基于TCGA和METABRIC数据分析同源重组修复基因对乳腺癌总生存期的影响

Analysis of expression and prognosis of homologous recombination repair genes in breast cancer based on TCGA and METABRIC data

目的应用公共数据库分析同源重组修复(homologous recombination repair,HRR)基因对乳腺癌总生存的影响。方法下载整理分析美国癌症基因组数据库(the cancer genome atlas,TCGA)和加拿大与英国合作的乳腺癌国际联盟分子分类数据库(the molecular taxonomy of breast cancer international consortium,METABRIC)乳腺癌HRR基因表达数据和临床数据。应该Cox风险比例回归模型进行单因素和多因素分析,探讨临床病理特征、HRR基因表达情况与总生存(overall survival,OS)的关系,应用Kaplan-Meier法绘制生存曲线。应用在线分析工具,探索候选目标HRR基因在TCGA数据中,乳腺癌组织与正常乳腺组织的基因表达差异。结果经过Cox风险比例回归模型的单因素和多因素分析,细胞周期蛋白依赖性激酶12(Cylin Dependent Kinase 12),范尼可贫血补充组Ⅰ(Fancoi Anemia Complementation GroupⅠ,CDK12)、FANCI、RAD54L以及BLM基因表达与乳腺癌总生存显著相关。新变量HRR高评分组预后更差,10年生存率为68.7%,而HRR低评分组为86.0%。在三阴特征、淋巴结转移阳性、病理类型和分期情况的亚组分析中,HRR高评分组与低评分组生存差异更明显。在TCGA数据库中,FANCI、RAD54L及BLM的基因表达情况在乳腺癌组织中较正常乳腺组织显著升高。结论基于公共数据库分析HRR基因能够预测乳腺癌总生存,尤其是四基因联合评估能够更好的预测预后,需要中国人乳腺癌样本进一步验证。

Objective To analyze the effect of homologous recombination repair(HRR)genes on overall survival of breast cancer in public databases.Methods HRR genes expression data and clinical data of breast cancer patients from TCGA and METABRIC was downloaded and analyzed.Cox proportional hazard regression model was used to explore the associations between clinicopathological features or HRR gene expression and overall survival(OS)by univariate and multivariate analysis.Kaplan-Meier method was used to plot survival curves.We also used the online analysis tool to explore the differences between breast cancer and normal breast tissues in gene expression of candidate target HRR genes based on TCGA data.Results The expressions of CDK12,FANCI,RAD54L and BLM genes were significantly associ⁃ated with OS in breast cancer patients using Cox proportional hazard regression model by univariate and multivariate anal⁃ysis.The new variable HRR high score group had a worse prognosis,with a 10-year survival rate of 68.7%compared with 86.0%in the HRR low score group.In the subgroup analysis of triple negative status,positive lymph node metas⁃tasis,histological type and TNM stage,the OS difference between the HRR high score group and the low score group was more significant.In TCGA data,the gene expressions of FANCI,RAD54L and BLM in breast cancer tissues were signif⁃icantly higher than those of normal breast tissues.Conclusion HRR gene analysis based on the public database might predict OS in breast cancer,especially the combined assessment of four candidate target genes can better forecast prog⁃nosis,which needs to be further verified by Chinese breast cancer samples study.