Constructing a Novel Prognostic Signature Based on TGF-β Signaling for Personalized Treatment in Pancreatic Adenocarcinoma

Background. Pancreatic adenocarcinoma (PAAD) shows significantly high mortality. Transforming growth factor-beta (TGF-β) signaling plays an important role in tumorigenesis and development. A prognostic model was conducted using transforming growth factor-beta (TGF-β) signaling for predicting PAAD prognosis and guiding personalized therapies. Methods. Datasets were grouped into test and training sets. Univariate Cox regression analysis and least absolute shrinkage and selection operator (LASSO) were applied and introduced for identifying prognostic genes associated with TGF-β. Risk score of each sample was calculated by the prognostic model. The difference in survival, clinical information, mutations, pathways, and chemotherapy and immunotherapy sensitivities between high-risk and low-risk groups was analyzed. Results. Based on TGF-β signaling, this work built a 7-gene prognostic model showing robustness in sample classification into low-risk and high-risk groups with differential prognoses. Oncogenic pathways like glycolysis, Notch signaling, and hypoxia were noticeably enriched in the group with high risk. Interferon and STAT1 were positively associated with risk score. Importantly, the low-risk group may develop a more favorable response to both chemotherapy and immunotherapy. The current work highlighted the significant function of TGF-β signaling in PAAD development and described the potential cross-links with other oncogenic pathways. Conclusion. Notably, the prognostic signature can act as a predictor of prognosis, but as a biomarker for optimizing personalized therapies in clinical practice.