多参数MRI对PSA为4-10μg/L前列腺癌的诊断价值分析

Evaluation of multiparametric MRI examination for detecting prostate cancer in men with PSA level of 4-10μg/L

目的：探讨联合应用磁共振成像（MRI）-T2WI、DWI和DCE-MRI检查对前列腺特异性抗原（PSA）为4~10μg/L的可疑前列腺癌人群的诊断价值。方法：回顾性分析2014年7月~2016年3月间于我院收治的PSA为4~10μg/L的242例前列腺穿刺患者的临床资料。单因素和多因素Logistic回归分析患者的年龄、体质指数（body mass index,BMI）、tPSA、游离与总PSA比值（f/tPSA）、前列腺体积（prostate volume,PV）、PSA密度（PSAD）、经直肠指检（DRE）、经直肠前列腺超声（TRUS）和MRI等指标与前列腺穿刺活检阳性的相关性,建立联合常规临床检查的Logistic回归预测模型,绘制各项指标及预测模型的受试者工作特征（receiver-operating characteristic,ROC）曲线,计算与比较MRI与其他检查的曲线下面积。结果：本组242例患者中,102（41.2%）例患者MRI判定为阳性,其中64.7%（66/102）最终经组织病理学检查诊断为前列腺癌。多因素Logistic回归分析结果显示MRI（OR：14.563;95%CI：6.363~33.329）是前列腺穿刺结果的独立预测因素（P〈0.05）。ROC曲线下面积MRI〉PSAD〉年龄〉f/tPSA〉DRE〉PSA〉TRUS,依次为0.813、0.726、0.723、0.657、0.642、0.598、0.569、0.568。MRI的ROC曲线下面积显著高于其他临床指标（P〈0.05）。联合MRI、PSAD、f/tPSA、DRE和TRUS的Logistic回归预测模型1,其ROC曲线下面积为0.892,与不包括MRI检查的预测模型2的ROC曲线下面积0.757相比,差异具有统计学意义（P〈0.05）。结论：联合MR-T2WI、DWI和DCE-MRI检查有利于对PSA水平在4~10μg/L前列腺癌的早期诊断,可以明显提高预测穿刺阳性的准确性。

Objective：To investigate the diagnostic value of routine MRI examination in patients with PSA level of 4-10μg/L.Method：We retrospectively analyzed the medical records of patients with PSA value of 4-10μg/L who underwent prostate biopsies in Renji hospital from July 2014 to March 2016.We analyzed the relationship between positive probability of prostate biopsy and other data including age,BMI,tPSA,f/tPSA,prostate volume（PV）,PSAD,digital rectal examination（DRE）,transrectal ultrasonography（TRUS）and MRI using univariate and multivariate logistic regression,and then built the prediction model of prostate cancer probability by logistic regression using routine clinical examinations.The area under the receiver-operating characteristic（ROC）curve of each parameter and prediction models were evaluated to determine their diagnostic value.Result：One hundred and two patients（41.2%）showed abnormal MRI findings,and 66（64.7%）of them were histopathologically diagnosed as prostate cancer.Multivariate logistic regression analysis demonstrated that MRI（OR：14.563;95%CI：6.363-33.329）was an independent factor in predicting positive prostate biopsy.The AUC of MRI,PSAD,age,f/tPSA,DRE,tPSA,and TRUS were 0.813,0.726,0.723,0.657,0.642,0.598,0.569 and 0.568,respectively.The logistic regression model one using MRI,PSAD,f/t PSA,DRE and TRUS to predict the probability of prostate cancer had a AUC of 0.892,which was significant higher（P〈0.05）than that of logistic regression model two only using PSAD,f/t PSA,DRE and TRUS（0.757）.Conclusion：Combined T2 weighted imaging,DWI and DCE-MRI facilitate the earlier diagnosis of prostate cancer in men with PSA level of 4-10μg/L,and it can significantly improve the accuracy of predicting prostate cancer.