Unit of Urology, Polytechnic University of the Marche Region, Azienda Ospedaliero-Universitaria Ospedali Riuniti di Ancona, Ancona, Italy.
Unit of Radiology, IRCSS INRCA, Ancona, Italy.
Urol Oncol. 2022 Aug;40(8):379.e1-379.e8. doi: 10.1016/j.urolonc.2022.04.011. Epub 2022 May 31.
To develop a nomogram incorporating clinical and multiparametric magnetic resonance imaging (mpMRI) parameters for the detection of clinically significant prostate cancer (csCaP) at radical prostatectomy (RP).
We retrospectively analyzed all consecutive patients who underwent robotic RP between 2016 and 2020. All patients underwent a 1.5-T mp-MRI according to the PI-RADS-v2 scoring system. RP specimens were examined with the whole-mount technique. csCaP definition: any tumor with a volume larger than 0.5 cm or with a Gleason score ≥7. Univariable logistic regression models explored the association between clinical and imaging data and the risk of csCaP. Significant variables (P < 0.05) were selected into multivariable regression models to identify independent predictors. A nomogram was designed to select the significant relevant predictors. The nomogram was internally validated in terms of discrimination and calibration. Receiver operating characteristics of the area under the curve was used to assess the discrimination ability of the nomogram. To assess the predictive performance of mpMRI, the accuracy of the mpMRI-based nomogram was compared with that excluding either PI-RADS score or mpMRI IL size.
The analysis involved 393 patients. The median age was 65(9) years. The median prostate specific antigen was 5.81(3.76) ng/ml. 363 had csCaP. PI-RADS v2 score of 4-5, prostate specific antigen density of 0.15 or more, and mpMRI index lesion (IL) size were significantly associated with csCaP in the multivariable regression analyses. Based on these variables, a diagnostic model was developed. The full model yielded an area under the curve of 0.77 (95%CI:0.75-0.80) which was significantly better than those excluding mpMRI findings (P = 0.02) Decision curve analysis showed a slight but significant net benefit associated with the use of the mp-MRI based nomograms compared with those excluding either PI-RADS score (Delta net benefit 0.0278) or mpMRI maximum IL size (Delta net benefit 0.0111).
The nomogram constructed in this study can assist urologists in assessing an individual's risk of csCaP at RP.
建立一种列线图，纳入临床和多参数磁共振成像（mpMRI）参数，用于检测根治性前列腺切除术（RP）时的临床显著前列腺癌（csCaP）。
我们回顾性分析了 2016 年至 2020 年间连续接受机器人 RP 的所有患者。所有患者均根据 PI-RADS-v2 评分系统行 1.5-T mp-MRI 检查。RP 标本采用全切片技术检查。csCaP 的定义：任何体积大于 0.5cm 或 Gleason 评分≥7 的肿瘤。单变量逻辑回归模型探讨了临床和影像学数据与 csCaP 风险之间的关联。选择有统计学意义的变量（P＜0.05）进入多变量回归模型，以确定独立预测因子。设计列线图以选择显著相关的预测因子。从区分度和校准度两方面对列线图进行内部验证。采用曲线下面积的受试者工作特征来评估列线图的区分能力。为了评估 mpMRI 的预测性能，将基于 mpMRI 的列线图的准确性与排除 PI-RADS 评分或 mpMRI IL 大小的准确性进行比较。
本研究共纳入 393 例患者。中位年龄为 65（9）岁。中位前列腺特异性抗原为 5.81（3.76）ng/ml。363 例患者患有 csCaP。多变量回归分析显示，PI-RADS v2 评分 4-5、前列腺特异性抗原密度≥0.15 和 mpMRI 指数病变（IL）大小与 csCaP 显著相关。基于这些变量，建立了一个诊断模型。全模型的曲线下面积为 0.77（95%CI：0.75-0.80），明显优于排除 mpMRI 结果的模型（P=0.02）。决策曲线分析显示，与排除 PI-RADS 评分（净获益差值 0.0278）或 mpMRI 最大 IL 大小（净获益差值 0.0111）相比，使用基于 mp-MRI 的列线图具有轻微但显著的净获益。
本研究构建的列线图可帮助泌尿科医生评估 RP 时个体 csCaP 的风险。
