背景:肾癌(KC)是成人最常见的恶性肿瘤之一,尤其影响老年患者的生存。我们旨在构建一个列线图来预测老年KC患者术后的总生存期(OS)。
方法:从监测中下载了2010年至2015年间所有年龄超过65岁并接受手术治疗的原发性KC患者的信息,流行病学,和结束结果(SEER)数据库。单因素和多因素Cox回归分析用于确定独立的预后因素。一致性指数(C-index),接收机工作特性曲线(ROC),曲线下面积(AUC),和校准曲线用于评估列线图的准确性和有效性。通过决策曲线分析(DCA)和时间依赖性ROC比较列线图和TNM分期系统的临床益处。
结果:共纳入15,989例接受手术的老年KC患者。将所有患者随机分为训练集(N=11,193,70%)和验证集(N=4796,30%)。列线图在训练集和验证集中产生0.771(95%CI0.751-0.791)和0.792(95%CI0.763-0.821)的C指数,分别,表明列线图具有出色的预测准确性。ROC,AUC,和校准曲线也显示出同样优异的结果。此外,DCA和时间依赖性ROC显示,列线图优于TNM分期系统,具有更好的净临床效益和预测功效。
结论:老年KC患者术后OS的独立影响因素为性别,年龄,组织学类型,肿瘤大小,grade,手术,婚姻,放射治疗,和T-,N-,M阶段。基于网络的列线图和风险分层系统可以帮助外科医生和患者进行临床决策。
BACKGROUND: Kidney cancer (KC) is one of the most common malignant tumors in adults which particularly affects the survival of elderly patients. We aimed to construct a nomogram to predict overall survival (OS) in elderly KC patients after surgery.
METHODS: Information on all primary KC patients aged more than 65 years and treated with surgery between 2010 and 2015 was downloaded from the Surveillance, Epidemiology, and End Results (SEER) database. Univariate and multivariate Cox regression analysis was used to identify the independent prognostic factors. Consistency index (C-index), receiver operating characteristic curve (ROC), the area under curve (AUC), and calibration curve were used to assess the accuracy and validity of the nomogram. Comparison of the clinical benefits of nomogram and the TNM staging system is done by decision curve analysis (DCA) and time-dependent ROC.
RESULTS: A total of 15,989 elderly KC patients undergoing surgery were included. All patients were randomly divided into training set (N = 11,193, 70%) and validation set (N = 4796, 30%). The nomogram produced C-indexes of 0.771 (95% CI 0.751-0.791) and 0.792 (95% CI 0.763-0.821) in the training and validation sets, respectively, indicating that the nomogram has excellent predictive accuracy. The ROC, AUC, and calibration curves also showed the same excellent results. In addition, DCA and time-dependent ROC showed that the nomogram outperformed the TNM staging system with better net clinical benefits and predictive efficacy.
CONCLUSIONS: Independent influencing factors for postoperative OS in elderly KC patients were sex, age, histological type, tumor size, grade, surgery, marriage, radiotherapy, and T-, N-, and M-stage. The web-based nomogram and risk stratification system could assist surgeons and patients in clinical decision-making.