Department of Urology, Medical University of Vienna, Vienna, Austria.
J Urol. 2010 Jul;184(1):53-8. doi: 10.1016/j.juro.2010.03.026. Epub 2010 May 15.
We developed and externally validated a prognostic nomogram specifically for papillary renal cell carcinoma.
We retrospectively studied 435 patients who underwent radical or partial nephrectomy for papillary renal cell carcinoma at 3 institutions. Slides were reviewed by 1 uropathologist per institution. We constructed a nomogram predicting 5-year disease specific survival as a graphic representation of significant variables on multivariate Cox proportional hazards regression analysis using data on 258 patients from 2 of the 3 institutions. Nomogram discrimination and calibration were assessed by bootstrapping to obtain relatively unbiased estimates. External validation was done in 177 patients from a third institution.
At a median 50.8-month followup 77 papillary renal cell carcinoma related deaths had occurred. In the multivariate Cox proportional hazards model incidental detection, T classification, M classification, vascular invasion and tumor necrosis extent were retained as independent prognostic factors of disease specific survival and formed the basis of the nomogram. The nomogram predicted well with a 93.6% bootstrapped corrected concordance index and showed good calibration. External independent validation revealed 94.2% predictive accuracy.
We developed a highly accurate tool specifically for papillary renal cell carcinoma using basic clinical and pathological information to predict disease specific survival. This tool should be helpful to identify papillary renal cell carcinoma with aggressive clinical behavior and may contribute to the ability to individualize postoperative surveillance and therapy.
我们开发并外部验证了一个专门针对乳头状肾细胞癌的预后列线图。
我们回顾性研究了 3 家机构的 435 例接受根治性或部分肾切除术的乳头状肾细胞癌患者。每家机构的 1 位泌尿病理学家对切片进行了复查。我们使用来自 3 家机构中的 2 家机构的 258 例患者的数据，通过多变量 Cox 比例风险回归分析构建了一个预测 5 年疾病特异性生存的列线图，作为显著变量的图形表示。通过自举法评估列线图的区分度和校准度，以获得相对无偏估计。在第 3 家机构的 177 例患者中进行了外部验证。
在中位随访 50.8 个月时，77 例乳头状肾细胞癌相关死亡。在多变量 Cox 比例风险模型中，偶然发现、T 分类、M 分类、血管侵犯和肿瘤坏死程度被保留为疾病特异性生存的独立预后因素，并构成了列线图的基础。该列线图预测效果良好，bootstrap 校正后的一致性指数为 93.6%，且校准效果良好。外部独立验证显示预测准确率为 94.2%。
我们使用基本的临床和病理信息为乳头状肾细胞癌开发了一个高度准确的工具，用于预测疾病特异性生存。该工具有助于识别具有侵袭性行为的乳头状肾细胞癌，并可能有助于实现个体化术后监测和治疗。
