Unit of Urology/Division of Oncology, URI, IRCCS Ospedale San Raffaele, Milan, Italy; Vita-Salute San Raffaele University, Milan, Italy.
Unit of Urology/Division of Oncology, URI, IRCCS Ospedale San Raffaele, Milan, Italy; Vita-Salute San Raffaele University, Milan, Italy; Cancer Prognostics and Health Outcomes Unit, University of Montreal Health Center, Montreal, Canada.
Eur Urol. 2017 Oct;72(4):632-640. doi: 10.1016/j.eururo.2017.03.049. Epub 2017 Apr 12.
Preoperative assessment of the risk of lymph node invasion (LNI) is mandatory to identify prostate cancer (PCa) patients who should receive an extended pelvic lymph node dissection (ePLND).
To update a nomogram predicting LNI in contemporary PCa patients with detailed biopsy reports.
DESIGN, SETTING, AND PARTICIPANTS: Overall, 681 patients with detailed biopsy information, evaluated by a high-volume uropathologist, treated with radical prostatectomy and ePLND between 2011 and 2016 were identified.
A multivariable logistic regression model predicting LNI was fitted and represented the basis for a coefficient-based nomogram. The model was evaluated using the receiver operating characteristic-derived area under the curve (AUC), calibration plot, and decision-curve analyses (DCAs).
The median number of nodes removed was 16. Overall, 79 (12%) patients had LNI. A multivariable model that included prostate-specific antigen, clinical stage, biopsy Gleason grade group, percentage of cores with highest-grade PCa, and percentage of cores with lower-grade disease represented the basis for the nomogram. After cross validation, the predictive accuracy of these predictors in our cohort was 90.8% and the DCA demonstrated improved risk prediction against threshold probabilities of LNI ≤20%. Using a cutoff of 7%, 471 (69%) ePLNDs would be spared and LNI would be missed in seven (1.5%) patients. As compared with the Briganti and Memorial Sloan Kettering Cancer Center nomograms, the novel model showed higher AUC (90.8% vs 89.5% vs 89.5%), better calibration characteristics, and a higher net benefit at DCA.
An ePLND should be avoided in patients with detailed biopsy information and a risk of nodal involvement below 7%, in order to spare approximately 70% ePLNDs at the cost of missing only 1.5% LNIs.
We developed a novel nomogram to predict lymph node invasion (LNI) in patients with clinically localized prostate cancer based on detailed biopsy reports. A lymph node dissection exclusively in men with a risk of LNI >7% according to this model would significantly reduce the number of unnecessary pelvic nodal dissections with a risk of missing only 1.5% of patients with LNI.
为了识别需要接受扩大盆腔淋巴结清扫术（ePLND）的前列腺癌（PCa）患者，术前评估淋巴结侵犯（LNI）的风险是强制性的。
利用详细的活检报告，更新一种预测当代 PCa 患者 LNI 的列线图。
设计、地点和参与者：共确定了 681 名接受高容量泌尿科医生评估、2011 年至 2016 年间接受根治性前列腺切除术和 ePLND 治疗的详细活检信息的患者。
拟合了一个预测 LNI 的多变量逻辑回归模型，并将其作为基于系数的列线图的基础。使用接受者操作特征（ROC）曲线下面积（AUC）、校准图和决策曲线分析（DCA）来评估该模型。
中位淋巴结切除数为 16 个。总体而言，79 名（12%）患者有 LNI。一个包含前列腺特异性抗原、临床分期、活检 Gleason 分级组、最高级别 PCa 核心百分比和低级别疾病核心百分比的多变量模型是该列线图的基础。经过交叉验证，这些预测因子在我们的队列中的预测准确性为 90.8%，DCA 表明在 LNI≤20%的阈值概率下改善了风险预测。使用 7%的截断值，可以避免 471 例（69%）ePLND，并使 7 例（1.5%）患者的 LNI 漏诊。与 Briganti 和 Memorial Sloan Kettering 癌症中心的列线图相比，新模型显示出更高的 AUC（90.8%比 89.5%比 89.5%）、更好的校准特征和 DCA 更高的净效益。
对于有详细活检信息且淋巴结受累风险低于 7%的患者，应避免进行 ePLND，以避免约 70%的 ePLND，而仅漏诊 1.5%的 LNI。
我们基于详细的活检报告，为临床局限性前列腺癌患者开发了一种新的列线图来预测淋巴结侵犯（LNI）。根据该模型，仅对 LNI 风险>7%的男性进行淋巴结清扫术，将显著减少不必要的盆腔淋巴结清扫术数量，仅漏诊 1.5%的 LNI 患者。
