转移性乳腺癌患者预后相关关键基因筛选、预后预测模型构建及验证.

Objective To screen out the key genes-related to the prognosis of patients with metastatic breast cancer, and to construct and verify the prognosis-prediction model of patients with metastatic breast cancer. Methods The gene expression dataset GSE124648 of metastatic breast cancer tissue was obtained from GEO database to screen out the differentially expressed genes ( DEGs) between metastatic breast cancer tissues and primary breast cancer tissues, and between metastatic breast cancer tissues and normal breast tissues. The functional enrichment of Gene Ontology (GO) and the anal ysis of Kyoto Encyclopedia of Genes and Genomes ( KEGG) signaling pathway were carried out. Data sets of 140 patients with metastatic breast cancer were randomly divided into the training set ( 72 cases) and test set ( 68 cases) . We used LASSO&COX regression model to screen out the key genes-related to the prognosis of patients with metastatic breast cancer in the training set, and constructed the prognosis prediction model of patients with metastatic breast cancer in the training set. According to the median risk value, the patients in the training set were divided into the high-risk group and low-risk group. The Kaplan-Meier survival curve was drawn to analyze the median survival time, and the ROC curve was drawn to evaluate the predictive effectiveness of the risk model. The prediction model was applied to the patients in the test set to de termine the prediction efficiency of the prediction model. Results A total of 287 DEGs of metastatic breast cancer tissues were screened out, including 29 high expression genes and 258 low expression genes. The functions of DEGs were mainly related to the proliferation and migration of breast cancer cells, the regulation and degradation of extracellular matrix, an giogenesis, and immune inflammatory reaction. Among them, 7 key genes-related to prognosis were EGFR, GEM, PT PRB, RARRESl, LAMA4, NFAT5 and LHFP. We constructed a prediction model for prognosis of metastatic breast cancer patients with training set: risk value = ( 0. 279 x EGFR) + ( 0. 704 x GEM) + ( 0. 326 x PTPRB) + ( 0. 138 x RARRESl) + ( -0. 570 x LAMA4) + (0. 262 x NFAT5) + ( -0. 555 x LHFP). In the training set, the median survival time of the high-risk group was significantly shorter than that of low-risk group (P <0. 001), and the area under the curve of 3-year survival rate of patients with metastatic breast cancer was 0. 787. In the test set, the median survival time of the high-risk group was significantly shorter than that of low-risk group ( P < 0. 05), and the area under the curve of 3-year survival rate of patients with metastatic breast cancer was 0. 785. Conclusion The prognosis-prediction model including seven genes EGFR, GEM, PTPRB, RARRESl, LAMA4, NFAT5 and LHFP is successfully constructed to predict the prognosis of pa tients with metastatic breast cancer. [ABSTRACT FROM AUTHOR]