Margue Gaëlle, Ferrer Loïc, Etchepare Guillaume, Bigot Pierre, Bensalah Karim, Mejean Arnaud, Roupret Morgan, Doumerc Nicolas, Ingels Alexandre, Boissier Romain, Pignot Géraldine, Parier Bastien, Paparel Philippe, Waeckel Thibaut, Colin Thierry, Bernhard Jean-Christophe
Bordeaux University Hospital, Urology department, Bordeaux, France.
Kidney Cancer group of the French Association of Urology Cancer Committee, Paris, France.
NPJ Precis Oncol. 2024 Feb 23;8(1):45. doi: 10.1038/s41698-024-00532-x.
Renal cell carcinoma (RCC) is most often diagnosed at a localized stage, where surgery is the standard of care. Existing prognostic scores provide moderate predictive performance, leading to challenges in establishing follow-up recommendations after surgery and in selecting patients who could benefit from adjuvant therapy. In this study, we developed a model for individual postoperative disease-free survival (DFS) prediction using machine learning (ML) on real-world prospective data. Using the French kidney cancer research network database, UroCCR, we analyzed a cohort of surgically treated RCC patients. Participating sites were randomly assigned to either the training or testing cohort, and several ML models were trained on the training dataset. The predictive performance of the best ML model was then evaluated on the test dataset and compared with the usual risk scores. In total, 3372 patients were included, with a median follow-up of 30 months. The best results in predicting DFS were achieved using Cox PH models that included 24 variables, resulting in an iAUC of 0.81 [IC95% 0.77-0.85]. The ML model surpassed the predictive performance of the most commonly used risk scores while handling incomplete data in predictors. Lastly, patients were stratified into four prognostic groups with good discrimination (iAUC = 0.79 [IC95% 0.74-0.83]). Our study suggests that applying ML to real-world prospective data from patients undergoing surgery for localized or locally advanced RCC can provide accurate individual DFS prediction, outperforming traditional prognostic scores.
肾细胞癌（RCC）最常于局部阶段被诊断出来，此时手术是标准治疗方法。现有的预后评分提供的预测性能一般，这给术后制定随访建议以及选择能从辅助治疗中获益的患者带来了挑战。在本研究中，我们利用机器学习（ML）对真实世界的前瞻性数据建立了一个用于预测个体术后无病生存期（DFS）的模型。我们使用法国肾癌研究网络数据库UroCCR，分析了一组接受手术治疗的RCC患者。参与研究的站点被随机分配到训练队列或测试队列，并在训练数据集上训练了多个ML模型。然后在测试数据集上评估最佳ML模型的预测性能，并与常用的风险评分进行比较。总共纳入了3372例患者，中位随访时间为30个月。使用包含24个变量的Cox PH模型在预测DFS方面取得了最佳结果，一致性AUC为0.81 [95%置信区间0.77 - 0.85]。该ML模型在处理预测变量中的不完整数据时，其预测性能超过了最常用的风险评分。最后，患者被分为四个具有良好区分度的预后组（一致性AUC = 0.79 [95%置信区间0.74 - 0.83]）。我们的研究表明，将ML应用于接受手术治疗的局部或局部晚期RCC患者的真实世界前瞻性数据，可以提供准确的个体DFS预测，优于传统的预后评分。
