A novel aging-associated lncRNA prognostic signature for predicting immunotherapy and chemotherapy response in pancreatic cancer

Abstract

Background: Tumors commonly develop as individuals age, and their relationship is intricately connected yet remarkably unclear). On one hand, long non-coding RNAs (lncRNAs) affect the proliferation, invasion, metastasis, prognosis, and drug resistance of most tumors, including pancreatic cancer. On the other hand, increasing evidence suggests that lncRNAs are also widely involved in the aging process. Nevertheless, it remains unknown whether aging-associated lncRNAs impact the prognosis and immune microenvironment of pancreatic cancer. Methods: Transcriptomic and clinical information of pancreatic cancer were obtained from TCGA databases. Aging-associated lncRNAs were enriched by co-expression analysis based on 500 aging-related genes in Aging Atlas database. Then aging-associated lncRNAs signature risk model was constructed by univariate, multivariate and Lasso Cox regression analysis. The receiver operating characteristic (ROC) curve, Kaplan-Meier analysis and nomogram were performed to evaluate the prognosis value and predictive ability. Patients in the PACA_PU databases were enrolled to verify the risk model as an external validation. ESTIMATE, single-sample gene set enrichment analysis (ssGSEA) and TIDE were used to analyze the immune microenvironment and immunotherapy response. And drug sensitivities were analyzed with the pRRophetic algorithm. Results: A risk signature containing six aging-associated lncRNAs (RP11-731F5.2, PVT1, RP1-239B22.5, LINC01004, RP11-61J19.5 and RP11-132A1.4) was established and shown to be significantly associated with overall survival in pancreatic cancer patients. The ROC curve, nomogram and PCA analysis demonstrated the accuracy and stability of the risk model’s ability to predict prognosis as well as its good clinical utility. The ESTIMATE and ssGSEA results suggested a strong relationship between the aging-associated lncRNAs risk signature and immune infiltration of pancreatic cancer. The analysis of PD1, TIDE, TMB and MSI further demonstrated the effectiveness of this risk signature for predicting immunotherapy response. In addition, this risk signature was also useful for predicting the sensitivity of chemotherapeutic agents. Conclusion: We constructed a novel aging-associated lncRNAs-derived risk signature with good prognostic value in pancreatic cancer that effectively predicted immunotherapy response and chemotherapeutic drug sensitivity.