VI-RADS-based algorithm for bladder cancer management randomized retrospective study

Abstract

Objective To evaluate if VI-RADS can distinguish between nonmuscle-invasive bladder cancers (NMIBC), muscle-invasive bladder cancer (MIBC), and high-risk nonmuscle-invasive bladder cancers (HR-NMIBCs). It is unclear if the Vesical Imaging-Reporting and Data System (VI-RADS) can replace repeated transurethral resection of bladder tumor (Re-TURBT) as in the new VI-RADS-based algorithm. Methods Sensitivity, specificity, and positive (PPV) and negative (NPV) predictive values of the VI-RADS score were calculated for mpMRI performance in patients undergoing TURBT and HR-NMIBC patients for only Re-TURBT. Results Of 283 cases, when VI-RADS ≥3 lesions were considered muscle-invasive, its sensitivity was 95.7% and specificity was 92.5%. PPV and NPV were 86.6% and 97.7%, respectively. The area under the curve (AUC) was 0.942 (P <.001). Of 89 patients undergoing post-Re-TURBT, 41 (46%) were tumor-free, 47 (50.5%) showed permanent HR-NMIBC, and 3 (2.2%) were upgraded to MIBC. Per the new VI-RADS-based approach, 73 (41%) of the 178 HR-NMIBCs with VI-RADS ≤2 would not undergo Re-TURBT. Of the 75 patients with VI-RADS ≥4, 6 (6) with HR-NMIBCs (8%) would not undergo Re-TURBT. When incomplete resections were excluded, 35 (60.3%) of the patients had complete resection, 23 (39.7%) had residual disease, and complete resection would not have been performed in these patients, and 2 (100%) still had residual disease. Conclusion The new VI-RADS-based algorithm helped VI-RADS ≥4 patients by switching to radical treatment. Since the residual disease is high in cases with VI-RADS ≤2, even if incomplete resections are excluded, TURBT should be continued.