基于PI-RADS v2.1和PSAD建立前列腺穿刺活检预测模型及其效能评估  

A prediction model for prostate needle biopsy based on PI-RADS v2.1 and PSAD and its efficacy assessment

作　　者：崔萍[1] 张继[1] 王松[1] 夏金翔[1] 吴慧[1] 吴苏桐 袁野 王若焜 钟星[1] 杨晓亮[1] CUI Ping;ZHANG Ji;WANG Song;XIA Jinxiang;WU Hui;WU Sutong;YUAN Ye;WANG Ruokun;ZHONG Xing;YANG Xiaoliang(MR Room,the Second People's Hospital of Hefei,Hefei 230011,China)

机构地区：[1]合肥市第二人民医院磁共振室,安徽合肥230011

出　　处：《实用放射学杂志》2023年第7期1127-1131,共5页Journal of Practical Radiology

基　　金：合肥市卫生健康委员会2020年度应用医学研究项目(Hwk2020zd0011)。

摘　　要：目的 探讨基于前列腺影像报告和数据系统2.1版(PI-RADS v2.1)的多参数磁共振成像(mpMRI)联合前列腺特异性抗原密度(PSAD)建立前列腺穿刺活检预测模型,评估其检出前列腺癌(PCa)的诊断效能.方法 回顾性分析170例经超声引导下前列腺穿刺活检或前列腺根治术后病理证实的患者相关临床影像及病理资料.对年龄、前列腺体积(PV)、前列腺特异性抗原(PSA)、游离前列腺特异性抗原(fPSA)、fPSA与PSA比值(fPSA/PSA)、PSAD及mpMRI评分进行单因素及多因素分析,确定临床显著性前列腺癌(csPCa)独立预测因子,并建立前列腺穿刺活检预测模型.绘制受试者工作特征(ROC)曲线评估各独立预测因子及预测模型对csPCa的诊断效能,通过约登指数计算预测模型的最佳诊断阈值.结果 单因素分析显示年龄、PV、PSA、fPSA、fPSA/PSA、PSAD、PI-RADS v2.1评分在csPCa组和非csPCa组间均有统计学差异(P<0.05).多因素分析显示PSAD和PI-RADS v2.1评分为csPCa的独立预测因子.利用上述2项指标建立预测模型P,计算公式为P=PSAD+0.14×3分(1,是;0,否)+0.50×4分(1,是;0,否)+0.86×5分(1,是;0,否).PSAD、PI-RADS v2.1评分、预测模型诊断csPCa的曲线下面积(AUC)分别为0.86、0.92、0.94,差异有统计学意义(P<0.01),预测模型诊断csPCa的临界值0.67,敏感性0.90,特异性0.93.结论 PSAD和PI-RADS v2.1评分为csPCa的独立预测因子,其联合预测模型诊断csPCa的效能较高,可进一步提高检出csPCa的准确性,减少不必要的前列腺穿刺活检,值得在临床推广应用.Objective To develop a prediction model for prostate needle biopsy based on the multiparametric magnetic resonance imaging(mpMRI)of prostate imaging reporting and data system version 2.1(PI-RADS v2.1)combined with prostate specific antigen density(PSAD)and to assess the diagnostic efficacy of this model for the detection of prostate cancer(PCa).Methods The clinical imaging and pathological data of 170 patients who were pathologically diagnosed after ultrasound-guided prostate needle biopsy or radical prostatectomy were analyzed retrospectively.Age,prostate volume(PV),prostate specific antigen(PSA),free PSA(fPSA),fPSA/PSA ratio,PSAD,and mpMRI scores were subjected to univariate and multivariate analysis to identify the independent predictors of clinically significant PCa(csPCa)and construct a prediction model for prostate needle biopsy.The diagnostic efficacy of each independent predictor and the constructed prediction model for csPCa was assessed by plotting receiver operating characteristic(ROC)curves,and the optimal diagnostic threshold of this prediction model was calculated by the Youden index.Results Univariate analysis results showed that age,PV,PSA,fPSA,fPSA/PSA ratio,PSAD,and PI-RADS v2.1 scores were all statistically different between the csPCa and non-csPCa groups(P<0.05).Multivariate analysis results revealed that PSAD and PI-RADS v2.1 scores were the independent predictors of csPCa.A prediction model P was developed based on the above two predictors with the following formula:P=PSAD+0.14X3 scores(1,yes;0,no)+0.50X4 scores(1,yes;0,no)+0.86X5 scores(1,yes;0,no).The area under the curve(AUC)values of the PSAD,PI-RADS v2.1 scores,and prediction model in diagnosing csPCa were 0.86,0.92,and 0.94,respectively,with statistically significant differences(P<0.01)between the two groups.Moreover,the threshold value,sensitivity,and specificity of the prediction model were 0.67,0.90,and 0.93,respectively.Conclusion PSAD and PI-RADS v2.1 scores are the independent predictors of csPCa.Moreover,the combined prediction

关 键 词：前列腺癌 前列腺影像报告和数据系统 多参数磁共振成像 预测模型 

分 类 号：R737.25[医药卫生—肿瘤] R445.2[医药卫生—临床医学]