We investigated the feasibility of magnetic resonance elastography (MRE) using a pelvic acoustic driver for the detection and classification of prostate cancer (PCa). A total of 75 consecutive patients (mean age, 70; range, 56–86) suspected of having PCa and who underwent multi-parametric MRI including MRE and subsequent surgical resection were included. The analyzed regions consisted of cancer (n = 69), benign prostatic hyperplasia (BPH) (n = 70), and normal parenchyma (n = 70). A histopathologic topographic map served as the reference standard for each region. One radiologist and one pathologist performed radiologic–pathologic correlation, and the radiologist measured stiffness values in each region of interest on elastograms automatically generated by dedicated software. Paired t-tests were used to compare stiffness values between two regions. ROC curve analysis was also used to extract a cutoff value between two regions. The stiffness value of PCa (unit, kilopascal (kPa); 4.9 ± 1.1) was significantly different to that of normal parenchyma (3.6 ± 0.3,p< 0.0001) and BPH (4.5 ± 1.4,p= 0.0454). Under a cutoff value of 4.2 kPa, a maximum accuracy of 87% was estimated, with a sensitivity of 73%, a specificity of 99%, and an AUC of 0.839 for discriminating PCa from normal parenchyma. Between PCa and BPH, a maximum accuracy of 62%, a sensitivity of 70%, a specificity of 56%, and an AUC of 0.598 were estimated at a 4.5 kPa cutoff. The stiffness values tended to increase as the ISUP grade increased. In conclusion, it is feasible to detect and classify PCa using pelvic MRE. Our observations suggest that MRE could be a supplement to multi-parametric MRI for PCa detection.
本研究探讨了利用盆腔声学驱动器进行磁共振弹性成像（MRE）在前列腺癌（PCa）检测与分类中的可行性。研究连续纳入75例疑似前列腺癌患者（平均年龄70岁，范围56-86岁），所有患者均接受了包含MRE的多参数磁共振成像检查并随后接受手术切除。分析区域包括癌变组织（n=69）、良性前列腺增生（BPH）（n=70）及正常腺体组织（n=70）。以组织病理学拓扑图作为各区域的参考标准。由一名放射科医师和一名病理医师进行影像-病理对照分析，放射科医师使用专用软件自动生成的弹性图测量各感兴趣区的硬度值。采用配对t检验比较两个区域间的硬度值差异，并通过ROC曲线分析提取区域间的临界值。前列腺癌组织的硬度值（单位：千帕kPa；4.9±1.1）与正常腺体组织（3.6±0.3，p<0.0001）及良性前列腺增生组织（4.5±1.4，p=0.0454）存在显著差异。以4.2 kPa为临界值时，鉴别前列腺癌与正常腺体组织的最大准确率达87%，敏感性73%，特异性99%，曲线下面积（AUC）为0.839。在前列腺癌与良性前列腺增生的鉴别中，以4.5 kPa为临界值可获得最大准确率62%，敏感性70%，特异性56%，AUC为0.598。硬度值随国际泌尿病理学会（ISUP）分级升高呈递增趋势。结论表明，盆腔MRE用于前列腺癌的检测与分类具有可行性。本研究提示MRE可作为多参数磁共振成像在前列腺癌检测中的补充手段。
Magnetic Resonance Elastography for the Detection and Classification of Prostate Cancer
肿瘤(癌症)患者之家