Ji Junjie, Wen Zengjin, Yao Yu, Jiang Lei, Yang Qingya, Zhang Guiming
Department of Urology, The Affiliated Hospital of Qingdao University, Qingdao, China.
School of Rehabilitation, Capital Medical University, Beijing, China.
Front Oncol. 2025 Jul 3;15:1571604. doi: 10.3389/fonc.2025.1571604. eCollection 2025.
This study aimed to construct and validate nomograms for the prediction of overall survival (OS), cancer-specific survival (CSS), and disease-free survival (DFS) in patients with resectable bladder urothelial carcinoma (BUC) after radical cystectomy (RC).
We retrospectively collected the demographic, pathological, imaging, and laboratory data from patients with BUC who underwent RC. The training cohort included patients from the Affiliated Hospital of Qingdao University from January 2018 to December 2021, while the test cohort included patients from the same hospital between January 2016 and December 2017, along with patients from Qilu Hospital of Shandong University. Univariate and multivariate Cox regression analyses were conducted to identify independent predictors of OS, CSS, and DFS. The performance of the nomograms was evaluated using Harrell's concordance index (-index), the area under the receiver operating characteristic (ROC) curve (AUC), the corrected AUC following 1,000 bootstrap resamplings with calibration curves, and decision curve analysis in both cohort validations.
A total of 393 patients were included in the training cohort, while 156 patients comprised the test cohort. Multivariate analyses revealed that age, tumor size, lymph node metastasis (LNM), lymphovascular invasion (LVI), urea nitrogen, creatinine, and the albumin/fibrinogen ratio (AFR) were independent predictors for OS. For CSS, the independent predictors were tumor size, LNM, LVI, urea nitrogen, and AFR. LNM and LVI were the independent predictors for DFS. The nomograms for OS and CSS demonstrated high predictive accuracy with robust C-indexes and ROC curves, along with reliable calibration curves with corrected AUCs and clinical utility in both cohorts. The DFS nomogram also showed high predictive accuracy with stable corrected AUCs in both cohorts.
We constructed OS, CSS, and DFS nomograms to predict prognosis in patients with BUC treated with RC. These nomograms exhibited high accuracy, reliability, and clinical utility in predicting outcomes in both cohorts.
本研究旨在构建并验证列线图，用于预测根治性膀胱切除术后可切除膀胱尿路上皮癌（BUC）患者的总生存期（OS）、癌症特异性生存期（CSS）和无病生存期（DFS）。
我们回顾性收集了接受根治性膀胱切除术的BUC患者的人口统计学、病理学、影像学和实验室数据。训练队列包括2018年1月至2021年12月期间青岛大学附属医院的患者，而测试队列包括2016年1月至2017年12月期间同一家医院的患者以及山东大学齐鲁医院的患者。进行单因素和多因素Cox回归分析，以确定OS、CSS和DFS的独立预测因素。使用Harrell一致性指数（C指数）、受试者操作特征（ROC）曲线下面积（AUC）、1000次自举重采样后校正的AUC以及校准曲线在两个队列验证中的决策曲线分析来评估列线图的性能。
训练队列共纳入393例患者，测试队列共156例患者。多因素分析显示，年龄、肿瘤大小、淋巴结转移（LNM）、淋巴管侵犯（LVI）、尿素氮、肌酐和白蛋白/纤维蛋白原比值（AFR）是OS的独立预测因素。对于CSS，独立预测因素为肿瘤大小、LNM、LVI、尿素氮和AFR。LNM和LVI是DFS的独立预测因素。OS和CSS的列线图显示出较高的预测准确性，具有稳健的C指数和ROC曲线，以及两个队列中具有校正AUC的可靠校准曲线和临床实用性。DFS列线图在两个队列中也显示出较高的预测准确性，校正后的AUC稳定。
我们构建了OS、CSS和DFS列线图，以预测接受根治性膀胱切除术治疗的BUC患者的预后。这些列线图在预测两个队列的结果方面表现出较高的准确性、可靠性和临床实用性。
