Evaluating biparametric MRI for diagnosing muscle-invasive bladder cancer with variant urothelial histology: a multicenter study.

Abstract

Background: Vesical Imaging-Reporting and Data System (VI-RADS) based on multiparametric MRI (mp-MRI) demonstrated excellent performance in diagnosing muscle-invasive bladder cancer (MIBC) in cases of pure urothelial carcinoma. However, the performance of VI-RADS based on mp-MRI and biparametric MRI (bp-MRI) in diagnosing urothelial carcinoma with variant histology (VUC) remains unknown.

Purpose: To evaluate the applicability of VI-RADS using mp-MRI and bp-MRI in diagnosing MIBC in patients with VUC.

Methods: A retrospective analysis was conducted on 86 patients with VUC from different medical centers. Each patient underwent mp-MRI, with images evaluated using VI-RADS scores. The acquired images were divided into two groups: the mp-MRI group and the bp-MRI group. The mp-MRI group was evaluated according to the VI-RADS protocol. For the bp-MRI group, two VI-RADS scoring criteria were established: bp-DWI, primarily driven by DWI, and bp-T2WI, primarily driven by T2WI. The bp-MRI group was evaluated based on these two criteria. Inter-reader agreement performance was evaluated using Kappa analysis. The evaluation methods were evaluated by receiver operating characteristic curve. Comparison of the area under the curve (AUC) was performed used DeLong's test. A p-value < 0.05 was considered significant.

Results: Inter-reader agreement was high across all evaluation methods, with Kappa values exceeding 0.80. The AUCs for mp-MRI, bp-DWI, and bp-T2WI were 0.934, 0.885, and 0.932, respectively. The diagnostic performance of bp-T2WI was comparable with that of mp-MRI (p = 0.682) and significantly higher than bp-DWI (p = 0.007). Both mp-MRI and bp-T2WI demonstrated high sensitivity and specificity.

Conclusion: VI-RADS based on mp-MRI demonstrates good diagnostic performance for MIBC in VUC patients. bp-T2WI may provide comparable diagnostic performance to mp-MRI.