Clinical value of VI-RADS score in predicting repeat transurethral resection in primary bladder cancer: a prospective observational study.

内容摘要

Repeat transurethral resection of bladder tumor (TURBT) is recommended in selected patients with newly diagnosed bladder cancer to improve staging accuracy and detect residual disease, but it also increases morbidity and treatment burden. The Vesical Imaging-Reporting and Data System (VI-RADS) has shown high accuracy for predicting muscle invasion, whereas its role in identifying patients who undergo repeat TURBT remains unclear. This study evaluated the clinical value of VI-RADS for predicting repeat TURBT status in primary bladder cancer. This single-center prospective observational study included 59 patients with newly diagnosed primary bladder tumors who underwent preoperative multiparametric magnetic resonance imaging and initial TURBT. VI-RADS scores were assigned by an experienced radiologist blinded to pathological findings. Associations between VI-RADS and clinicopathological variables were examined using Spearman's correlation analysis. Diagnostic performance for predicting repeat TURBT status was assessed by receiver operating characteristic analysis. Univariable and multivariable logistic regression models were used to identify independent predictors of undergoing repeat TURBT. Repeat TURBT was performed in 31 patients (52.5%). Higher VI-RADS categories were more frequently observed among patients who underwent repeat TURBT, whereas lower categories predominated in the single-TURBT group. Among patients undergoing repeat TURBT, residual malignancy was identified in 66.7% of VI-RADS 4 lesions and 88.9% of VI-RADS 5 lesions. VI-RADS correlated significantly with tumor grade, invasion depth, tumor size, and number of surgical procedures. ROC analysis showed modest discriminatory ability for predicting repeat TURBT status (AUC 0.635, 95% CI 0.491-0.778; p = 0.066). At a cutoff of VI-RADS ≥ 3, sensitivity was 70.4% and specificity was 59.4%. In multivariable analysis, only invasion depth at initial TURBT independently predicted repeat TURBT. VI-RADS was associated with pathological tumor severity and with repeat TURBT status, but it did not independently predict undergoing repeat TURBT. These findings support VI-RADS as an adjunctive preoperative tool rather than a stand-alone determinant for repeat resection decisions.