目的 探讨磁共振扩散加权成像(diffusion-weighted imaging,DWI)的表观扩散系数(apparent diffusion coefficient,ADC)值和相对ADC值对移行带高危前列腺癌(high-risk prostate cancer,hPCa)的鉴别诊断价值及与国际泌尿病理学会(International Society of Urological Pathology,ISUP)前列腺癌分级分组(grading group,GG)的相关性.材料与方法 回顾性分析经病理证实的40例移行带前列腺癌患者的双参数MRI资料,分别测量移行带癌灶和基质型增生结节的平均ADC(mean ADC,ADCmean)值和最小ADC(minimum ADC,ADCmin)值,并计算移行带癌灶与基质型增生结节ADC比值的相对ADCmean(relative ADCmean,rADCmean)值和相对ADCmin(relative ADCmin,rADCmin)值.比较hPCa组与低危前列腺癌(low-risk prostate cancer,lPCa)组之间ADCmean、ADCmin、rADCmean和rADCmin值的差异.绘制受试者工作特征(receiver operating characteristic,ROC)曲线评估ADC各参数对移行带hPCa的诊断效能,并根据约登指数确定最佳截断值.采用DeLong检验比较ROC曲线下面积(area under the curve,AUC)的差异.Spearman相关分析ADC各参数与ISUP GG之间的相关性.结果 hPCa组的ADCmean、ADCmin、rADCmean和rADCmin值均低于lPCa组(P均<0.05).ADCmean、ADCmin、rADCmean和rADCmin鉴别诊断移行带hPCa的AUC分别为0.775[95%置信区间(confidence interval,CI):0.615～0.892]、0.879(95%CI:0.736～0.960)、0.751(95%CI:0.589～0.874)和0.914(95%CI:0.782～0.979),rADCmin的AUC最大.rADCmin与ADCmean和rADCmean的AUC差异均有统计学意义(P均<0.05),但与ADCmin的AUC差异无统计学意义(P>0.05).当rADCmin最佳截断值取0.664×10-3 mm2/s,约登指数最大(0.783),诊断移行带hPCa的敏感度和特异度分别为100.00%、78.26%.ADCmean、ADCmin、rADCmean和rADCmin值与ISUP GG均呈负相关[r=-0.486(95%CI:-0.755～-0.151)、-0.613(95%CI:-0.769～-0.365)、-0.553(95%CI:-0.745～-0.260)、-0.678(95%CI:-0.810～-0.474),P均≤0.001].结论 rADCmin鉴别诊断移行带hPCa的效能高,并且能够无创预测移行带PCa的ISUP GG,有助于为患者的个性化治疗决策提供支持.
Differential diagnosis of MRI apparent diffusion coefficient for high-risk prostate cancer in the transition zone and its correlation with pathological grading group
Objective:To investigate the differential diagnostic value of apparent diffusion coefficient(ADC)and relative ADC values of diffusion weighted imaging(DWI)for high-risk prostate cancer(hPCa)in the transition zone and their correlation with International Society of Urological Pathology(ISUP)grading group(GG).Materials and Methods:Retrospective analysis was performed on biparametric MRI data from 40 patients with transition zone prostate cancer confirmed by pathology.This analysis involved measuring the mean ADC(ADCmean)and minimum ADC(ADCmin)of transition zone prostate cancer and stromal hyperplastic nodules.Additionally,it calculated the relative ADCmean(rADCmean)and relative ADCmin(rADCmin),defined as the ratio of ADC values between transition zone carcinoma foci and stromal hyperplastic nodules.The receiver operating characteristic(ROC)curve was used to assess the diagnostic efficacy of each ADC parameter for hPCa in the transition zone and to determine the optimal cutoff value based on the Youden's index.DeLong's test was used to compare the differences in area under the curve(AUC)of the ROC curve.Spearman correlation analysis was performed to analyze the correlation between each of the ADC parameters and ISUP GG.Results:The values of ADCmean,ADCmin,rADCmean and rADCmin in the hPCa group were lower than those in the lPCa group(all P<0.05).The AUCs for the differential diagnosis of hPCa in the transition zone were 0.775[95%confidence interval(CI):0.615-0.892]、0.879(95%CI:0.736-0.960)、0.751(95%CI:0.589-0.874)and 0.914(95%CI:0.782-0.979)for ADCmean,ADCmin,rADCmean and rADCmin,respectively.The maximum AUC was observed with rADCmin.rADCmin showed statistically significant differences in AUC compared to both ADCmean and rADCmean(all P<0.05),but not with ADCmin(P>0.05).When the optimal cutoff value of rADCmin was taken as 0.664×10-3 mm2/s with the highest Youden's index(0.783),the sensitivity and specificity of diagnosing hPCa in the transition zone were 100.00%and 78.26%,respectively.ADCmean,ADCmin,rADCmean and rADCmin values were all negatively correlated with ISUP GG[r=-0.486(95%CI:-0.755--0.151),-0.613(95%CI:-0.769--0.365),-0.553(95%CI:-0.745--0.260)and-0.678(95%CI:-0.810--0.474,all P≤0.001].Conclusions:The efficacy of rADCmin in differential diagnosing hPCa in the transition zone was high.rADCmin was able to noninvasively predict ISUP GG of PCa in the transition zone,which can help to provide personalized treatment decision support for patients.
prostate neoplasmsprostate hyperplasiatransition zonemagnetic resonance imagingapparent diffusion coefficientgrading group
万方数据
维普
