Identification And Validation Of An Immune Gene Prognostic Signature Based On TCGA Database In Breast Cancer

Purpose: The diagnosis and treatment of breast cancer has initially entered the era of immunotherapy.In this paper,we aims to construct a prognostic risk model of immune genes based on the TCGA database to predict the prognosis of breast cancer.Analyze the correlation between immune-related prognostic characteristics and clinicopathological or biological characteristics of breast cancer,so as to provide a basis for the prediction of prognosis,diagnosis and treatment of clinical breast cancer.Methods: The gene expression data and complete clinical data of TCGA-BRCA were downloaded directly from The Cancer Genome Atlas(TCGA)online portal.The corresponding expression matrix was extracted by Perl language in the cmd environment,and DEGs related to overall survival(OS)were determined by differential expression gene(DEGS)analysis and univariate Cox regression analysis.Lasso regression and multivariate Cox regression analysis were performed to construct the immune-related prognostic model,and the risk score of each sample was calculated according to the model formula.Patients were classified into high and low risk groups according to the median risk score.And the difference in survival between the two groups was compared.The predictive power of the prognostic model in 5-year survival period was evaluated using time-dependent ROC curve.The factors affecting the survival of breast cancer patients were analyzed by incorporating clinicopathological factors and prognostic scores into Cox regression model.Cox regression analysis was also used to determine independent prognostic factors for BRCA.In addition,we investigated the association between immune-related signature and clinical features and immune cells infiltration.Finally,the prognostic risk signature was validated in an external independent dataset.Results: Transcriptome data of 1222 breast cancer samples and clinical data of 1085 female breast cancer patients were acquired and integrated from the TCGA database.After differential gene analysis,a total of 366 differentially expressed immune genes were screened,among which 193 were up-regulated and 173 were down-regulated.Univariate Cox regression analysis of differentially expressed genes showed that 59 immune genes including IGHE,ULBP2,ADM,SEMA6 D et.al were closely related to the prognosis of breast cancer patients.To explore the regulatory mechanisms of these genes,we constructed a TF-mediated regulatory network here.Lassso Cox regression analysis was performed on 59 immune genes.According to Lambda value,35 genes were filtered.Combined with multivariate Cox analysis,13 immune-related genes(ULBP2,HLA-G,CXCL13,TINGAGL1,NFKBIE,IGHE,SEMA3 F,SEMA6D,CLEC11 A,FGF7,TSLP,FLT3,TRBJ2)were finally screened out to construct an immune-related prognosis model for breast cancer.According to the median value of1.32,844 female breast cancer patients with survival time ≥ 90 days were classified into the high-risk(422)or the low-risk group(422).Survival analysis confirmed that the overall survival rate was significantly higher in patients with low prognostic score(P=1.332e-15).Clinical correlation analysis revealed significant differences in the expressions of CXCL13,FLT3,NFKBIE,SEMA3 F,TINAGL1,TRBJ2-7 and TSLP at different ages,pathological stages and TNM stages.The relationship between risk score and immune cell infiltration indicated that the infiltration density of CD4+T cells decreased significantly in the high-risk group.The prognostic value of this immune-related prognostic model was further successfully validated in ICGC database and 5-year AUC reached up to 0.823.Conclusion: The prognostic signature of immune-related genes can effectively predict the prognosis of patients with breast cancer.Combined with the clinicopathological characteristics of breast cancer patients,high-risk groups can be screened out,and a personalized clinical diagnosis and treatment scheme can be proposed for them.