Eleven immune-gene pairs signature associated with TP53 predicting the overall survival of gastric cancer: a retrospective analysis of large sample and multicenter from public database

Abstract Background Growing attention have been paid to the relationship between TP53 and tumor immunophenotype, but there are still lacking enough search on the field of gastric cancer (GC). Materials and methods We identified differential expressed immune-related genes (DEIRGs) between the TP53-altered GC samples (n = 183) and without TP53-altered GC samples (n = 192) in The Cancer Genome Atlas and paired them. In the TCGA cohort (n = 350), a risk score was determined through univariate and multivariate cox regression and Lasso regression analysis. Patients were divided into two groups, high-risk and low-risk, based on the median risk score. Four independent cohorts (GSE84437,n = 431; GSE62254, n = 300; GSE15459, n = 191; GSE26901, n = 100) from the Gene Expression Omnibus (GEO) database were used to validate the reliability and universal applicability of the model. Results The signature contained 11 gene pairs showed good performance in predicting progression-free survival (PFS), disease-free survival (DFS), disease special survival (DSS), and the overall survival (OS) for GC patients in the TCGA cohort. The subgroup analysis showed that the signature was suitable for GC patients with different characteristics. The signature could capable of distinguish GC patients with good prognosis and poor prognosis in all four independent external validation cohorts. The high- and low-risk groups differed significantly in the proportion of several immune cell infiltration, especially for the T cells memory resting, T cells memory activated and follicular helper, and Macrophage M0, which was also related to the prognosis of GC patients. Conclusion The present work proposed an innovative system for evaluating the prognosis of gastric cancer. Considering its stability and general applicability, which may become a widely used tool in clinical practice.

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.