Correlation of Multiparametric MRI-estimated Vesical Imaging Reporting and Data System (VI-RADS) Score with Muscle Invasion and Tumor Grade on TURBT in Patients with Urothelial Bladder Cancer: A Prospective Cross-sectional Study

Abstract

Introduction and Objective: Differentiation between non-muscle-invasive bladder cancer (NMIBC) and muscle-invasive BC (MIBC) is one of the main challenges in BC management. Tumor grade is another parameter guiding treatment. Transurethral resection of bladder tumor (TURBT) bears an inherent risk of understaging, especially if the sample does not contain detrusor muscle. Hence, an adjunctive modality for local staging such as CT or MRI is desirable. Multi-parametric MRI (mpMRI), and the Vesical Imaging Reporting and Data System (VI-RADS) score derived therefrom is one such tool for BC local staging. The objective of the present study was to prospectively evaluate the accuracy the VI-RADS score in assessing muscle-invasiveness and tumor grade on TURBT, by using histopathological examination as the reference standard. Methods: 50 patients with bladder tumor were included in this study conducted at the Department of Urology, Gauhati Medical College and Hospital, Guwahati, Assam (India) over 13 months from November 2020 to November 2021. All patients underwent a mpMRI of the urinary bladder prior to TURBT and a five-point VI-RADS score was obtained. The VI-RADS score was correlated with the histopathological muscle invasion and tumor grade obtained on TURBT. Results: There was a positive correlation between the VI-RADS score and the pathological results with respect to both muscle invasion (r = 0.401, p < 0.05) and tumor grade (r = 0.386, p < 0.05). For muscle-invasion, the AUC of the VI-RADS score was 0.722, and for tumor grade, the AUC was 0.709. For a VI-RADS score of ≥ 3, the sensitivity and specificity to predict muscle-invasiveness were 84% and 44% respectively, and for a VI-RADS score of ≥ 4, the sensitivity and specificity to predict muscle-invasiveness were 68% and 72% respectively. Conclusion: The reliability of VI-RADS for BC local staging and discrimination between NMIBC and MIBC has been widely reported. However, the lack of accuracy of correlation in the present study and the lack of identification of important prognostic markers for BC on MRI, such as carcinoma in-situ, lymphovascular invasion and variant histology relegates the VI-RADS score to being only an adjunct in the workup of patients with BC.