儿童横纹肌肉瘤预后模型的构建及验证  

Construction and validation of a prognostic nomogram for children with rhabdomyosarcoma

作　　者：张琳[1] 熊昊[1] 李建新[1] 李晖[1] 陶芳 王卓 聂应明 陈智[1] 唐威 滕菊仙 ZHANG Lin;XIONG Hao;LI Jianxin;LI Hui;TAO Fang;WANG Zhuo;NIE Yingming;CHEN Zhi;TANG Wei;TENG Juxian(Department of Hematology and Oncology,Wuhan Children's Hospital,Tongji Medical College,Huazhong University of Science and Technology,Wuhan 430019,China)

机构地区：[1]华中科技大学同济医学院附属武汉儿童医院血液肿瘤科,武汉430019

出　　处：《临床肿瘤学杂志》2021年第7期622-628,共7页Chinese Clinical Oncology

基　　金：湖北省卫计委面上资助项目(WJ2017M195);湖北省科技厅面上资助项目(2012FFB05302);武汉市卫生计生委临床研究课题资助项目(WX13B19);武汉市中青年医学骨干人才培养计划(第二批)资助项目(2013⁃2020)。

摘　　要：目的分析影响儿童横纹肌肉瘤(RMS)预后的因素并构建列线图预测模型。方法从监测、流行病学、最终结果(SEER)数据库中收集2004年至2016年年龄≤18岁且分期明确的RMS患者1212例,并随机分为测试集(n=865)和验证集(n=347);同时收集2013年1月至2018年12月经免疫组化染色确诊的25例RMS患者作为外部验证集。采用Cox风险比例回归模型分析筛选影响儿童RMS患者生存的独立预后因素并构建预后列线图模型。采用Bootstrap法对模型进行内部验证及外部验证,并绘制校准曲线。采用Kaplan-Meier法绘制生存曲线,采用时间依赖受试者工作特征(tdROC)曲线评估该模型预测RMS患者3、5、10年生存率的准确性。结果多因素Cox风险比例回归模型分析结果显示,总分期、手术、原发部位均为影响RMS患者总生存时间的独立因素。将这些因素纳入并成功构建了列线图模型。该预测模型在测试集、内部验证集及外部验证集中C指数分别为0.762、0.833、0.792, tdROC曲线评估该模型对RMS患者预后预测的准确性,结果显示,在SEER测试集、SEER验证集、外部验证集中,列线图预测RMS患者3、5、10年生存率的曲线下面积(AUC)分别为0.753、0.768、0.736,0.818、0.831、0.808和0.839、0.794。结论通过多因素Cox风险比例回归模型筛选影响儿童RMS预后的变量,并对相关独立危险因素进行可视化构建,具有较高的临床价值,对高危人群的筛选及后续个体化诊疗方案的制定具有重要意义。Objective To analyze the prognostic factors for pediatric patients with rhabdomysarcoma(RMS)and construct a reasonable prognostic model.Methods A total of 1212 patients with RMS aged 18 years or less and with definite staging from 2004 to 2016 were collected from the Surveillance,Epidemiology and End Results(SEER)database.They were randomly divided into a test set(n=865)and a validation set(n=347).Twenty-five pediatric patients with RMS confirmed by immunohistochemical staining from January 2013 to December 2018 were collected as the external validation set.Cox proportional risk regression model was used to analyze and screen the independent prognostic factors affecting the survival of children with RMS and to construct the prognostic array model.The Bootstrap method was used for internal and external validation of the model,and the calibration curve was drawn.Kaplan-Meier method was used to plot survival curves.Time-dependent receiver operating characteristic(tdROC)curve was used to evaluate the accuracy of the model in predicting 3-,5-and 10-year survival rates for patients with RMS.Results The results of multivariate Cox proportional risk regression model showed that total stage,surgery and primary site were independent prognostic factors for overall survival of patients with RMS.These factors were incorporated and the model of the histogram was successfully constructed.The C index of the prediction model in the test set,internal validation set,and external validation set were 0.762,0.833,and 0.792,respectively.The tdROC curve evaluated the accuracy of the prediction model for the prognosis of patients with RMS,and the results showed that in the SEER test set,SEER validation set,and external validation set the area under the curve(AUC)of 3-,5-and 10 year were 0.753,0.768,0.736,and 0.818,0.831,0.808,and 0.839,0.794,respectively.Conclusion The multi-factor Cox risk proportional regression model is used to screen the variables affecting the prognosis of children’s RMS,and the visual construction of related inde

关 键 词：横纹肌肉瘤 预后 危险因素 列线图 

分 类 号：R738􀆰[医药卫生—肿瘤]