Department of Urology, Peking University People's Hospital, Beijing, China.
The Institute of Applied Lithotripsy Technology, Peking University, Beijing, China.
Ann Surg Oncol. 2024 Dec;31(13):8967-8977. doi: 10.1245/s10434-024-16226-4. Epub 2024 Sep 16.
To develop a prognostic model to manage patients with muscle-invasive bladder cancer (MIBC) undergoing radical cystectomy (RC) and chemotherapy.
Clinicopathologic characteristics and survival data were collated from a North American database to develop a model. Genomic and clinicopathologic data were also obtained from European and Asian databases to externally validate the model. Patients were classified as either "primary" or "progressive" MIBC according to non-muscle invasive stage history. Optimized cancer-specific survival (CSS) models, based on MIBC types, were constructed using Cox's proportional hazard regression. Differences of biological function and tumor immunity, between two risk-based groups stratified according to the prognostic model, were estimated.
There were 2631 participants in the American cohort, 291 in the European cohort and 142 in the Asian cohort. Under Cox's regression analysis, tumor stage, lymph node stage, age, ethnicity, and MIBC types were independent CSS predictors (all p < 0.05). The constructed nomogram, which integrated these variables, improved the predictive power. This model had good discrimination and calibration. Patients were categorized into high- or low-risk groups according to the total points calculated. Kaplan-Meier curves revealed that patients in the high-risk group had poorer survival (p < 0.001). This was confirmed with two external validation cohorts (both with p < 0.001). Higher stromal scores and increased M0 and M2 macrophage numbers were observed in samples from the high-risk group, whereas regulatory T cells had lower infiltration in these populations (all with p < 0.05).
This MIBC type-based nomogram provides accurate CSS predictions, which could help improve patient management and clinical decision-making.
开发一种预测模型，以管理接受根治性膀胱切除术（RC）和化疗的肌层浸润性膀胱癌（MIBC）患者。
从北美数据库中收集临床病理特征和生存数据以建立模型。还从欧洲和亚洲数据库中获得了基因组和临床病理数据，以对模型进行外部验证。根据非肌肉浸润阶段史，患者被分类为“原发性”或“进展性”MIBC。使用 Cox 比例风险回归构建基于 MIBC 类型的优化癌症特异性生存（CSS）模型。根据预后模型分层的两个风险组，估计生物学功能和肿瘤免疫之间的差异。
美国队列中有 2631 名参与者，欧洲队列中有 291 名参与者，亚洲队列中有 142 名参与者。在 Cox 回归分析中，肿瘤分期、淋巴结分期、年龄、种族和 MIBC 类型是 CSS 的独立预测因子（均 p<0.05）。该模型整合了这些变量的构建列线图，提高了预测能力。该模型具有良好的区分度和校准度。根据计算的总分，患者被分为高风险或低风险组。Kaplan-Meier 曲线显示，高风险组患者的生存率更差（p<0.001）。这在两个外部验证队列中得到了证实（均 p<0.001）。在高风险组的样本中观察到更高的基质评分和更多的 M0 和 M2 巨噬细胞数量，而调节性 T 细胞在这些人群中的浸润较低（均 p<0.05）。
这种基于 MIBC 类型的列线图可提供准确的 CSS 预测，有助于改善患者管理和临床决策。
