摘要
目的探讨双参数MRI(bpMRI)联合临床特征预测前列腺癌包膜外侵犯(EPE)的价值。方法回顾性分析2022年1月—2023年12月本院收治的经手术治疗的前列腺癌患者110例的临床资料,术前均采用bpMRI进行检查,纳入MRI前列腺癌包膜外侵犯评估系统(MRI-EPE)、PI-RADS 3~5类评分、总前列腺特异性抗原(tPSA)、病灶最大径、前列腺体积、年龄因素进行分析,分别构建临床特征、bpMRI特征、临床+bpMRI特征的Logistic模型预测EPE。采用受试者工作特征曲线(ROC)和Delong检验验证各模型以及比较模型间的差异。结果110例前列腺癌患者中,EPE阴性组59例、EPE阳性组51例。分析结果显示MRI-EPE(χ^(2)=26.326,P<0.001)、PI-RADS 3~5类评分(2=16.683,P<0.001)、tPSA(t=-3.269,P=0.002)、病灶最大径(t=-2.704,P=0.008)与EPE情况具有相关性,基于上述参数进一步建立临床特征(tPSA)、bpMRI特征(病灶最大径+MRI-EPE+PI-RADS)、临床+bpMRI特征(tPSA+病灶最大径+MRI-EPE+PI-RADS)模型。临床+bpMRI特征模型的AUC为0.841(95%CI:0.767~0.915),分别高于临床和bpMRI特征模型的0.672(95%CI:0.571~0.774)和0.782(95%CI:0.697~0.867),Delong检验显示临床+bpMRI特征模型的AUC分别与临床、bpMRI特征模型具有统计学差异(Z=3.056、1.982,P=0.002、0.047),临床、bpMRI特征模间不具有统计学差异(Z=1.449,P=0.147)。临床+bpMRI特征模型的准确性、特异性、敏感性、阳性预测值以及阴性预测值分别为0.800、0.780、0.824、0.764、0.836。结论基于临床+bpMRI特征所构建的模型对预测EPE具有很高的参考价值。
Objective To investigate the value of combining biparametric magnetic resonance imaging(bpMRI)with clinical features in predicting extraprostatic extension(EPE)in prostate cancer.Methods A retrospective study was conducted on 110 patients with prostate cancer who underwent surgical treatment at the First Affiliated Hospital of Bengbu Medical University from January,2022 to December,2023.All patients underwent preoperative bpMRI examination.The MRI-EPE,PI-RADS 3~5 score,total prostate-specific antigen(tPSA),maximum lesion diameter,prostate volume,and age were included in the analysis.Logistic models were constructed for clinical,bpMRI,and clinical+bpMRI features to predict EPE.The receiver operating characteristic curve(ROC)and Delong test were used to validate the models and compare their differences.Results Among the 110 patients with prostate cancer,there were 59 cases in the EPE-negative group and 51 cases in the EPE-positive group.The analysis showed that MRI-EPE(χ^(2)=26.326,P<0.001),PI-RADS 3~5 score(χ^(2)=16.683,P<0.001),tPSA(t=-3.269,P=0.002),and maximum lesion diameter(t=-2.704,P=0.008)were correlated with EPE.Based on these parameters,clinical features(tPSA),bpMRI features(maximum lesion diameter+MRI-EPE+PI-RADS),and clinical+bpMRI features(tPSA+maximum lesion diameter+MRI-EPE+PI-RADS)models were further established.The AUC of the clinical+bpMRI features model was 0.841(95%CI:0.767~0.915),which was higher than the 0.672(95%CI:0.571~0.774)of the clinical features model and the 0.782(95%CI:0.697~0.867)of the bpMRI features model.The Delong test showed that the AUC of the clinical+bpMRI features model was statistically different from that of the clinical and bpMRI features models(Z=1.982,3.056;P=0.002,0.047).At the same time,there was no statistically significant difference between the clinical and bpMRI features models(Z=1.449,P=0.147).The accuracy,specificity,sensitivity,positive predictive value,and negative predictive value of the clinical+MRI features model were 0.800,0.780,0.824,0.764,and 0.836.Conclusions The model based on clinical and bpMRI features has a high reference value for predicting EPE.
作者
周牧野
苏蕾
陈艾琪
陈岩
马宜传
陈刘成
Zhou Muye;Su Lei;Chen Aiqi;Chen Yan;Ma Yichuan;Chen Liucheng(Department of Radiology,the First Affiliated Hospital of Bengbu Medical University,Bengbu,Anhui 233004,China;Department of Ultrasound,the First Affiliated Hospital of Bengbu Medical University,Bengbu,Anhui 233004,China)
出处
《齐齐哈尔医学院学报》
2024年第14期1355-1359,共5页
Journal of Qiqihar Medical University
基金
蚌埠医学院自然科学重点项目(2022byzd063)。
关键词
前列腺癌
包膜外侵犯
双参数磁共振
联合模型
Prostate cancer
Extraprostatic extension
Biparametric magnetic resonance imaging
Combined model
作者简介
通信作者:陈刘成,Email:chenliu1385521@163.com。
