The diagnosis of muscle layer invasion in bladder cancer is of immense significance due to its direct impact on treatment decisions and prognostic outcomes. Local staging of bladder cancer is primarily based on pathology specimens obtained by TURBT. In some cases, a second TURBT is recommended when there is a T1 high-grade or when no muscular layer sample is present in the specimen at the first TURBT. Initial TURBT is sometimes inadequate, with residual tumors reported in 33%–76% of cases and up-staging in 0%–32% of cases [6]. In this study, second TURBT was performed in 54 out of 121 cases (44.6%). Residual tumor was discovered in 22 cases (39.3%), and up-staging occurred in two cases (3.7%). A second TURBT is effective for resection of residual tumor and obtaining a muscular layer sample but is not without the risk of bleeding and perforation. To mitigate the risk of potential complications of TURBT, imaging tools are recommended. CT can detect tumor invasion outside the bladder, but it cannot provide detailed images of the border between the tumor and the muscle layer. Therefore, CT cannot be used for local diagnosis of bladder cancer.

Since the VI-RADS was first reported in 2018, it has become a popular diagnostic method for predicting muscle invasion in bladder cancer. Several meta-analyses of the diagnostic performance of the VI-RADS have been reported [2]. In 2020, Woo et al. reported a high diagnostic performance of the VI-RADS with 83% sensitivity, 90% specificity, and the area under the hierarchical summary receiver operating characteristic (HSROC) curve was 0.94 [6].

VI-RADS is a system that predicts muscle layer invasion, with its scores reflecting the extent of invasion. In the study by Panebianco et al., no cut-off score was set [3]. The present study examined the diagnostic performance of a cut-off VI-RADS score as a predictive tool for MIBC and found the use of a cut-off score to enhance the usefulness of the VI-RADS in clinical practice.

In 2022, Del Giudice et al. reported a high diagnostic performance in their meta-analysis of the VI-RADS score; with a cut-off value of 3, it showed a sensitivity of 87%, a specificity of 86%, and an area under the HSROC curve of 0.93. A cut-off value of 4 showed a sensitivity of 78%, a specificity of 94%, and an area under the HSROC curve of 0.91 [2]. In comparison, this study showed a sensitivity of 89.3%, a specificity of 92.5%, a positive predictive value of 78.1%, and an AUC of 0.91. The diagnostic performance of VI-RADS in this study was slightly lower than that reported in other reports. The reasons for this are that the present study evaluated VI-RADS using 1.5 T-MRI and bp-MRI, and that the study population was smaller than that in other studies. [7]

Furthermore, in the present study, the VI≧4 group exhibited larger tumor diameters than the VI≦3 group. A large tumor diameter was identified as a preoperative factor associated with VI≧4. The relationship between tumor diameter and pathological findings is further discussed below.

Subsequently, by comparing the clinicopathological characteristics of each group, we retrospectively examined the characteristics of the VI≧4 group. This study revealed a higher prevalence of high-grade (G3) tumors in the VI≧4 group than in the VI≦3 group. The VI≧4 group was diagnostic for MIBC and suggested the possibility of predicting high-grade bladder cancer. Until recently, it was believed that low-grade bladder tumors progressed to high-grade tumors through the accumulation of genetic abnormalities [8, 9]. However, low-grade and high-grade tumors are genetically exclusive [10,11,12]. Therefore, low-grade tumors rarely become high-grade tumors. From this perspective, almost all cases in the VI≧4 group exhibited high-grade tumors (G3) in the results of the present study. The VI≧4 group had significantly more high-grade bladder tumors than the VI≦3 group, suggesting that the VI≧4 group could be associated with a diagnosis of high-grade tumors. Therefore, the VI≧4 cases should be considered for a treatment plan that includes total cystectomy and systemic treatment.

Regarding tumor grade, there were significantly more cases of urine cytology class≧4 in the VI≧4 group than in the VI≦3 group. Urine cytology is a pathological method used to diagnose atypia of urothelial exfoliated cells in the urine. Its sensitivity for low-grade NMIBC is low, while it is high for high-grade urothelial carcinoma [13]. In this study, we observed a higher prevalence of urine cytology class ≧4 in the VI≧4 group, which contains more high-grade tumors.

In relation to the tumor grade, there were significantly more cases of tumor necrosis in the VI≧4 group than in the VI≦3 group. In this case, the mean diameter of the MRI of the VI≧4 group was 31.8 mm (range: 10—59 mm, median: 30 mm). It was significantly larger than that of the VI≦3　group. Armin et al. suggested that tumors > 3 cm in diameter are high-grade tumors with necrosis [14]. This study suggests that patients in the VI≧4 group may be diagnosed with high-grade UC based on tumor size and necrosis.

The major pathological type of bladder cancer is UC; non-UC is considered rare, accounting for approximately 5% of all cases [15]. Further, urothelial bladder carcinoma has several subtypes. The presence of the UC variant is thought to be relatively rare, but recent reports have shown that it is present in approximately 20–25% of all cases [16, 17]. In this study, UC variants were identified in 16 out of 121 patients (11.6%) with high-grade tumors, and the VI≧4 group showed significantly more UC variants than the VI≦3 group.

The most common UC variant in this study was the lipid cell variant, which often occurs with high-grade UC or the micropapillary variants of UC [18]. Lopez et al. reported that the lipid cell variant is diagnosed at an advanced stage, with high grade and high mortality. [18] In this study, eight of 12 lipid cell variants cases were VI≧4, and nine of 12 cases were advanced stage (> pT2). All lipid cell variant cases were high-grade tumors, and three of 12 cases had tumor necrosis. Therefore, this study suggested that the VI≧4 group contains a considerable number of highly malignant lipid cell variants. Lopez et al. also suggest the potential presence of a lipid cell variant, which can pose diagnostic challenges. We believe that our study successfully diagnosed a latent lipid cell variant [18].

The micropapillary variant often occurs with lymph vascular invasion and is also diagnosed at an advanced stage. Kamat et al. report a 5-year overall survival rate of 54% and a 10-year survival rate of 27% for the micropapillary variant, which is a poor prognosis. To improve its prognosis, immediate total cystectomy is a viable option [19]. In this study, the micropapillary variant had lymphovascular invasion in two of four cases and VI≧4 in three of four cases. Because the VI≧4 group may include the micropapillary variant, lymph node metastasis must be carefully evaluated during diagnosis.

The microcystic variant is a relatively rare UC variant, and thus, its clinical characteristics and prognostic outcomes are controversial [20]. Lopez et al. concluded that there was no significant difference in prognosis when the microcystic variant was compared with conventional UC [21]. In this study, the microcystic variant cases were all NMIBC cases. However, three of four cases were high-grade tumors, and thus, this variant should be followed up for recurrence and metastasis.

Finally, two patients with the plasmacytoid variants were included in the VI≧4 group. The plasmacytoid variant has a poor prognosis and requires a treatment plan that includes surgery and chemotherapy [22]. Therefore, immediate treatment is recommended and possible if it can be predicted by preoperative MRI.

This study examined the possibility that the VI≧4 group may provide predictive insights on the clinicopathological features. The VI≧4 group had a higher incidence of MIBC and high-grade tumors. Furthermore, this study suggests that the VI≧4 group may have potentially high-grade bladder tumors with tumor necrosis and UC variants. The potential significance of the VI ≧4 group is that it has the added value of predicting not only MIBC but also some pathological features.