切换至 "中华医学电子期刊资源库"
高级检索
中华腔镜泌尿外科杂志(电子版)

中华腔镜泌尿外科杂志(电子版) ›› 2023, Vol. 17 ›› Issue (04) : 343 -350. doi: 10.3877/cma.j.issn.1674-3253.2023.04.007

临床研究

乳头状肾细胞癌预后预测模型的构建及验证
张磊磊, 蒋方, 徐疾飞, 周真文, 郭志文, 毕满华†()   
  1. 246003　安徽医科大学附属安庆第一人民医院泌尿外科
  • 收稿日期:2022-10-28 出版日期:2023-08-01
  • 通信作者: 毕满华

Construction and validation of prognosis prediction model for papillary renal cell carcinoma

Leilei Zhang, Fang Jiang, Jifei Xu, Zhenwen Zhou, Zhiwen Guo, Manhua Bi†()   

  1. Department of Urology, Anqing First People's Hospital of Anhui Medical University, Anqing 246003, China
  • Received:2022-10-28 Published:2023-08-01
  • Corresponding author: Manhua Bi
全文 ( ) 下载PDF ( ) 导出引用
引用本文：

张磊磊, 蒋方, 徐疾飞, 周真文, 郭志文, 毕满华. 乳头状肾细胞癌预后预测模型的构建及验证[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2023, 17(04): 343-350.

Leilei Zhang, Fang Jiang, Jifei Xu, Zhenwen Zhou, Zhiwen Guo, Manhua Bi. Construction and validation of prognosis prediction model for papillary renal cell carcinoma[J/OL]. Chinese Journal of Endourology(Electronic Edition), 2023, 17(04): 343-350.

目的

构建乳头状肾细胞癌患者的预后预测模型并进行验证，为该疾病患者的治疗和预后预测提供参考。

方法

从SEER数据库（美国国立癌症研究数据库） （2005~2015年）中收集筛选出共计11 396例乳头状肾细胞癌患者的数据，使用单因素Cox回归分析预后的风险因素，将具有统计学意义的风险因素纳入多因素Cox回归分析其预后的独立影响因素。依据Cox回归结果绘制列线图预测模型，同时将列线图总分值作为风险分层依据进行统计分析。使用C指数（C-index）和校正曲线确定列线图的预测准确度，使用ROC曲线下面积即AUC值判断该模型的区分度。

结果

成功构建了基于年龄、Fuhrman分级、TNM分期、肿瘤大小、手术方式等因素的乳头状肾细胞癌预后预测模型，C指数以及校准曲线验证表明，该列线图模型的预测结果与实际观测结果吻合较好。所有ROC曲线的AUC值均大于0.7，模型建立组和验证组的校准图显示OS和CSS的预测列线图之间具有良好的一致性。

结论

多因素Cox回归分析表明，年龄≥65岁、Fuhrman Ⅲ及Ⅳ级、T3及T4分期、N1及N2分期、M1分期、肿瘤大小≥40 mm、未手术均是影响乳头状肾细胞癌预后的独立危险因素。基于以上参数建立的预后预测模型，能够较为准确的预测乳头状肾细胞癌的预后风险，为临床诊疗提供统计学证据。

关键词: 乳头状肾细胞癌, 预后风险模型, 列线图
Objective

Constructing and validating of the prognositic prediction model of patients with papillary renal cell carcinoma (pRCC), which can provide a reference for the treatment and prognositic prediction of patients with pRCC.

Methods

A total of 11 396 patients with pRCC were selected from SEER database (2005-2015). Univariate Cox regression was used to analyze the risk factors of prognosis, and multivariate Cox regression was used to analyze the independent factors of prognosis. Based on the results of Cox regression, a nomogram colum line chart prediction model would be constructed, and the total score of the nomogram would be used as the basis for risk stratification in siatistical analysis. The c-index and correction curve were used to determine the prediction accuracy of nomogram, and the area under the ROC curve (AUC) was used to judge the discrimination of the model.

Results

A prognostic risk prediction model based on age, Fuhrman grade, TNM stage, tumor size and surgical procedure was successfully established. The c-index and calibration curve fully verified that the prediction results of the nomogram model were in good agreement with the actual observation results. In all ROC curves, all AUC values were greater than 0.7. The calibration plots of the model establishment group and the validation group showed good consistency between the predicted nomograms of overall survival and cancer specific survival.

Conclusions

The multivariate Cox regression analysis showed that age ≥65, Fuhrman Ⅲ and Ⅳ grade, T3 and T4 stage, N1 and N2 stage, M1 stage, tumor size ≥40 mm and unoperated were independent risk facters affecting prognostic of pRCC. Prognostic predictien model based on above parameters has ability to predict the prognosis of pRCC patients, and can be used in clinical work.

Key words: Papillary renal cell carcinoma, Prognostic risk model, Nomogram
图1 乳头状肾细胞癌（pRCC）患者年龄及肿瘤大小数据分割示意图
表1 乳头状肾细胞癌（pRCC）患者临床基线资料
项目 总体数据[例(%)]n=11 396 建模组[例(%)]n=7 977 验证组[例(%)]n=3 419 P值
年龄  
< 65岁 6357 (55.8) 4 433(55.7) 1 914(56.0) 0.738
≥65岁 5039 (44.2) 3 534(44.3) 1 505(44.0)
人种  
白人 7 796(68.4) 5 449(68.3) 2 347(68.6) 0.855
黑人 3 226(28.3) 2 269(28.4) 0957(28.0)
其它 .374(3.3) 259(3.2) 115(3.4)
性别  
男 8 733(76.6) 6100(76.5) 2 633(77.0) 0.532
女 2 663(23.4) 1877(23.5) 786(23.0)
Fuhrman分级  
Ⅰ 1 403(12.3) 0.963(12.1) 440(12.9) 0.136
Ⅱ 6 004(52.7) 4 169(52.3) 1 835(53.7)
Ⅲ 3 535(31.9) 2 595(32.5) 1 040(30.4)
Ⅳ 354(3.3) 250(3.1) 104(3.0)
肿瘤位置  
左肾 5 679(49.8) 3 981(49.9) 1 698(49.7) 0.813
右肾 5 717(50.2) 3 996(50.1) 1 721(50.3)
T分期  
T1 8 875(77.9) 6 206(77.8) 2 669(78.1) 0.915
T2 1 270(11.1) 0.896(11.2) 0.374(10.9)
T3 1 209(10.6) 0.844(10.6) 0.365(10.7)
T4 042(0.4) 0.31(0.4) 0.11(0.3)
N分期  
N0 10 999(96.5)0 7 707(96.6) 3 292(96.3) 0.494
N1 220(1.9) 146(1.8) 0.74(2.2)
N2 177(1.6) 124(1.6) 0.53(1.6)
M分期  
M0 11 133(97.7)0 7 789 (97.6) 3 344(97.8) 0.595
M1 263(2.3) 188 (2.4) 0.75(2.2)
肿瘤大小(mm)  
<40 6 194(54.4) 4 303(53.9) 1 891(55.3) 0.335
40~77 3 705(32.5) 2 626(32.9) 1 079(31.6)
>77 1 497(13.1) 1 048(13.2) 449(13.1)
手术方式  
未手术 213(1.9) 138(1.7) 0.75(2.2) 0.375
部分切除 5 025(44.1) 3 512(44.0) 1 513(44.3)
单纯切除 922(8.1) 0659(8.3) 263(7.7)
根治切除 4 826(42.3) 3 387(42.5) 1 439(42.1)
其它方式 410(3.6) 0281(3.5) 129(3.8)
表2 pRCC患者OS和CSS单因素及多因素Cox回归分析结果
项目 总生存率（OS） pRCC特异性生存率（CSS）
单因素回归分析 多因素回归分析 单因素回归分析 多因素回归分析
HR(95%CI) P值 HR(95%CI) P值 HR(95%CI) P值 HR(95%CI) P值
年龄  
< 65岁 1.00(参考值) 1.00(参考值)
≥65岁 2.267(2.073~2.479) <0.001 2.220(2.028~2.430) <0.001 1.845(1.604~2.122) <0.001 1.806(1.567~2.082) <0.001
人种  
白人 1.00(参考值) 1.00(参考值)
黑人 0.999(0.906~1.102) 0.987 / 0.877(0.747~1.031) 0.112 /
其他 0.885(0.679~1.153) 0.366 / 1.381(0.986~1.934) 0.061 /
性别  
男 1.00(参考值) 1.00(参考值)
女 0.862(0.775~0.960) 0.007 0.810(0.728~0.903) <0.001 0.873(0.737~1.035) 0.117 /
Fuhrman分级  
Ⅰ 1.00(参考值) 1.00(参考值)
Ⅱ 1.079(0.931~1.252) 0.313 1.102(0.948~1.280) 0.206 1.298(0.973~1.731) 0.076 1.260(0.941~1.687) 0.121
Ⅲ 1.517(1.303~1.766) <0.001 1.284(1.096~1.504) 0.002 2.796(2.105~3.714) <0.001 1.834(1.363~2.468) <0.001
Ⅳ 3.403(2.734~4.236) <0.001 1.807(1.428~2.285) <0.001 08.997(6.436~12.578) <0.001 2.549(1.770~3.670) <0.001
肿瘤位置  
左肾 1.00(参考值) 1.00(参考值)
右肾 0.932(0.854~1.017) 0.113 / 0.845(0.736~0.971) 0.017 0.889(0.774~1.023) 0.100
T分期  
T1 1.00(参考值) 1.00(参考值)
T2 1.563(1.374~1.778) <0.001 0.752(0.613~0.923) 0.006 3.067(2.536~3.708) <0.001 1.057(0.797~1.402) 0.702
T3 2.741(2.447~3.071) <0.001 1.295(1.112~1.507) 0.001 6.722(5.737~7.875) <0.001 2.058(1.643~2.578) <0.001
T4 10.855(7.377~16.059) <0.001 1.870(1.229~2.846) 0.003 030.578(19.852~47.100) <0.001 2.761(1.700~4.485) <0.001
N分期  
N0 1.00(参考值) 1.00(参考值)
N1 6.572(5.370~8.042) <0.001 2.115(1.667~2.666) <0.001 013.908(11.031~17.536) <0.001 2.553(1.937~3.365) <0.001
N2 08.283(6.792~10.101) <0.001 2.646(2.082~3.363) <0.001 019.466(15.651~24.211) <0.001 3.155(2.396~4.154) <0.001
M分期  
M0 1.00(参考值) 1.00(参考值)
M1 10.610(9.003~12.504) <0.001 3.593(2.911~4.435) <0.001 23.352(19.391~28.121) <0.001 4.496(3.511~5.758) <0.001
肿瘤大小(mm)  
<40 1.00(参考值) 1.00(参考值)
40~77 1.985(1.665~2.366) <0.001 1.172(1.054~1.304) 0.003 1.495(1.353~1.650) <0.001 1.231(1.017~1.490) 0.033
>77 6.290(5.300~7.465) <0.001 1.701(1.401~2.067) <0.001 2.624(2.340~2.942) <0.001 2.134(1.631~2.793) <0.001
手术方式  
未手术 1.00(参考值) 1.00(参考值)
部分切除 0.150(0.118~0.191) <0.001 0.229(0.178~0.294) <0.001 0.091(0.062~0.133) <0.001 0.159(0.106~0.239) <0.001
单纯切除 0.348(0.268~0.451) <0.001 0.490(0.373~0.644) <0.001 0.301(0.200~0.453) <0.001 0.418(0.272~0.645) <0.001
根治切除 0.403(0.320~0.506) <0.001 0.464(0.364~0.592) <0.001 0.478(0.338~0.676) <0.001 0.431(0.297~0.625) <0.001
其它方式 0.269(0.194~0.373) <0.001 0.391(0.280~0.547) <0.001 0.161(0.089~0.291) <0.001 0.336(0.183~0.618) <0.001
图2 基于多因素Cox回归分析结果的pRCC总生存率（OS）列线图模型
图3 基于多因素Cox回归分析结果的pRCC肿瘤特异性生存率（CSS）列线图模型
图4 pRCC患者计算列线图模型总分后风险分层图
图5 pRCC患者风险分层的OS及CSS生存曲线
图7 列线图模型预测pRCC患者预后的ROC曲线
[1]
Krawczyk KM, Nilsson H, Allaoui R, et al. Papillary renal cell carcinoma-derived chemerin, IL-8, and CXCL16 promote monocyte recruitment and differentiation into foam-cell macrophages[J]. Lab Invest, 2020, 97(11): 1296-1305.
[2]
Delahunt B, Bethwaite P, Nacey JJU. Renal cell carcinoma in new zealand: a national survival study[J]. Urology, 1994, 43(3): 300-309.
[3]
Negrier S, Rioux-Leclercq N, Ferlay C, et al. Axitinib in first-line for patients with metastatic papillary renal cell carcinoma: Results of the multicentre, open-label, single-arm, phase II AXIPAP trial[J]. Eur J Cancer, 2020, 129:107-116.
[4]
Dunnick NR. Renal cell carcinoma: staging and surveillance[J]. Abdom Radiol (NY), 2020, 41(6):1079-1085.
[5]
Gao Z, Zhang D, Duan Y, et al. A five-gene signature predicts overall survival of patients with papillary renal cell carcinoma[J]. PLoS One, 2019, 14(3): e0211491.
[6]
Klatte T, Pantuck AJ, Said JW, et al. Cytogenetic and molecular tumor profiling for type 1 and type 2 papillary renal cell carcinoma[J]. Clin Cancer Res, 2021, 15(4): 1162-1169.
[7]
Borgmann H, Musquera M, Haferkamp A, et al. Prognostic significance of fuhrman grade and age for cancer-specific and overall survival in patients with papillary renal cell carcinoma: results of an international multi-institutional study on 2189 patients[J]. World J Urol, 2019, 35(12): 1891-1897.
[8]
Sukov WR, Lohse CM, Leibovich BC, et al. Clinical and pathological features associated with prognosis in patients with papillary renal cell carcinoma[J]. J Urol, 2020, 187(1): 54-59.
[9]
Ravaud A, Oudard S, De Fromont M, et al. First-line treatment with sunitinib for type 1 and type 2 locally advanced or metastatic papillary renal cell carcinoma: a phase II study (SUPAP) by the French Genitourinary Group (GETUG) dagger[J]. Ann Oncol, 2020, 26(6):1123-1128.
[10]
Leibovich BC, Lohse CM, Cheville JC, et al. Predicting oncologic outcomes in renal cell carcinoma after surgery[J]. Eur Urol, 2018, 73(5): 772-780.
[11]
Jindal T, Jain N, Agarwal A. Type II papillary renal cell carcinoma with heterotrophic ossification: a case report[J]. Ann R Coll Surg Engl, 2018, 100(3): e49-e50.
[12]
Ronnen EA, Kondagunta GV, Ishill N, et al. Treatment outcome for metastatic papillary renal cell carcinoma patients[J]. Cancer, 2020, 107(11): 2617-2621.
[13]
Suzuki K, Nishiyama T, Hara N, et al. Kattan postoperative nomogram for renal cell carcinoma: predictive accuracy in a Japanese population[J]. Int J Urol, 2021, 18(3): 194-199.
[14]
Stenman M, Staehler M, Szabados B, et al. Metastatic papillary renal cell carcinoma in the era of targeted therapy - a retrospective study from three European academic centres[J]. Acta Oncol, 2019, 58(3): 306-312.
[15]
Tsui KH, Shvarts O, Smith RB, et al. Prognostic indicators for renal cell carcinoma: a multivariate analysis of 643 patients using the revised 1997 tnm staging criteria[J]. J Urol, 2020, 163(4): 1090-1095.
[16]
Guo S, Yao K, He X, et al. Prognostic significance of laterality in renal cell carcinoma: A population-based study from the surveillance, epidemiology, and end results (SEER) database[J]. Cancer Med, 2019, 8(12): 5629-5637.
[1] 胡可, 鲁蓉. 基于多参数超声特征的中老年女性压力性尿失禁诊断模型研究[J/OL]. 中华医学超声杂志（电子版）, 2024, 21(05): 477-483.
[2] 余晓青, 高欣, 罗文培, 杨露. BI-RADS 4类结节患者的乳腺癌风险预测模型[J/OL]. 中华乳腺病杂志(电子版), 2024, 18(04): 217-223.
[3] 蒲彦婷, 吴翠先, 兰玉梅. 类风湿关节炎患者骨质疏松症风险预测列线图模型构建[J/OL]. 中华关节外科杂志(电子版), 2024, 18(05): 596-603.
[4] 庄燕, 戴林峰, 张海东, 陈秋华, 聂清芳. 脓毒症患者早期生存影响因素及Cox 风险预测模型构建[J/OL]. 中华危重症医学杂志(电子版), 2024, 17(05): 372-378.
[5] 杜佳丽, 鲍睿, 乔春红, 韩伟. 中孕期宫颈功能不全孕妇经阴道紧急宫颈环扎术后不良妊娠结局预测模型构建[J/OL]. 中华妇幼临床医学杂志(电子版), 2024, 20(04): 403-409.
[6] 白香妮, 孙巨军, 谢鹤, 李宏斌. 急性胰腺炎患者血清微小RNA-142-3p和磷脂酰肌醇3-激酶水平变化及对并发腹腔感染风险预测[J/OL]. 中华实验和临床感染病杂志(电子版), 2024, 18(04): 222-228.
[7] 蔡大明, 陆晓峰, 王行舟, 王萌, 刘颂, 夏雪峰, 沈晓菲, 杜峻峰, 管文贤. 三级淋巴结构在胃神经内分泌瘤中的预后价值及预后预测模型构建[J/OL]. 中华普外科手术学杂志(电子版), 2024, 18(04): 401-405.
[8] 王淑贤, 张良灏, 王利君, 张慧, 郭源, 许传屾, 李志强, 蔡金贞, 解曼, 饶伟. 成人肝移植围手术期严重心血管事件危险因素分析及预测模型研究[J/OL]. 中华移植杂志(电子版), 2024, 18(04): 222-229.
[9] 屈勤芳, 束方莲. 盆腔器官脱垂患者盆底重建手术后压力性尿失禁发生的影响因素及列线图预测模型构建[J/OL]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(06): 606-612.
[10] 张瑜, 姜梦妮. 基于DWI信号值构建局部进展期胰腺癌放化疗生存获益预测模型[J/OL]. 中华肝脏外科手术学电子杂志, 2024, 13(05): 657-664.
[11] 杨秀君, 崔梦莹, 刘水, 盛基尧, 张丹. 基于SEER数据库胰头部胰腺神经内分泌癌患者预后列线图构建与验证[J/OL]. 中华肝脏外科手术学电子杂志, 2024, 13(04): 520-525.
[12] 单良, 刘怡, 于涛, 徐丽. 老年股骨颈骨折术后患者心理弹性现状及影响因素分析[J/OL]. 中华老年骨科与康复电子杂志, 2024, 10(05): 294-300.
[13] 刘燚隆, 党荣广, 艾蓉, 张凯. 肝硬化合并静脉曲张出血患者内镜治疗后再出血风险的模型建立与验证[J/OL]. 中华消化病与影像杂志(电子版), 2024, 14(04): 336-342.
[14] 韦巧玲, 黄妍, 赵昌, 宋庆峰, 陈祖毅, 黄莹, 蒙嫦, 黄靖. 肝癌微波消融术后中重度疼痛风险预测列线图模型构建及验证[J/OL]. 中华临床医师杂志(电子版), 2024, 18(08): 715-721.
[15] 颜世锐, 熊辉. 感染性心内膜炎合并急性肾损伤患者的危险因素探索及死亡风险预测[J/OL]. 中华临床医师杂志(电子版), 2024, 18(07): 618-624.
阅读次数
全文


摘要

版权所有 中华医学会 中华医学电子音像出版社有限责任公司  京ICP备14006079号-1  网络出版服务许可证：(署)网出证(京)字第075号