基于SEER数据库的老年肺癌术后患者预后模型构建与内部验证

Establishment and internal validation of a prognostic model for elderly patients after lung cancer surgery based on SEER

目的·运用回顾性研究的分析方法,探讨术后老年肺癌患者预后的独立预测因素,并建立个体化预测预后的列线图模型。方法·通过回顾美国癌症监测、流行病学和结果(surveillance, epidemiology and end results,SEER)数据库获取术后老年肺癌患者临床资料,采用Kaplan-Meier法计算总体生存率(overall survival,OS)和癌症特异性生存率(lung cancer-specific survival,LCSS)。采用Cox单因素与多因素联合分析确定术后老年肺癌患者的独立预测因素,并选取相应独立预测因素绘制列线图。使用自助抽样法(Bootstrap)、一致性指数(concordance index,C-index)和接受者操作特征曲线(receiver operating characteristic curve,ROC)对列线图模型进行内部验证和模型效能评估,并就其区分度与TNM分期(tumor node metastasis classification,TNM classification)系统进行比较。结果·共有14 891例患者纳入研究,年龄、性别、肿瘤原发部位、肿瘤病理分型、肿瘤TNM分期、手术区域淋巴清扫的数目为术后老年肺癌患者的独立预测因素,生成的预测模型可准确预测患者预后,且准确性高于TNM分期系统(C-index=0.720 vs C-index=0.686)。结论·采用美国SEER数据库初步建立了可靠的术后老年肺癌患者预后模型,可为后续国内相关临床研究提供参考;经过内部验证,其预测效能较好。

Objective·To explore the prognostic factors for elderly patients after lung cancer surgery,and establish a prognostic nomogram to predict their prognosis with multiple factors.Methods·The surveillance,epidemiology and end results(SEER)database was used to obtain clinical data of elderly patients after lung cancer surgery,and Kaplan-Meier method was performed to calculate overall survival(OS)and lung cancer-specific survival(LCSS).Cox univariate and multivariate analysis was displayed to determine the independent predictors of postoperative lung cancer in elderly patients,and the independent predictors were selected to draw nomogram.In addition,the internal validation and effectiveness of the model were performed by Bootstrap method,concordance index(C-index)and ROC curve,the discriminative ability was compared with tumor node metastasis classification(TNM classification)system.Results·A total of 14891 patients were included in the study.The age,gender,the primary site of tumor,tumor pathological type,TNM classification and the number of regional lymph nodes cleaning were independent predictors of elderly postoperative patients with lung cancer.The prediction model could accurately predict the prognosis of patients,and the accuracy was higher than that of TNM classification system(C-index=0.720 vs C-index=0.686).Conclusion·This study establishes a reliable prognostic model for elderly patients after lung cancer surgery based on SEER.The predictive performance is excellently demonstrated by internal validation.