Abstract / Summary
Background/Objectives: We developed and internally validated the Bladder Risk Index using Clinical and Systemic biomarkers (BRICS), a preoperative prediction model designed to estimate the likelihood of high-grade (HG) disease and muscle-invasive bladder cancer (MIBC) in patients undergoing their first transurethral resection of bladder tumor (TURBT). Methods: This retrospective single-center study included 161 patients with newly diagnosed urothelial bladder cancer. The BRICS-HG model included tumor size, neutrophil-to-lymphocyte ratio (NLR), and creatinine-to-hemoglobin ratio (CHR), while the BRICS-MIBC model was based on tumor size and NLR. Model performance was assessed in terms of discrimination, calibration, and Brier score, together with exploratory decision curve analysis. Internal validation was performed using 2000 bootstrap resamples. Results: MIBC was identified in 20 patients (12.4%). BRICS-HG demonstrated limited discriminatory ability, with an apparent AUC of 0.649 and an optimism-corrected AUC of 0.623 (95% bootstrap interval: 0.541–0.703). BRICS-MIBC achieved moderate discrimination, with an apparent AUC of 0.735 and an optimism-corrected AUC of 0.709 (95% bootstrap interval: 0.588–0.826). Across the bootstrap resamples, the model combining tumor size and NLR was selected in 53.9% of cases, reflecting uncertainty in predictor selection. Conclusions: BRICS-HG should be regarded as an exploratory model, and its current performance does not support clinical application. BRICS-MIBC demonstrated moderate discrimination after internal validation; however, its performance should be interpreted cautiously given the small number of MIBC events. Validation in larger, independent cohorts, ideally across multiple centers, is necessary before the model can be considered for clinical implementation.