^(68)Ga-PSMA PET/CT预测PSA<20.0μg/L有临床意义前列腺癌的效能和价值

A SUV_(max) model based on ^(68)Ga-PSMA PET/CT to predict clinically significant prostate cancer in patients with PSA levels<20.0μg/L

目的 探讨基于^(68)Ga-PSMA PET/CT构建SUV_(max)多参数前列腺癌(PCa)风险模型预测PSA<20μg/L有临床意义前列腺癌(csPCa)的效能和价值。方法 回顾分析2018年10月至2020年9月南京医科大学附属南京第一医院60例接受^(68)Ga-PSMA PET/CT检查且PSA 4.0~20.0μg/L患者的临床资料,采用经直肠超声穿刺活检(TRUS-GB)或PSMA PET/CT靶向穿刺活检(PET-TB)。PET-TB采用俯卧位、经臀肌群入路单针穿刺新模式。应用Logistic回归分析建立基于SUV_(max)多参数csPCa预测模型,绘制ROC曲线下面积。结果 PCa诊断率41.7%(25/60),csPCa诊断率35.0%(21/60)。SUV_(max)诊断csPCa cut-off值为10.075,经log_(2)正态转化后的SUV_(max)是csPCa独立预测因素。以年龄、PSA和SUV_(max)构建csPCa风险预测模型,ROC曲线下面积0.927,诊断敏感性88.2%,特异性81.8%。将logit(P)≥-0.906定义为高风险,logit(P)<-0.906定义为低风险,本组23例患者为高风险,37例为低风险,高风险患者csPCa检出率显著高于低风险患者(70.83%vs.8.11%,P<0.001)。结论 基于SUV_(max)构建PCa风险模型可预测PSA 4.0~20.0μg/L人群穿刺阳性获益,提升csPCa诊断效能。

Objective To explore the predictive value of a SUV_(max) based model constructed by ^(68)Ga-PSMA PET/CT for clinically significant prostate cancer in patients with PSA<20.0μg/L.Methods This retrospective analysis allocated 60 patients with PSA levels between 4.0-20.0μg/L that underwent ^(68)Ga-PSMA PET/CT examination in Nanjing First Hospital Affiliated to Nanjing Medical University from October 2018 to September 2020.Patients underwent transrectal ultrasound biopsy(TRUS-GB)or PSMA PET/CT targeted needle biopsy(PET-TB).PET-TB adopted a new mode of single-puncture percutaneous transgluteal approach.A multi-parameter SUV_(max),model for csPCa was established by Logistic regression analysis,and the area under the ROC curve was drawn.Results The diagnosis rate of PCa was 41.7%(25/60),and csPCa 35%(21/60).The cut-off value of SUV_(max) for diagnosing csPCa was 10.075.SUV_(max) being log_(2) normal transformated was an independent predictive factor for csPCa.The multi-parameter model for csPCa was constructed by age,PSA and SUV_(max).The area under the ROC curve was 0.927 with diagnostic sensitivity in 88.2%and specificity in 81.8%.We defined patients with Logit(P)≥-0.906 as high risk and logit(P)<-0.906 as low risk.The results showed that there were 23 high-risk patients and 37 low-risk patients.The detection rate of csPCa in the high-risk patients was significantly higher than that in low-risk patients(70.83%vs.8.11%,P<0.001).Conclusions A SUV_(max) model based on ^(68)Ga-PSMA PET/CT could effectively predict the biopsy outcome in patients with PSA 4.0-20.0μg/L and improve the diagnostic efficiency of csPCa.