A prognostic and immunotherapy effectiveness model for pancreatic adenocarcinoma based on cuproptosis-related lncRNAs signature

Abstract

Pancreatic adenocarcinoma (PAAD) results in one of the deadliest solid tumors with discouraging clinical outcomes. Growing evidence suggests that long non-coding RNAs (lncRNAs) play a crucial role in altering the growth, prognosis, migration, and invasion of pancreatic cancer cells. Cuproptosis is a novel type of cell death induced by copper (Cu) and is associated with mitochondrial respiration during the tricarboxylic acid cycle. However, the relationship between lncRNAs related to cuproptosis and PAAD is poorly studied. In this study, we investigated the association between a signature of cuproptosis-related lncRNAs and the diagnosis of PAAD. Genomic data and clinical information were obtained using the TCGA dataset, while cuproptosis-related genes (CRGs) from previous studies. Co-expression analysis was utilized to identify lncRNAs associated with cuproptosis. We developed and verified a prognostic risk model following a classification of patients into high- and low-risk categories. The prediction capacity of the risk model was assessed using a number of methods including Kaplan–Meier analysis, receiver operating characteristic (ROC) curves, nomograms, and principal component analysis (PCA). Furthermore, differentially expressed genes (DEGs) were used to perform functional enrichment analyses, and to examine the behaviors of various risk groups in terms of immune-related activities and medication sensitivity. We identified 7 cuproptosis-related lncRNA signatures, including CASC19, FAM83A-AS1, AC074099.1, AC007292.2, AC026462.3, AL358944.1, and AC009019.1, as overall survival (OS) predictors. OS and progression-free survival (PFS) showed significant differences among patients in different risk groups. Independent prognostic analysis revealed that the cuproptosis-related lncRNA signatures can independently achieve patient prognosis. The risk model demonstrated strong predictive ability for patient outcomes, as evidenced by ROC curves, nomograms, and PCA. Higher tumor mutation burden (TMB) and lower tumor immune dysfunction and exclusion (TIDE) scores were observed in the high-risk group. Additionally, the low-risk group was hypersensitive to 3 anti-cancer medications, whereas the high-risk group was hypersensitive to one. A prognostic risk model with a good predictive ability based on cuproptosis-related lncRNAs was developed, providing a theoretical basis for personalized treatment and immunotherapeutic responses in pancreatic cancer.