基于SEER数据库老年肺鳞癌患者的预后分析

Prognostic analysis of elderly patients with lung squamous cell carcinoma based on SEER database

目的基于SEER数据库分析影响老年肺鳞癌患者预后的危险因素,并构建预测模型。方法基于SEER数据库,选取2004—2015年确诊为肺鳞癌的25602例患者作为研究对象,按照7∶3随机分为训练集(17921例)与验证集(7681例);通过Kaplan-Meier生存曲线分析肺鳞癌患者基本资料与预后的相关性;利用Cox回归和LASSO回归分析确定影响老年肺鳞癌患者预后的独立危险因素并构建列线图模型;以C-index、受试者工作特征(ROC)曲线、决策曲线评估列线图模型的预测性能。结果25602例患者中死亡17717例(69.2%);多因素Cox回归分析结果显示,年龄越大、高T分期、高N分期、高M分期、未接受手术是影响老年肺鳞癌患者预后不良的危险因素(P<0.05);训练集和验证集中用于评估列线图模型的C-index分别为0.732(95%CI:0.727~0.736)和0.733(95%CI:0.725~0.739),训练集列线图模型预测半年、1年、3年肺鳞癌患者预后的曲线的曲线下面积(AUC)分别为0.772(95%CI:0.764~0.779)、0.795(95%CI:0.789~0.802)、0.846(95%CI:0.839~0.852)。验证集列线图模型预测半年、1年、3年肺鳞癌患者预后的AUC分别为0.774(95%CI:0.763~0.785)、0.805(95%CI:0.795~0.815)、0.839(95%CI:0.829~0.850)。决策曲线结果显示,列线图模型的临床实用性良好。结论老年肺鳞癌患者年龄、T分期、N分期、M分期、手术情况对其预后有显著影响,构建的列线图模型能较为直观、准确地预测患者生存率,可为评估老年肺鳞癌患者预后提供科学依据。

Objective To analyze the risk factors for the prognosis of elderly patients with lung squamous cell carcinoma and construct a prediction model based on SEER database.Methods Based on the SEER database,a total of 25602 patients diagnosed with lung squamous cell carcinoma from 2004 to 2015 were selected as the research subjects.The patients were randomly divided into training set(17921 cases)and validation set(7681 cases)according to 7∶3.Kaplan-Meier survival curve was used to analyze the correlation between basic data and prognosis of patients with lung squamous cell carcinoma.Cox regression and LASSO regression analysis were used to determine the independent risk factors for the prognosis of elderly patients with lung squamous cell carcinoma and a nomogram model was constructed.The predictive performance of the nomogram model was evaluated by C-in dex,receiver operating characteristic(ROC)curve and decision curve.Results Among the 25602 patients,17717(69.2%)died.Multivariate Cox regression analysis showed that older age,higher T stage,higher N stage,higher M stage and without surgery were risk factors for poor prognosis of elderly patients with lung squamous cell carcinoma(P<0.05).The C-index used to evaluate the nomogram model in the training set and the validation set were 0.732(95%CI:0.727-0.736)and 0.733(95%CI:0.725-0.739)respectively.The area under the curve(AUC)of the training set model for predicting the prognosis of patients with lung squamous cell carcinoma at half a year,one year and three years was 0.772(95%CI:0.764-0.779),0.795(95%CI:0.789-0.802)and 0.846(95%CI:0.839-0.852)respectively.The area under the AUC of the validation set model for predicting the prognosis of patients with lung squamous cell carcinoma at half a year,one year,and three years was 0.774(95%CI:0.763-0.785),0.805(95%CI:0.795-0.815)and 0.839(95%CI:0.829-0.850)respectively.The decision analysis curve showed that the nomogram model had good clinical practicability.Conclusion The age,T stage,N stage,M stage and operation status have significant influence on the prognosis of elderly patients with lung squamous cell carcinoma.The constructed nomogram model can intuitively and accurately predict the survival rate of patients,which can provide scientific basis for evaluation of the prognosis of elderly patients with lung squamous cell carcinoma.