Migrasome-related LncRNA features predict immune microenvironment and prognosis in pancreatic cancer

TL;DR

This study identifies specific long non-coding RNAs linked to migrasomes in pancreatic cancer, which can predict patient survival and immune response, offering new insights for treatment.

Contribution

The study introduces a novel risk model based on migrasome-related lncRNAs for predicting prognosis and immune features in pancreatic cancer.

Findings

A risk model using four migrasome-related lncRNAs accurately predicts pancreatic cancer prognosis with high AUC values.

The model correlates with immune infiltration, tumor mutation burden, and drug sensitivity in pancreatic cancer patients.

High-risk patients show lower survival rates, and the model effectively distinguishes clinical outcomes.

Abstract

The onset of pancreatic cancer is insidious, and the early symptoms are similar to those of common gastrointestinal diseases, which leads to easy neglect and misdiagnosis, which greatly affects the accuracy of survival prediction. Cell migration is the hallmark of malignant tumor and the key step of metastasis. Migrasome are involved in embryonic development, immune response, angiogenesis, inflammatory response, wound healing, and cancer metastasis in vivo. Considering the unknown association between migrasome and lncRNAs in pancreatic cancer, the purpose of this study was to identify migrasome-related lncRNAs (MRLs) and explore their prognostic value. In this study, we first analyzed the Pancreatic adenocarcinoma (PAAD) data in The Cancer Genome Atlas(TCGA) database and identified the correlation between MRLs and pancreatic cancer prognosis and immune infiltrating landscape. Secondly, four MRLs (MED14OS, AC141930.2, Z97832.2, LINC01091) were selected to construct a risk model as a prognostic feature. Kaplan-Meier survival analysis, Cox regression analysis, Nomogram and Time - dependent Receiver Operating Characteristic (ROC) Curve were then used to verify the accuracy of the model. And then, the Prognostic Risk Model were used in clinical to validate the accuracy. Finally, the correlation of immune score, tumor immune cell infiltration, tumor mutation load, tumor immune escape, and drug sensitivity of the risk model was systematically analyzed. The risk-prognosis model of MRLs was constructed. Survival analysis showed that the survival rate of high-risk subtypes was lower than that of low-risk subtypes. MRL features were an independent prognostic predictor, and the area under the subject working curve (AUC) for 1-year, 3-year, and 5-year were 0.667, 0.780, and 0.865, respectively. Prognosis MRLs is related to immune infiltrating landscape and can reflect the immune status, immune response, tumor mutation burden and drug sensitivity of pancreatic cancer patients. At the same time, this model can distinguish clinical patients well. The results of this study construct a predictive model of pancreatic cancer associated with migrasome, and clarify the relevance of this model to immunotherapy and so on. It provides a new idea for improving immunotherapy and drug therapy.

Figures