Department of Urology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu, China.
Department of Radiology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu, China.
Int J Urol. 2022 Mar;29(3):186-195. doi: 10.1111/iju.14748. Epub 2021 Dec 19.
The objective of this study is to systematically evaluate the diagnostic performance of the Vesical Imaging-Reporting and Data System for predicting muscle-invasive bladder cancer. Embase, PubMed and Web of Science were systematically searched from 1 September 2018 to 30 July 2021 to include proper studies. We included studies that included data on Vesical Imaging-Reporting and Data System and their associated pathological findings, and we assessed their quality using the Quality Assessment of Diagnostic Accuracy Studies-2 tool. The pooled sensitivity and specificity were calculated and plotted using hierarchical summary receiver operating characterisijutic modeling. Meta-regression analysis was carried out to detect heterogeneity. A total of 20 studies with 2725 patients were included. When the cut-off point was 3, the pooled sensitivity and specificity were 0.92 (0.89-0.94) and 0.85 (0.78-0.90), respectively, and 0.82 (0.75-0.88) and 0.95 (0.91-0.97), respectively, when the cut-off point was 4. The area under the curve was 0.95 and 0.95, respectively. Heterogeneity was substantially considerable in sensitivity and specificity. All subgroup variables, including patient number, study design, magnetic resonance imaging field strength, number of radiologists, surgery pattern, diffusion-weighted imaging, and dynamic contrast-enhanced magnetic resonance imaging, contributed to sensitivity heterogeneity when the cut-off point was 3 and specificity heterogeneity when the cut-off point was 4. Multiple image acquisition plane of diffusion-weighted imaging achieved a higher sensitivity than single image acquisition plane of diffusion-weighted imaging in both the Vesical Imaging-Reporting and Data System 3 and 4 groups, and higher specificity in the Vesical Imaging-Reporting and Data System 4 group. Another significant source of heterogeneity was the cut-off point. The diagnostic performance of the Vesical Imaging-Reporting and Data System for predicting muscle-invasive bladder cancer was excellent in both cut-off points of the Vesical Imaging-Reporting and Data System 3 and 4. Multiple image acquisition planes of diffusion-weighted imaging should be given more attention in the Vesical Imaging-Reporting and Data System.
本研究旨在系统评估膀胱影像学报告和数据系统（Vesical Imaging-Reporting and Data System，VI-RADS）预测肌层浸润性膀胱癌的诊断性能。我们系统性检索了 Embase、PubMed 和 Web of Science 数据库，检索时间从 2018 年 9 月 1 日至 2021 年 7 月 30 日，以纳入合适的研究。我们纳入了包含 VI-RADS 及其相关病理结果数据的研究，并使用诊断准确性研究质量评估工具（Quality Assessment of Diagnostic Accuracy Studies-2）评估其质量。使用分层汇总受试者工作特征建模计算并绘制合并敏感性和特异性。进行 Meta 回归分析以检测异质性。共纳入 20 项研究，包含 2725 例患者。当截断值为 3 时，合并敏感性和特异性分别为 0.92（0.89-0.94）和 0.85（0.78-0.90），当截断值为 4 时，分别为 0.82（0.75-0.88）和 0.95（0.91-0.97）。曲线下面积分别为 0.95 和 0.95。敏感性和特异性的异质性均较大。包括患者数量、研究设计、磁共振成像场强、放射科医生数量、手术方式、弥散加权成像、动态对比增强磁共振成像等所有亚组变量，当截断值为 3 时，对敏感性异质性有影响，当截断值为 4 时，对特异性异质性有影响。在 VI-RADS 3 组和 4 组中，扩散加权成像的多图像采集平面比单图像采集平面的敏感性更高，而在 VI-RADS 4 组中，特异性更高。另一个显著的异质性来源是截断值。VI-RADS 预测肌层浸润性膀胱癌的诊断性能在 VI-RADS 3 和 4 的截断值下均表现出色。在 VI-RADS 中应更加重视扩散加权成像的多图像采集平面。
