Xu Lili, Zhang Gumuyang, Zhang Xiaoxiao, Bai Xin, Yan Weigang, Xiao Yu, Sun Hao, Jin Zhengyu
Department of Radiology, Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of Medical Sciences, Beijing, China.
Department of Urology, Peking Union Medical College Hospital, Peking Union Medical College, Chinese Academy of Medical Sciences, Beijing, China.
Front Oncol. 2021 Apr 1;11:655093. doi: 10.3389/fonc.2021.655093. eCollection 2021.
To externally validate the extraprostatic extension (EPE) grade criteria on MRI and analyze the incremental value of EPE grade to clinical models of prostate cancer.
A consecutive 130 patients who underwent preoperative prostate MRI followed by radical prostatectomy between January 2015 to January 2020 in our institution were retrospectively enrolled. The EPE grade, Cancer of the Prostate Risk Assessment (CAPRA), and Memorial Sloan Kettering Cancer Center nomogram (MSKCCn) score for each patient were assigned. Significant clinicopathological factors in univariate and multivariate analyses were combined with EPE grade to build the Clinical + EPE grade model, and the CAPRA and MSKCCn score were also combined with EPE grade to build the CAPRA + EPE grade and MSKCCn + EPE grade model, respectively. The area under the curve (AUC), sensitivity and specificity of these models were calculated to evaluate their diagnostic performance. Calibration and decision curve analyses were used to analyze their calibration performance and clinical utility.
The AUC for predicting EPE was 0.767-0.778 for EPE grade, 0.704 for CAPRA, and 0.723 for MSKCCn. After combination with EPE grade, the AUCs of these clinical models increased significantly than using clinical models along ( < 0.05), but was comparable with using EPE grade alone ( > 0.05). The calibration curves of EPE grade, clinical models and combined models showed that these models are well-calibrated for EPE. In the decision curve analysis, EPE grade showed slightly higher net benefit than MSKCCn and CAPRA.
The EPE grade showed good performance for evaluating EPE in our cohort and possessed well clinical utility. Further combinations with the EPE grade could improve the diagnostic performance of clinical models.
对外验证MRI上前列腺外扩展（EPE）分级标准，并分析EPE分级对前列腺癌临床模型的增量价值。
回顾性纳入2015年1月至2020年1月在本机构接受术前前列腺MRI检查并随后进行根治性前列腺切除术的连续130例患者。为每位患者确定EPE分级、前列腺癌风险评估（CAPRA）和纪念斯隆凯特琳癌症中心列线图（MSKCCn）评分。单因素和多因素分析中的显著临床病理因素与EPE分级相结合构建临床+EPE分级模型，CAPRA和MSKCCn评分也分别与EPE分级相结合构建CAPRA+EPE分级模型和MSKCCn+EPE分级模型。计算这些模型的曲线下面积（AUC）、敏感性和特异性以评估其诊断性能。采用校准和决策曲线分析来分析其校准性能和临床效用。
EPE分级预测EPE的AUC为0.767 - 0.778，CAPRA为0.704，MSKCCn为0.723。与EPE分级相结合后，这些临床模型的AUC比单独使用临床模型时显著增加（<0.05），但与单独使用EPE分级时相当（>0.05）。EPE分级、临床模型和联合模型的校准曲线表明这些模型对EPE具有良好的校准。在决策曲线分析中，EPE分级显示出比MSKCCn和CAPRA略高的净效益。
EPE分级在我们的队列中对评估EPE表现良好且具有良好的临床效用。与EPE分级的进一步结合可提高临床模型的诊断性能。
