| Background: Breast cancer(BRCA)is one of the common malignant tumors in women,the incidence of which ranks the first among female malignant tumors,and causes serious harm to women’s physical and mental health.Anoikis plays a key role in tumor metastasis,promoting the detachment and survival of cancer cells from the primary tumor site.However,there are few systematic studies on genes related to anoikis in breast cancer.Based on The Cancer Genome Atlas(TCGA)database and high throughput Gene Expression Omnibus(GEO)database,this study downloaded and screened the risk genes related to anoikis in breast cancer.A prognostic risk model of anoikis-related genes(ARGs)in breast cancer was constructed and the immune infiltration status was evaluated.Methods:1.In this study,1231 transcriptome data and 1097 clinical data sets of BRCA were downloaded from TCGA database,and GSE159956 data set and GPL2567 platform files of BRCA were downloaded from GEO database.Then,we downloaded ARGs from the Genecards database and the Harmonizome database.2.In this study,significant difference analysis of ARGs in BRCA patients was performed(P < 0.05),and the samples of BRCA patients were classified according to the obtained 32 ARGs expression levels.Survival analysis,UMAP analysis,difference analysis,heat map mapping and immune cell difference analysis were performed for the obtained subtypes.3.The clinical data sets of patients were randomly divided into the training set and the test set by 1:1.In this study,there was at least absolute shrinkage and selection operator(Lasso)regression for the training set data of BRCA,and then cross-validation for the lasso regression model to construct a prognosis model of breast cancer based on ARGs.4.BRCA patients were divided into high risk group and low risk group according to the median risk score,and the data set samples combined with TCGA and GEO were equally divided into the training set and the test set.Kaplan-Meier(K-M)survival curve and ROC curve were used to examine the survival difference between the high-low risk groups in the training set and the test set to evaluate the accuracy of the prediction model.5.In this study,an independent prognostic analysis was performed on the constructed ARGs prognostic model of breast cancer.A nomograph was constructed based on multifactor COX regression to predict the survival of patients.The model was evaluated by Decision curve analysis(DCA).6.In this study,CIBERSORT algorithm was used to perform immune infiltration analysis and Tumor microenvironment(TME)difference analysis in high and low risk groups of breast cancer patients.The expression of ARGs in different cells in the prognostic model was analyzed through BRCA-GSE161529 of the Tumor Immune Single-Cell Hub(TISCH)database.Results:1.We downloaded 652 ARGs from Genecards and Harmonizome,and analyzed 139 Differential expression genes(DEGs)from tumor tissue and paracancer tissue in the TCGA database.Through single factor regression analysis,32 anoikis-related genes related to breast cancer prognosis were obtained(P<0.05).2.We divided BRCA patients into five subtypes using 32 significantly different genes associated with breast cancer prognosis.Survival analysis found that there was A difference in survival among the different subtypes(P<0.001),and that the prognosis of subtypes A and C was generally better than that of subtypes B,D and E.UMAP analysis showed that five types of ARGs could be distinguished according to their expression levels.3.The difference analysis of immune cells showed that there were significant differences in immune infiltration status among different types of breast cancer(P<0.001).The proportion of neutrophils in type A was high.The proportion of immature dendritic cells in type B was high.The proportion of natural killer cells,eosinophils and mast cells in type C was high.The proportion of CD56 dim natural killer cells was higher in type D.The proportion of activated B cells,activated dendritic cells,CD56 bright natural killer cells,activated CD4 T cells.The T cells was higher in type E.4.This study further constructed a risk scoring model based on 5 ARGs through LassoCox proportional risk analysis and cross-validation.The risk score of this model was0.240* E2F1 expression level +0.106* CD24 expression level +0.238* PTK6 expression level-0.196* NTRK 3 expression level-0.128* MAOA expression level.NTRK3 and MAOA were identified as protective genes with Hazard Ratio(HR)< 1,while E2F1,CD24 and PTK6 were identified as risk genes with HR > 1.5.In this study,the constructed prognostic model composed of five ARGs was verified.The K-M survival curve showed that the overall survival rate of patients in the high-risk group was significantly lower than that in the low-risk group(P<0.01),and ROC curve results show that this model has good predictive ability.Independent prognostic analysis showed that the five ARGs we constructed could be used as prognostic factors for BRCA independently of other clinical traits(P<0.05).6.The content of immune infiltration cells was different in different risk subgroups of breast cancer patients.There were significant differences in naive B cells,CD4 cells,M0 macrophages,M1 macrophages,monocytes,regulatory T cells and resting mast cells between high risk group and low risk group(P<0.05).Conclusion: In this study,a risk prognostic model composed of E2F1,CD24,PTK6,NTRK3 and MAOA was constructed,which can be used to predict the prognosis of patients with adenocarcinoma.There are significant differences in immune infiltration status among different types of breast cancer based on anoikis-related genes and the content of immune infiltration cells is different in different risk subgroups of breast cancer patients. |
PDF Full Text Request