Gondo Tatsuo, Hricak Hedvig, Sala Evis, Zheng Junting, Moskowitz Chaya S, Bernstein Melanie, Eastham James A, Vargas Hebert Alberto
Urology Service, Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY, USA.
Eur Radiol. 2014 Dec;24(12):3161-70. doi: 10.1007/s00330-014-3367-7. Epub 2014 Aug 7.
The aim of this study was to assess the diagnostic performance of pre-treatment 3-Tesla (3T) multiparametric magnetic resonance imaging (mpMRI) for predicting Gleason score (GS) downgrading after radical prostatectomy (RP) in patients with GS 3 + 4 prostate cancer (PCa) on biopsy.
We retrospectively reviewed 304 patients with biopsy-proven GS 3 + 4 PCa who underwent mpMRI before RP. On T2-weighted imaging and three mpMRI combinations (T2-weighted imaging + diffusion-weighted imaging [DWI], T2-weighted imaging + dynamic contrast-enhanced-MRI [DCE-MRI], and T2-weighted imaging + DWI + DCE-MRI), two radiologists (R1/R2) scored the presence of a dominant tumour using a 5-point Likert scale (1 = definitely absent to 5 = definitely present). Diagnostic performance in identifying downgrading was evaluated via areas under the curves (AUCs). Predictive accuracies of multivariate models were calculated.
In predicting downgrading, T2-weighted imaging + DWI (AUC = 0.89/0.85 for R1/R2) performed significantly better than T2-weighted imaging alone (AUC = 0.72/0.73; p < 0.001/p = 0.02 for R1/R2), while T2-weighted imaging + DWI + DCE-MRI (AUC = 0.89/0.84 for R1/R2) performed no better than T2-weighted imaging + DWI (p = 0.48/p > 0.99 for R1/R2). On multivariate analysis, the clinical + mpMRI model incorporating T2-weighted imaging + DWI (AUC = 0.92/0.88 for R1/R2) predicted downgrading significantly better than the clinical model (AUC = 0.73; p < 0.001 for R1/R2).
mpMRI improves the ability to identify a subgroup of patients with Gleason 3 + 4 PCa on biopsy who are candidates for active surveillance. DCE-MRI (compared to T2 + DWI) offered no additional benefit to the prediction of downgrading.
Diagnostic performance of T2-weighted-imaging + DWI was better than T2-weighted-imaging alone. Diagnostic performance of T2-weighted-imaging + DWI was similar to T2-weighted-imaging + DWI + DCE-MRI. Combining clinical and T2-weighted-imaging + DWI features best predicted GS downgrading. mpMRI might prevent overtreatment by increasing eligibility for PCa active surveillance.
本研究旨在评估治疗前3特斯拉（3T）多参数磁共振成像（mpMRI）对活检确诊为Gleason评分（GS）3 + 4前列腺癌（PCa）患者根治性前列腺切除术（RP）后GS降级的预测诊断性能。
我们回顾性分析了304例活检证实为GS 3 + 4 PCa且在RP前接受mpMRI检查的患者。在T2加权成像和三种mpMRI组合（T2加权成像 + 扩散加权成像 [DWI]、T2加权成像 + 动态对比增强磁共振成像 [DCE-MRI]、T2加权成像 + DWI + DCE-MRI）上，两名放射科医生（R1/R2）使用5分李克特量表（1 = 肯定不存在至5 = 肯定存在）对优势肿瘤的存在情况进行评分。通过曲线下面积（AUC）评估识别降级的诊断性能。计算多变量模型的预测准确性。
在预测降级方面，T2加权成像 + DWI（R1/R2的AUC分别为0.89/0.85）的表现明显优于单独的T2加权成像（AUC分别为0.72/0.73；R1/R2的p < 0.001/p = 0.02），而T2加权成像 + DWI + DCE-MRI（R1/R2的AUC分别为0.89/0.84）的表现并不优于T2加权成像 + DWI（R1/R2的p = 0.48/p > 0.99）。多变量分析显示，纳入T2加权成像 + DWI的临床 + mpMRI模型（R1/R2的AUC分别为0.92/0.88）预测降级的能力明显优于临床模型（AUC = 0.73；R1/R2的p < 0.001）。
mpMRI提高了识别活检确诊为Gleason 3 + 4 PCa且适合进行主动监测的患者亚组的能力。DCE-MRI（与T2 + DWI相比）在预测降级方面没有额外益处。
T2加权成像 + DWI的诊断性能优于单独的T2加权成像。T2加权成像 + DWI的诊断性能与T2加权成像 + DWI + DCE-MRI相似。结合临床和T2加权成像 + DWI特征能最好地预测GS降级。mpMRI可能通过增加PCa主动监测的适用性来避免过度治疗。
