基于竞争风险模型构建胃癌特异性生存预后列线图

Constructing a nomogram of cancer-specific survival prognosis for gastric cancer patients based on competing risk model

目的 基于竞争风险模型,探究胃癌肿瘤特异性生存(cancer-specific survival,CSS)的影响因素并构建预后模型,评估该模型的临床应用价值。方法 通过监测、流行病学和最终结果(surveillance,epidemiology,and end results,SEER)数据库提取2010至2021年确诊的胃癌患者2961例,按7∶3的比例随机分为训练集(n=2 073)和验证集(n=888)。利用单因素和多因素竞争风险分析筛选影响胃癌CSS的独立危险因素并构建预后模型,利用受试者工作特征(receiver operating characteristic,ROC)曲线、校准曲线、决策曲线分别评估模型的区分度、校准度和临床实用性。结果 多因素竞争风险分析结果显示肿瘤分级、TNM分期、T分期、N分期、原发部位手术、区域淋巴结清扫、化疗,骨转移、肺转移、肿瘤大小是胃癌CSS的独立危险因素(P均<0.05)。基于以上因素构建预后模型。训练集和验证集1年、3年、5年CSS的ROC曲线下面积分别为0.858、0.872、0.872和0.848、0.867、0.871。校准曲线结果显示模型预测1年、3年、5年CSS的校正曲线与理想曲线吻合度良好。决策曲线结果显示该模型具有较好的临床实用性。结论 本研究构建的胃癌特异性生存预后模型具有较高的预测效能及临床应用价值,有助于医师进行临床决策,从而改善患者预后,提高生存质量。

Objective To investigate the factors influencing gastric cancer-specific survival(CSS) based on competing risk model,construct a prognostic model and evaluate the clinical application value of the model.Methods A total of 2961 patients with gastric cancer from 2010 to 2021 were extracted from the Surveillance,Epidemiology and End Results(SEER) database,and then randomly allocated to the training set(n=2 073) and the validation set(n=888) in a ratio of 7∶3.Univariate and multivariate competing risk analyses were used to screen the independent influencing factors of CSS,and a prognostic model was constructed.Receiver operating characteristic(ROC) curve,calibration curve and decision curve analysis were used to evaluate the discriminative ability,calibration,and clinical utility of the model.Results Multiple competing risk analysis showed that tumor grade,TNM stage,T stage,N stage,primary tumor resection,regional lymph node dissection,chemotherapy,bone metastasis,lung metastasis,and tumor size were independent risk factors of CSS in gastric cancer(all P<0.05).A prognostic model was constructed based on these influencing factors.The area under the ROC curve(AUC) for 1-,3-and 5-year CSS in the training set and validation set were 0.858,0.872,0.872 and 0.848,0.867,0.871,respectively.Calibration curve results showed good agreement between the model's predicted 1-,3-,and 5-year CSS and the ideal curve.The results of the decision curve analysis demonstrated that the model possessed good clinical utility.Conclusion The CSS prognostic model for gastric cancer constructed in this study demonstrates high predictive efficacy and clinical utility,which could help physicians making clinical decisions,thus improving patients' prognosis and enhancing the quality of life.