Validation of vesical imaging-reporting and data system (VI-RADS) and its use in clinical management of bladder cancer

ABSTRACT In our study, we aimed to investigate the success of the VI-RADS scoring system in predicting muscle invasion in bladder cancer. Additionally, we aimed to explore the relationships between clinical data such as tumor location, diameter, localization, and lymphovascular invasion, and radiological data such as the presence of hydronephrosis and lymph node involvement, with the VI-RADS scoring system and muscle invasion. After obtaining ethical approval, patients who presented to our clinic between January 2023 and June 2024 with suspected bladder tumors, underwent diagnostic cystoscopy and MR imaging, and met inclusion criteria without exclusion criteria were included in the study. As a result, the outcomes of 74 patients were analyzed. Demographic data such as patients' names, file numbers, and genders, cystoscopy data such as the number, localization, morphology, and diameter of tumors, TUR-M or radical cystectomy pathology data and indications, and radiological data such as hydronephrosis and lymph node involvement were recorded. All these data were compared with patients' MpMRI data. In the VI-RADS> 3 group, compared to the VI-RADS≤ 3 group, the rates of incomplete resection, the presence of tumors in the trigone, tumor size, tumor stage and grade, and the presence of lymphovascular invasion were significantly higher (all p< 0.001). In the VI-RADS> 3 group, compared to the VI-RADS≤ 3 group, the distances of tumor contact, the largest tumor diameter, lymph node invasion, the presence of hydroureteronephrosis, and the rate of non-exophytic morphology of tumors were significantly higher (all p< 0.001). A statistically significant relationship was found between MRI lymph node involvement and TUR-M lymphovascular invasion, and between MRI lymph node involvement and radical cystectomy pathological lymph node invasion (p<0.001 and p = 0.015, respectively). A statistically significant relationship was found between the pathological T stage of TUR-M and the VI-RADS score (p< 0.001). In the logistic regression analysis conducted to predict the presence of muscle invasion based on the TUR-M pathology results, the model predicting muscle tissue invasion was significant (χ2(8) = 1.9, p = 0.983) and could explain 84% of the variance in reencapsulation (Nagelkerke R2 = 0.840). The model correctly predicted 96.4% of those without muscle invasion and 89.5% of those with muscle invasion (total 94.6%). The most important predictor for muscle invasion was found to be a high VI-RADS score. In the correlation analysis conducted between clinical and radiological parameters, a high level of positive correlation was observed between the VI-RADS score and cystoscopy tumor diameter, MRI tumor contact length, MRI tumor diameter, and MRI tumor axial largest diameter, and this correlation was found to be statistically significant (r = 0.720 p < 0.001, r = 0.811 p < 0.001, r = 0.724 p < 0.001, r = 0.800 p < 0.001, respectively). In our study, we concluded that the VI-RADS scoring system could strongly predict muscle invasion in bladder cancer. Additionally, we identified clinical and radiological factors affecting the VI-RADS score and muscle invasion, presenting the methods of using the VI-RADS scoring system in the clinical management of bladder cancer. Further studies with larger patient series are needed on this subject. Keywords: Muscle invasion, Bladder cancer, VI-RADS