| Abstract: |
Background: The diagnostic performance of contrast medium–free biparametric magnetic resonance imaging (bpMRI; combining T2-weighted imaging [T2WI] and diffusion-weighted imaging [DWI]) for evaluating variant-histology urothelial carcinoma (VUC) remains unknown. Objective: To compare the diagnostic performance of bpMRI and multiparametric MRI (mpMRI; combining T2WI, DWI, and dynamic contrast-enhanced MRI]) for assessing muscle invasion of VUC. Design, setting, and participants: This multi-institution retrospective analysis included 118 patients with pathologically verified VUC who underwent bladder mpMRI before transurethral bladder tumor resection between 2010 and 2019. Outcome measurements and statistical analysis: Three board-certified radiologists separately evaluated two sets of images, set 1 (bpMRI) and set 2 (mpMRI), in accordance with the Vesical Imaging Reporting and Data System (VI-RADS). The histopathology results were utilized as a reference standard. Receiver operating characteristic curve analysis, Z test, and Wald test were used to assess diagnostic abilities. Results and limitations: Sixty-six (55.9%) and 52 (44.1%) of the 118 patients with VUC included in the analysis (mean age, 71 ± 10 yr; 88 men) had muscle-invasive bladder cancer (MIBC) and non-MIBC, respectively. For the diagnosis of MIBC, the areas under the curve for bpMRI were significantly smaller than those for mpMRI (0.870–0.884 vs 0.902–0.923, p < 0.05). The sensitivity of bpMRI was significantly lower than that of mpMRI for all readers with a VI-RADS cutoff score of 4 (65.2–66.7% vs 77.3–80.3%, p < 0.05). The specificity of bpMRI and mpMRI did not differ significantly for all readers (88.5–90.4 vs 88.5–92.3, p > 0.05). A limitation of the study is the limited sample size because of the rarity of VUC. Conclusions: In patients with VUC, on applying VI-RADS, the diagnostic results of bpMRI were inferior to those of mpMRI for evaluating muscle invasion. Therefore, mpMRI-based methods are recommended for evaluating muscle invasiveness of VUC. Patient summary: Contrast medium–free biparametric magnetic resonance imaging (bpMRI)-based Vesical Imaging Reporting and Data System (VI-RADS) can accurately diagnose pure urothelial carcinomas, similar to conventional multiparametric magnetic resonance imaging–based VI-RADS. However, bpMRI-based VI-RADS may misdiagnose muscle invasiveness of urothelial carcinoma with variant histology, particularly when its cutoff score is 4. © 2023 European Association of Urology |
