Establishment of an immune-related LncRNA based prognostic risk assessment model for pancreatic cancer according to TCGA database
- VernacularTitle:基于TCGA数据库构建免疫相关LncRNA的胰腺癌预后风险评估模型
- Author:
Zhenchao GAO
1
;
Yiqun SONG
;
Xinlong CHEN
;
Ze'en ZHU
;
Zheng WANG
;
Weikun QIAN
Author Information
- Publication Type:Journal Article
- Keywords: pancreatic cancer; prognosis; long non-coding RNA(LncRNA); immune reaction
- From: Journal of Xi'an Jiaotong University(Medical Sciences) 2025;46(4):663-670
- CountryChina
- Language:Chinese
- Abstract: Objective To screen immune-related long non-coding RNAs(LncRNAs)in the TCGA database pancreatic cancer dataset and construct a prognostic risk assessment model with immune-related LncRNAs to explore prognosis-related potential molecular mechanisms.Methods RNA-seq data of 171 pancreatic cancer samples and corresponding clinical information were obtained by The Cancer Genome Atlas(TCGA)database,and two classical immune-related gene datasets(GO0006955/IMMUNE RESPONSE and GO0002376/IMMUNE SYSTERM PROCESS)and gene annotation information were used to identify immune-related LncRNAs.The immune-related LncRNAs associated with pancreatic cancer prognosis were used for univariate and multivariate Cox analyses to establish a model for the assessment of pancreatic cancer prognostic risk based on immune-associated LncRNAs.This risk model was used for survival analysis,clinical correlation analysis,immune cell infiltration analysis,pathway enrichment analysis,and prognostic column line plot modeling.Results We screened 119 immune-related LncRNAs in pancreatic cancer,and five immune-related LncRNAs(AC064836.3,LINC00941,ZNF236-DT,TMEM161B-AS1 and AC068580.2)were identified for the development of pancreatic cancer prognostic risk assessment model.According to the prognostic risk assessment model,pancreatic cancer patients were divided into low-risk group(n=86)and high-risk group(n=85).Compared with the low-risk group,the high-risk group showed a significant negative enrichment trend for immune-related signaling pathways,the 5-year overall survival of pancreatic cancer patients was significantly increased in the low-risk group compared with the high-risk group.The expression of low-risk immune-related LncRNAs(AC064836.3,ZNF236-DT and TMEM161B-AS1)gradually decreased with increasing clinical stage of pancreatic cancer patients.Patient age(P=0.031,risk ratio and 95％CI:1.025/1.002-1.048)and prognostic risk score(P＜0.001,risk ratio and 95％ confidence interval 1.801/1.465-2.215)could be used as independent prognostic risk factors for overall survival in pancreatic cancer.In addition,the prognostic risk assessment model had better predictive efficiency(area under the curve=0.695)compared with the disease predictive ability of common clinical characteristics.Steroid biosynthesis,pentose phosphate pathway,intercellular linkage,cytoskeletal rearrangement and other pathways related to energy metabolism and invasive migration of pancreatic cancer cells were significantly activated in the high-risk group.Meanwhile,pancreatic cancer patients in the high-risk group had lower levels of naive B cells,plasma cells and neutrophils with anti-tumor activity,but their macrophage infiltration levels were significantly higher than those in the low-risk group.Conclusion The prognostic risk assessment model constructed based on five immune-related LncRNAs can effectively predict the survival status,clinical characteristics,molecular pathways,and immune cell infiltration differences of pancreatic cancer patients.Meanwhile,relying on this model,the prognosis of pancreatic cancer patients can be prospectively predicted,which enhances the usefulness of this risk prediction model.
