Department of Diagnostic and Interventional Radiology, Eberhard-Karls-University, Hoppe-Seyler-Str.3, 72076 Tübingen, Germany.
Department of Urology, Eberhard-Karls-University, Hoppe-Seyler-Str.3, 72076 Tübingen, Germany.
Eur J Radiol. 2022 Jan;146:110059. doi: 10.1016/j.ejrad.2021.110059. Epub 2021 Nov 20.
To assess the clinical applicability of local tumor staging in urinary bladder cancer (BC) with preoperative multiparametric MRI (mpMRI) using the five-point Vesical Imaging-Reporting and Data System (VI-RADS) scoring system and to compare it to dual-phase contrast-enhanced computed tomography (CECT).
33 patients with primary untreated bladder cancer underwent CECT followed by preoperative multiparametric 3.0 T MRI between July 2019 and August 2020 and were enrolled in this retrospective study. Two radiologists initially performed staging on the CECT image data sets and - blinded to CT results - on subsequent mpMRI. BCs were staged according to the VI-RADS scoring system. Postoperative pathology was correlated to the VI-RADS score and the CECT results. The performance of VI-RADS in determining detrusor muscle invasion was analyzed using a receiver operating characteristic curve. Based on the histopathology, sensitivity, specificity and accuracy for muscle invasiveness between both image modalities were compared using the Chi square test.
A total of 33 patients (29 male, median age 70 years, IQR: 59-81 years) were included. 10 tumors were categorized as non-muscle invasive (30%) and 23 as muscle invasive BC (70%) in final histology. Tumor stages were correctly assigned as being either muscle invasive or non-muscle invasive on both CECT and mpMRI with regard to both early and late stages of BC (Ta-Tis and T3a-T4b). T-stages bordering the histopathologic limits of muscle invasiveness (T1-T2a-b) resulted in overestimation of muscle invasion in 43% of cases (VI-RADS 3-4) for the mpMRI image data sets and in an underestimation of muscle invasion in up to 55.5% of cases analysing the CECT data. Sensitivity and specificity for the determination of muscle invasion in CECT and mpMRI were 80%/80% and 74%/61% for Radiologist#1 and 70%/90% and 83%/70% for Radiologist#2, respectively.
There are advantages and disadvantages of both CECT and mpMRI when used in the clinical assessment of BC muscular tumor invasion. In borderline cases, only the combination of cross-sectional imaging and histopathological staging may help in making the optimal treatment decisions.
使用五点膀胱成像报告和数据系统 (VI-RADS) 评分系统评估术前多参数 MRI (mpMRI) 在膀胱癌 (BC) 中的局部肿瘤分期的临床适用性，并将其与双期对比增强 CT (CECT) 进行比较。
2019 年 7 月至 2020 年 8 月期间，33 例原发性未经治疗的膀胱癌患者接受了 CECT 检查，随后进行了术前 3.0 T MRI 检查，这些患者被纳入本回顾性研究。两位放射科医生最初对 CECT 图像数据集进行分期，然后在随后的 mpMRI 上进行分期 - 对 CT 结果不知情。BC 根据 VI-RADS 评分系统进行分期。术后病理与 VI-RADS 评分和 CECT 结果相关。使用受试者工作特征曲线分析 VI-RADS 对逼尿肌浸润的诊断性能。基于组织病理学，使用卡方检验比较两种影像学方式对肌肉侵袭性的敏感性、特异性和准确性。
共纳入 33 例患者（29 例男性，中位年龄 70 岁，IQR：59-81 岁）。10 例肿瘤在最终组织学中被归类为非肌肉浸润性（30%），23 例为肌肉浸润性 BC（70%）。对于 BC 的早期和晚期，CECT 和 mpMRI 均正确地将肿瘤分期归类为肌肉浸润性或非肌肉浸润性（Ta-Tis 和 T3a-T4b）。在组织病理学上逼尿肌浸润的边界（T1-T2a-b）处的 T 分期导致 mpMRI 图像数据集中肌肉浸润的高估，占 43%（VI-RADS 3-4），而 CECT 数据分析中肌肉浸润的低估高达 55.5%。CECT 和 mpMRI 对肌肉浸润的确定的敏感性和特异性分别为放射科医生#1 的 80%/80%和 74%/61%，以及放射科医生#2 的 70%/90%和 83%/70%。
CECT 和 mpMRI 在膀胱癌肌肉肿瘤浸润的临床评估中各有优缺点。在边界病例中，只有横断面成像和组织病理学分期的结合才能帮助做出最佳治疗决策。
