DOI: 10.4103/iju.iju_128_26 ISSN: 0970-1591

The diagnostic value of biparametric magnetic resonance imaging vesical imaging-reporting and data system combined with tumor contact length for assessing muscle invasion in bladder cancer lesions scored vesical imaging-reporting and data system 3: A

Wenbin Bao, Yangwenyi Liu, Jincui Yang, Xuebing Ma
Show PDF Cite

Objective:

To explore the diagnostic value of biparametric magnetic resonance imaging (bp-MRI) combined with tumor contact length (TCL) for identifying muscle-invasive bladder cancer (MIBC) among lesions scored 3 on the vesical imaging-reporting and data system (VI-RADS).

Materials and Methods:

This single-center retrospective study enrolled patients with pathologically confirmed bladder urothelial carcinoma and a preoperative VI-RADS overall score of 3 (October 2022–October 2025). Two radiologists independently assessed biparametric VI-RADS score and TCL. Inter-observer agreement was evaluated. Based on the postoperative pathology, patients were divided into nonmuscle-invasive bladder cancer (NMIBC) and MIBC groups. Diagnostic performance of individual indicators and combined models (logistic regression and clinical rule) was assessed using receiver operating characteristic analysis. Bootstrap validation (1000 resamples) was performed.

Results:

One hundred and three patients (46 NMIBC, 57 MIBC) were included. Inter-observer agreement was excellent for biparametric VI-RADS (weighted Kappa = 0.82) and TCL (intraclass correlation coefficient = 0.91). TCL and biparametric score was significantly higher in the MIBC group (both P < 0.001). TCL alone (cutoff >2.5 cm) yielded an area under the curve (AUC) of 0.708; for the biparametric score alone (cutoff >3), AUC was 0.743. The combined logistic model (bp-MRI + TCL) achieved the highest AUC of 0.821 (optimism-corrected AUC of 0.798), sensitivity of 96.5%, and specificity of 65.2%. Proposed clinical rule (downgrade biparametric score 3-2 if TCL ≤2.5 cm) achieved an AUC of 0.814 (sensitivity 91.2% and specificity 69.6%). Multiparametric model (including dynamic contrast-enhanced) was not superior (AUC 0.810, P > 0.05).

Conclusion:

Integrating TCL with biparametric VI-RADS score improved the diagnostic accuracy for muscle invasion in VI-RADS 3 lesions. The simple clinical rule (TCL ≤2.5 cm downgrades the score to 2) showed promising efficacy and may assist in preliminary decision-making, pending external validation.

More from our Archive