食管癌预后相关的生物标志物及预后模型的建立

Establishment of biomarkers and prognostic models related to the prognosis of esophageal cancer

目的利用肿瘤基因组图谱(the cancer genome atlas, TCGA)数据库中的食管癌样本数据,应用生物信息学分析方法构建基于mRNA、lncRNA、miRNA的预后风险评分系统,为食管癌的预后、治疗决策提供理论依据。方法下载TCGA数据库中食管癌mRNA、lncRNA、miRNA和临床病理数据,应用拉索(least absolute shrinkage and selection operator, LASSO)回归、单变量和多变量Cox回归确定一组与食管癌预后相关的RNA并建立食管癌的预后风险评分。利用RNA预后风险评分和临床病理特征构建列线图模型,通过受试者工作特征(receiver operating characteristic,ROC)曲线下面积(area under curve, AUC)和决策曲线分析法(decision curve analysis, DCA)对列线图和肿瘤分期(tumor node metastasis, TNM)模型的优劣进行评估,并使用bootstrap重抽样对列线图模型进行内部验证。结果经过筛选共得到12个与预后相关的RNA,包括SLC26A9、COX6B2、RP13-672B3.2、RXFP3、OSM,5个mRNA;BLACAT1、CTD-2034I21.2、RP11-60A24.3、RP11-1123I8.1, 4个lncRNA;hsamir-1269a、hsa-mir-135b、hsa-mir-935, 3个miRNA。模型评估显示,基于TNM分期模型的AUC为0.653,RNA预后风险评分和TNM分期联合的列线图模型的AUC为0.815,后者显示出良好的区分能力。DCA显示列线图模型的临床净收益比单纯的TNM分期模型高,且内部验证结果显示列线图模型预测值和实际值具有良好的一致性。结论基于12个RNA的预后风险评分具有良好的区分能力,有助于食管癌的临床治疗和预后决策。

Objective This study uses esophageal cancer sample data in the The Cancer Genome Atlas(TCGA)database and uses bioinformatics analysis methods to construct a prognostic risk scoring system based on mRNA,lncRNA,and miRNA,which provides a theoretical basis for the prognosis and treatment decisions of esophageal cancer.Methods Download esophageal cancer mRNA,lncRNA,miRNA and clinicopathological data in the TCGA database,use Least absolute shrinkage and selection operator(LASSO)regression,univariate and multivariate Cox regression to determine a set of RNAs related to the prognosis of esophageal cancer and establish a prognostic risk score for esophageal cancer.The RNA prognostic risk score and clinical pathological characteristics were used to construct a nomogram.The ROC Area Under Curve(AUC)and the Decision Curve Analysis(DCA)curve were used to evaluate the prognostic and poor models of RNA prognostic risk score and Tumor Node Metastasis(TNM)staging.Bootstrap resampling was used to internally verify the model.Results After screening,a total of 12 RNAs related to prognosis were obtained,including SLC26 A9,COX6 B2,OSM,RXFP3,RP13-672 B3.2,5 mRNA;BLACAT1,CTD-2034 I21.2,RP11-60 A24.3,RP11-1123 I8.1,4 lncRNA;hsa-mir-1269 a,hsa-mir-135 b,hsa-mir-935,3 miRNA.The evaluation of the nomogram model shows that the AUC based on the TNM staging model is 0.653,the AUC of the RNA prognostic risk score and the TNM staging combined model is 0.815,and the latter shows good discrimination ability.The DCA curve shows that the clinical net benefit of the model after increasing the RNA risk score is higher than that of the simple TNM staging model.The internal verification results show that the predicted and actual values of the joint model are in good agreement.Conclusion The prognostic risk score based on 12 RNAs has a good discriminating ability,which is helpful for clinical treatment and prognostic decision of esophageal cancer.