基于膀胱影像报告与数据系统评分的多参数磁共振成像对膀胱癌分级诊断的应用价值

Value of multi-parameter magnetic resonance imagin based on bladder imaging report and data system (VIRADS) score for bladder cancer, grading, diagnosis, and application

摘要:

目的 评估基于膀胱影像报告与数据系统（VIRADS）评分的多参数磁共振成像（MRI）对膀胱癌分级诊断的应用价值。方法 回顾性分析2018年3月—2020年12月宜宾市第一人民医院经病理确诊的120例膀胱癌患者临床资料。术前均接受T₂WI加权成像、动态增强成像（DCE）、扩散加权成像（DWI）多参数MRI检查，由2位影像科医生基于VIRADS系统对多参数MRI进行评分。以病理结果为金标准，分析VIRADS评分诊断膀胱癌分期准确率，采用Kappa检验分析VIRADS评分与病理结果的一致性；并比较不同VIRADS评分截断值对肌层浸润性膀胱癌的诊断效能，绘制受试者工作特征（ROC）曲线分析VIRADS评分诊断肌层浸润性膀胱癌的曲线下面积（AUC）。结果 病理结果显示，120例膀胱癌均为尿路上皮细胞癌，其中单发79例，多发41例；肌层浸润性膀胱癌42例，非肌层浸润性膀胱癌78例；分期分布：T_a期8例，T₁期70例，T₂期20例，T₃期14例，T₄期8例；将病理结果作为金标准，VIRADS评分诊断膀胱癌分期的准确率为90.83%（95% CI：0.760，0.976），与病理结果的一致性较好（κ=0.868）；VIRADS评分≥3分为截断值时诊断肌层浸润性膀胱癌的准确度（90.00%）比其他分值高（P <0.05），且约登指数最高，为0.780；VIRADS评分≥ 3分为截断值时诊断肌层浸润性膀胱癌的AUC=0.917（95% CI：0.854，0.976）。结论 基于VIRADS评分的多参数MRI在膀胱癌中具有较高的诊断效能，可准确评估膀胱癌分期，且VIRADS评分≥3分时诊断肌层浸润性膀胱癌准确度及约登指数最高。

Abstract:

Objective To evaluate the application value of multi-parameter magnetic resonance imaging (MRI) based on the bladder imaging report and data system (VIRADS) score in the diagnosis and grading of bladder cancer.Methods The clinical data of 120 patients with bladder cancer from March 2018 to January 2020 were retrospectively analyzed. The clinical data of 120 patients with bladder cancer diagnosed by pathology in Yibin First People's Hospital from March to January 2018 were analyzed retrospectively. All patients received T₂WI, DCE, and DWI before operation, and multi-parameter MRI was scored by two radiologists based on VIRADS system. Pathological results were used as the gold standard. The accuracy of VIRADS score in the staging of bladder cancer was analyzed, and the consistency between VIRADS score and pathological results was analyzed by Kappa Test. The area under the curve (AUC) of VIRADS score in the diagnosis of muscularis invasive bladder cancer was analyzed by ROC.Results Pathological results showed that 120 cases of bladder cancer were all urothelial cell carcinomas, of which 79 cases were single and 41 cases were multiple; 42 cases were muscular invasive bladder cancer, 78 cases were non-muscular invasive bladder cancer; staging distribution: Ta stage 8 cases, 70 cases in T₁ stage, 20 cases in T₂ stage, 14 cases in T₃ stage, and 8 cases in T₄ stage; Using pathological results as the gold standard, the diagnostic accuracy of VIRADS was 90.83% (95% CI: 76.0%, 97.6% ) , which was in good agreement with pathological results (κ = 0.868); when VIRADS score ≥ 3 is divided into cut-off value, the accuracy of diagnosing muscle-invasive bladder cancer (90.00%) is higher than other scores (P < 0.05), and the Youden index was the highest (0.780); when ROC was plotted, the AUC of VIRADS score ≥ 3 was divided into cut-off values for diagnosis of muscular invasive bladder cancer was 0.917 (95% CI: 0.854, 0.976).Conclusion Multi-parameter MRI based on VIRADS score has high diagnostic efficiency in bladder cancer, and can accurately assess the staging of bladder cancer, and the diagnosis of myometrial invasive bladder cancer has the highest accuracy and Youden index when VIRADS score ≥ 3.