Academic In-depth Report
Editor’s Note: During a recent session on urological oncology, Professor A. Ślusarczyk from the Department of General, Oncological, and Functional Urology at the Medical University of Warsaw, Poland, presented the results of an international multicenter retrospective study. The study aimed to develop and validate a clinical predictive model to identify patients with high-grade non-muscle-invasive bladder cancer (NMIBC) who are at risk of BCG-unresponsiveness or disease progression following treatment.
01 Research Design and Cohort Characteristics
The study analyzed data from 3,800 patients across 13 European centers. After screening, a total of 2,211 BCG-naive patients with high-grade Ta or T1 NMIBC (with or without concomitant CIS) were enrolled. All participants underwent complete transurethral resection of bladder tumor (TURBT), followed by a re-TURBT where indicated. All patients received adequate BCG induction and maintenance therapy. The median duration of BCG maintenance was one year (equivalent to three maintenance courses).
02 Identification of Five Core Risk Factors
The research team utilized the Cox proportional hazards model to develop a clinical risk score. Multivariate analysis identified five independent risk factors significantly associated with the primary endpoint—defined as BCG-unresponsiveness (per International Bladder Cancer Group [IBCG] criteria) or progression to muscle-invasive bladder cancer (MIBC) or metastatic disease:
- T1 High-Grade Status
- Tumor Multiplicity
- Concomitant Carcinoma in Situ (CIS)
- Presence of Residual High-Grade or T1 Tumor at re-TURBT
- Grading: A combined grading system incorporating both WHO 1973 and the 2022 standards.
03 Clinical Endpoints and Statistical Outcomes
Within the study cohort, 22.8% of patients reached the primary endpoint (BCG-unresponsiveness or progression). Notably, 12.5% of patients progressed to MIBC, a figure consistent with expectations for this high-risk population. Based on the five risk factors, a nomogram was constructed to categorize patients into four distinct risk groups.
04 Prognostic Assessment by Risk Stratification
The risk model demonstrated significant variance in 5-year event-free survival (EFS) across the groups: • Lowest Risk Group: 5-year EFS was approximately 90%. • Highest Risk Group: 5-year EFS decreased to approximately 50%. The data indicates that the model effectively distinguishes patients with poor prognosis under standard BCG therapy. For those in the highest risk category, more aggressive clinical interventions may be required.
05 Model Validation and Clinical Significance
To ensure the robustness of the model, the dataset was split into a 70% training set and a 30% validation set. Internal validation was performed using bootstrapping and 10-fold cross-validation. The results showed high predictive consistency. Professor A. Ślusarczyk concluded: While internal validation has been completed, external validation using data from additional centers is the necessary next step. The establishment of this predictive model provides a foundation for personalized medicine in high-grade NMIBC. It assists clinicians in the early identification of patients likely to derive limited benefit from BCG, allowing for optimized follow-up strategies or timely adjustments to treatment plans.
