VI-RADS评分在TURBT术后的表现:探讨修改的必要性
这项回顾性单中心研究纳入123例膀胱癌患者,评估TURBT术后MRI的VI-RADS评分性能。以再TURBT或膀胱切除病理为参考,DWI主导评分AUC达0.86-0.95,准确率86%-96%。TURBT后≥2-4周检查且图像质量达标可提高可靠性。
要点
- DWI主导VI-RADS评分在TURBT术后诊断性能数值上更高
- TURBT后2周内检查准确率最低,2-4周明显改善
- 图像质量评分提高可改善VI-RADS诊断准确性
核心结论
- TURBT术后膀胱MRI在考虑检查时机、序列权重和图像质量时仍保持准确。
- DWI主导VI-RADS评分显示数值上更高的诊断估计值。
研究设计与样本
- 回顾性单中心研究,最终纳入123例膀胱癌患者,TURBT后行MRI。
- 三名放射科医生分别分配标准、DWI主导和DCE主导VI-RADS评分,以≥3为截断值。
- 以再TURBT或膀胱切除组织病理为参考标准,按时间间隔和VI-RADS质量评分分层分析。
主要发现
- 标准和DCE主导VI-RADS结果相同:AUC=0.81,敏感性95%,特异性67%,准确性77%。
- DWI主导VI-RADS的AUC为0.86-0.95,准确性86%-96%(跨读者)。
- 读者间一致性优秀:总体Fleiss κ=0.91-0.96,成对Cohen κ最高0.98。
- 诊断准确性随时间间隔变化,2周内最低(特异性13%,准确性26%),2-4周改善(准确性94-98%)。
- 准确性随VI-RADS质量评分提高而改善。
关键数值与推荐
- VI-RADS评分截断值为≥3。
- DWI主导VI-RADS的AUC范围0.86-0.95,准确性86%-96%。
- 标准/DCE主导VI-RADS:AUC=0.81,敏感性95%,特异性67%,准确性77%。
- 2周内检查特异性13%、准确性26%;2-4周准确性94-98%。
临床怎么用
- 建议TURBT后≥2-4周再行膀胱MRI检查以提高诊断可靠性。
- 采用DWI主导VI-RADS评分可能获得更高的诊断准确性。
- 确保图像质量达标(质量评分高)以改善VI-RADS评分可靠性。
局限性
- 回顾性研究设计。
- 单中心研究。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND: The Vesical Imaging-Reporting and Data System (VI-RADS) was introduced to standardize bladder magnetic resonance imaging (MRI) interpretation, mainly in treatment-naïve patients. However, most patients undergo MRI after transurethral resection of bladder tumor (TURBT), where postprocedural changes compromise diagnostic accuracy. This study aimed to identify the dominant sequence for post-TURBT VI-RADS scoring, define optimal MRI timing, and evaluate the effect of image quality. METHODS: This retrospective single-center study included patients with bladder cancer who underwent TURBT followed by MRI. Three radiologists assigned standard, diffusion-weighted imaging (DWI)-dominant, and dynamic contrast-enhanced (DCE)-dominant VI-RADS scores using a cutoff of ≥3. Histopathology from re-TURBT or cystectomy served as the reference standard. Diagnostic performance was assessed with receiver operating characteristic analysis and stratified by time interval (<2 wk, 2-4 wk, >4 wk) and VI-RADS Quality Score (QS). Inter-reader agreement was evaluated using pairwise Cohen's κ and overall Fleiss' κ. RESULTS: The final cohort comprised 123 patients. Standard and DCE-dominant VI-RADS produced identical results (area under the curve [AUC] = 0.81; sensitivity = 95%; specificity = 67%; accuracy = 77%). For DWI-dominant VI-RADS, AUC ranged from 0.86 to 0.95 and accuracy from 86% to 96% across readers. Inter-reader agreement was excellent (overall Fleiss' κ = 0.91-0.96; pairwise Cohen's κ = up to 0.98). Diagnostic accuracy varied by time interval, reaching its lowest within 2 wk (specificity = 13%, accuracy = 26%) and improving at 2-4 wk (accuracy = 94-98%). Accuracy also improved with increasing QS. The study is limited by its retrospective design and single-center setting. CONCLUSION: Bladder MRI after TURBT remains accurate when timing, sequence weighting, and image quality are considered. DWI-dominant VI-RADS shows numerically higher diagnostic estimates. Scheduling MRI ≥2-4 wk after TURBT and ensuring adequate image quality may improve reliability.
引用
- Dehghanpour A, Pecoraro M, Zhang GM, et al. VI-RADS Score Performance in the Post-TURBT Setting: Exploring the Need for Modification. Eur Urol Oncol. 2026. PMID: 42469151. doi:10.1016/j.euo.2026.05.012
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。