An immune evasion molecular subgroup predicts prognosis in lung adenocarcinoma

Abstract

Abstract The formation and propagation of lung cancer are closely linked to immunoevasion (IEV). However, few studies have examined IEV-related genes to predict prognoses. Thus, Using the Lasso Cox regression models, we created the IEV-related gene signature (IEVSig) comprising six prognostic IEV-related genes (AHSA1, TNFRSF1A, FADD, CEP55, VDAC2, EMC6) based on The Cancer Genome Atlas databases and then validated by the Gene Expression Omnibus database. According to our findings, IEVSig is an independent prognostic factor. Patients with a higher IEVSig have shorter overall survival. Further, Assessments were conducted using multivariate Cox regression, nomogram, and Kaplan-Meier analysis. The areas under the ROC curve of GEO and TCGA databases at 1, 3, and 5 years are 0.63, 0.60, and 0.55, and 0.67, 0.65, and 0.66, respectively. In addition, the correlations between the IEVSig and the immune score were analyzed with Spearman. Furthermore, Gene ontology (GO) analyses indicated that IEVSig was enriched in death-inducing signaling. moreover, we evaluated the genomic alteration and somatic mutation status between patients with high and low IVESig in the TCGA-LUAD cohort. We noted distinct somatic mutation profiles among these subtypes Despite TP53, TTN, and MUC16 being the most frequent mutation, the relative frequencies varied among different subtypes. As a result, LUAD patients with an IEV-related prognostic signature were identified and the prediction marker can be used to offer insight into therapeutic approaches.