未分化型骨肉瘤患者临床预后模型的开发和应用  

Development and application of a clinical prognostic model for patients with undifferentiated osteosarcoma

作　　者：贺希希 韦庆军[1] He Xixi;Wei Qingjun(Department of Orthopedics Trauma and Hand Surgery,The First Affiliated Hospital of Guangxi Medical University,Nanning 530021,China)

机构地区：[1]广西医科大学第一附属医院创伤手外科,南宁530021

出　　处：《广西医科大学学报》2021年第1期112-118,共7页Journal of Guangxi Medical University

基　　金：国家自然科学基金资助项目(No.81560371);广西财政科技计划项目资助(No.815603712017AB45095)

摘　　要：目的:探讨未分化型骨肉瘤(OS)患者生存相关因素并构建生存预测模型。方法:使用检测、流行病学和最终结果(SEER)数据库识别并收集2004—2015年诊断为原发OS的所有患者信息。单因素和多因素Cox回归分析确定各预后因素与OS患者生存的关系。将预后因素纳入并构建列线图,通过受试者工作特征(ROC)曲线、一致性指数(C-index)和校准曲线验证预后模型的可靠性和准确性。所有统计分析均使用R 3.5.3软件进行分析。结果:共有1056例OS患者符合标准。多因素回归分析表明,20~59岁,≥60岁、T_1期、T_2期、N_1期、M_1期是OS患者的独立危险因素(P<0.05),肿瘤原发部位四肢骨和手术是独立保护因素(P<0.05)。Kaplan-Meier生存分析显示,所有预后相关因素与骨肉患者生存时间均存在统计学意义(P<0.05)。C-index为0.753,校准图证明实际值和模型预测结果存在良好的一致性,1年、3年、5年特异性生存率ROC曲线下面积(AUC)分别为0.890、0.754和0.735。结论:基于统计的预后影响因素,成功构建未分化型OS患者较为准确的预后列线图预测模型,能够帮助临床医生参考和指导治疗,有利于患者的个体化治疗。Objective:To investigate the factors associated with survival in patients with undifferentiated osteosarcoma and to construct a survival prediction model.Methods:The Surveillance,Epidemiology,and End Results(SEER)was used to identify and collect information on all patients diagnosed with primary osteosarcoma between 2004 and 2015.Univariate and multivariate Cox regression analyses were performed to determine the relationship between each prognostic factor and the survival of osteosarcoma patients.Prognostic factors were included and nomograms were constructed to validate the reliability and accuracy of the prognostic model by ROC(Receiver Operating Characteristic)curve,concordance index(C-index),and calibration curve(Calibration curve).All statistical analyses were performed by R3.5.3 software.Results:A total of 1056 osteosarcoma patients met the criteria.Multivariate regression analysis indicated that 20-59 years old,≥60 years old,T1 stage,T2 stage,N1 stage,and M1 stage were independent risk factors for osteosarcoma patients(P<0.05),and limb bone at the primary tumor site and surgery were independent protective factors(P<0.05).Kaplan-Meier analysis of survival showed statistical significance between all prognostic factors and survival time of osteosarcoma patients(P<0.05).The C-index was 0.753,and the calibration curve demonstrated good agreement between the actual value and the model prediction results,with areas under the ROC curves for 1-,3-,and 5-year specific survival rates of 0.890,0.754,and 0.735,respectively.Conclusion:Based on the statistical prognostic factors,a relatively accurate prognostic nomograms prediction model was successfully constructed for patients with undifferentiated osteosarcoma,which can help clinicians to guide the treatment and facilitate the individualized treatment of patients.

关 键 词：未分化型骨肉瘤 SEER数据库 预后模型 

分 类 号：R738.1[医药卫生—肿瘤]