The LncRNA Prognostic Model Of ER-positive Breast Cancer Was Constructed Based On The TCGA Database

Objective: In women worldwide,breast cancer is the most common malignancy oestrogen receptor(estrogen receptor,ER)-positive breast cancer accounts for about70% of breast cancer,and endocrine therapy combined with surgery or chemotherapy has significantly improved the prognosis of patients with ER-positive breast cancer.However,a proportion of ER positive breast cancer patients still have disease recurrence and metastasis in the course of endocrine or chemotherapy disease,leading to poor prognosis.At present,clinicopathological factors such as tumor size,pathological grade,lymph node metastasis,and HER 2 expression are often used to judge the prognosis of patients,but they are lack of targeted prognostic indicators.Therefore,further exploration of biomarkers of prognosis in ER positive breast cancer patients are needed to facilitate the selection ofmore individualized treatment options for ER positive breast cancer patients.Methods:In the first instance,We downloaded breast cancer of RNA expression data of from the Cancer Genome Atlas(The cancer genome atlas,TCGA)database,after that from the RNAseq expression matrix,wr differentiated the expression profile of long non-coding RNA(long non-coding RNA,lnc RNA).The included samples were divided into ER positive group,ER negative group and normal breast samples according to the expression status of ER.The ER positive group samples were analyzed as cancer group and ER negative group samples as control group(p <0.01).These differential genes were screened so that the unique differential genes between ER-positive breast cancer and normal breast tissue could be identified.The selected differential genes were divided into multiple modules using weighted gene co-expression network analysis(Weighted Gene Co-Expression Network Analysis,WGCNA).According to the correlation between module and subtype,the gene module with the most relevant ER positive breast cancer was identified,and the genes in the module were analyzed by univariate COX analysis and multivariate COX analysis to select lnc RNA related to overall survival(Overall survival,OS).According to the regression coefficient of each lnc RNA,the risk score formula based on lnc RNA expression was constructed.After calculating the risk score for each sample,the median risk number was selected as the cut-off value to divide ER positive breast cancer patients into two group,the group of high relapse risk and the group of low relapse risk.Make manufacture the survival curve(Kaplan-Meier curve,K-M curve)was drawn for survival analysis,and the model was evaluated according to the area under the subject operating characteristic curve(receiver operating characteristic curve,ROC curve).Group the samples according to different clinicopathological characteristics to observe whether the predictive power of the model in different subgroups remains valid.Combining patient risk score with clinicopathological information for Single factor and multiple factor COX analysis to assess whether the model that we set was the independent factor can make independent predictions.Gene Set Enrichment Analysis,GSEA was used on the prognostic model to explore the mechanism of the prognostic model leading to poor prognosis.Conclusion:In this study,we used bioinformatics analysis to screen out six lnc RNA associated with the prognosis of ER-positive breast cancer and constructed a prognostic model for ER-positive non-metastatic breast cancer.Our study provides new biological markers for the prognostic evaluation of non-metastatic ER-positive breast cancer,especially for the selection of individualized treatment options postoperative adjuvant therapy for stage I-II ER-positive early breast cancer.