Aerial View of the Association Between m6A-Related LncRNAs and Clinicopathological Characteristics of Pancreatic Cancer

Pancreatic cancer is a highly malignant tumor with a poor survival prognosis. We attempted to establish a robust prognostic model to elucidate the clinicopathological association between lncRNA, which may lead to poor prognosis by influencing m6A modification, and pancreatic cancer. We investigated the lncRNAs expression level and the prognostic value in 440 PDAC patients and 171 normal tissues from GTEx, TCGA, and ICGC databases. The bioinformatic analysis and statistical analysis were used to illustrate the relationship. We implemented Pearson correlation analysis to explore the m6A-related lncRNAs, univariate Cox regression and Kaplan-Meier methods were performed to identify the seven prognostic lncRNAs signatures. We inputted them in the LASSO Cox regression to establish a prognostic model in the TCGA database, verified in the ICGC database. The AUC of the ROC curve of the training set is 0.887, while the validation set is 0.711. Each patient has calculated a risk score and divided it into low-risk and high-risk subgroups by the median value. Moreover, the model showed a robust prognostic ability in the stratification analysis of different risk subgroups, pathological grades, and recurrence events. We established a ceRNA network between lncRNAs and m6A regulators. Enrichment analysis indicated that malignancy-associated biological function and signaling pathways were enriched in the high-risk subgroup and m6A-related lncRNAs target mRNA. We have even identified small molecule drugs, such as Thapsigargin, Mepacrine, and Ellipticine, that may affect pancreatic cancer progression. We found that seven lncRNAs were highly expressed in tumor patients in the GTEx-TCGA database, and LncRNA CASC19/UCA1/LINC01094/LINC02323 were confirmed in both pancreatic cell lines and FISH relative quantity. We provided a comprehensive aerial view between m6A-related lncRNAs and pancreatic cancer’s clinicopathological characteristics, and performed experiments to verify the robustness of the prognostic model.