基于SEER数据库和国内队列构建及验证非转移性同时性散发性双肾细胞癌预后模型

Construction and validation of a prognostic model for non-metastatic synchronous sporadic bilateral renal cell carcinoma based on SEER database and domestic cohort

目的:构建非转移性同时性散发性双肾细胞癌(synchronous sporadic bilateral renal cell carcinoma,SSBRCC)的预后模型,并利用多中心队列评估其临床应用价值。方法:回顾性分析2010—2015年录入SEER数据库的非转移性SSBRCC患者554例(SEER队列)及2007年4月—2021年11月南京大学医学院附属金陵医院和山东大学齐鲁医院收治的非转移性SSBRCC患者111例(验证队列)的临床资料。通过Kaplan-Meier生存分析比较对于SSBRCC不同手术治疗方式之间的术后5年总体生存率(overall survival,OS)差异。利用单因素和多因素Cox回归分析,筛选出SSBRCC的独立预后因素,并根据这些独立预后因素构建Nomogram预后模型。结果:与至少一侧行根治性肾切除术(radical nephrectomy,RN)的患者比较,SSBRCC患者双侧均行部分肾切除术(partial nephrectomy,PN)可以带来更好的远期生存收益。初诊时年龄≥60岁,TNM分期处于T3、T4期,组织学病理出现肉瘤样变特征,核分级Ⅲ、Ⅳ级以及至少一侧行RN是SSBRCC术后生存的独立危险因素。基于以上独立预后因素构建的5年OS Nomogram预后模型具有较好的预测效能(SEER队列:C-index=0.773,验证队列:C-index=0.918)。结论:本研究建立的预后模型可有效预测非转移性SSBRCC患者的总体预后,对于临床上简易评估临床预后具有一定的参考意义。

Objective: To investigate the construction of a survival model for non-metastatic synchronous sporadic bilateral renal cell carcinoma(SSBRCC) and evaluate its clinical application with multicenter cohorts. Methods: We retrospectively analyzed the clinical data of 554 patients with non-metastatic SSBRCC(SEER cohort) enrolled in the SEER database from 2010 to 2015 and 111 patients with non-metastatic SSBRCC(validation cohort) admitted to Jinling Hospital, Medical School of Nanjing University and Qilu Hospital of Shandong University from April 2007 to November 2021. We compared the difference in postoperative 5-year overall survival(OS) between different surgical treatment techniques for SSBRCC by Kaplan-Meier survival analysis. Independent prognostic factors in SSBRCC were screened using univariate and multifactorial Cox regression analysis, and Nomogram prognostic models were constructed based on these independent prognostic factors. Results: Partial nephrectomy(PN) on both sides in patients with SSBRCC provided better long-term survival benefits than radical nephrectomy(RN) on at least one side. Age at initial diagnosis ≥60 years, TNM staging at T3 or T4, sarcomatous features in histological pathology, nuclear classification Ⅲ or Ⅳ, and RN on at least one side were independent risk factors in survival after SSBRCC surgery. The 5-year OS Nomogram prognostic model constructed based on the above independent prognostic factors had good predictive validity(SEER cohort: C-index=0.773,Validation cohort: C-index=0.918).Conclusion: Our multicenter-based survival model can effectively predict 5-year OS in patients with non-metastatic SSBRCC, which is important for assessing clinical treatment prognosis.