rates of genetic alterations, particularly in the TP53 and KRAS genes. Furthermore, these patients exhibited increased levels of immunosuppression, characterized by an increased presence of macrophages, neutrophils, Th2 cells, and regulatory T cells. Additionally, high-risk patients showed increased sensitivity to Sabutoclax and Venetoclax. Conclusion By utilizing a gene signature associated with mitochondrial metabolism, a prognostic model has been established which could be a highly efficient method for predicting the outcomes of PAAD patients.

Keywords

• Pancreatic ductal adenocarcinoma
• Mitochondrial metabolism related genes
• Risk model
• Bioinformatic