Molecular Classifications and Candidate Biomarkers for Personalizing BCG Therapy in Non-Muscle-Invasive Bladder Cancer
TL;DR
A narrative review summarizes molecular classifications and candidate biomarkers associated with BCG response and assesses their potential clinical utility, based on a focused search of PubMed, Scopus, and Web of Science through July 2026, supplemented by ClinicalTrials.gov.
Abstract
Intravesical Bacillus Calmette–Guérin (BCG) remains the standard adjuvant treatment for intermediate- and high-risk non-muscle-invasive bladder cancer. However, 30–50% of patients fail to respond or experience recurrence within two years, while clinical and pathological risk models have limited accuracy in predicting individual response. This narrative review summarizes molecular classifications and candidate biomarkers associated with BCG response and assesses their potential clinical utility, based on a focused search of PubMed, Scopus, and Web of Science through July 2026, supplemented by ClinicalTrials.gov, regulatory documents, and current guidelines. Transcriptomic classifications, including UROMOL2021, BCG Response Subtypes (BRS), MSP888, and subtypes defined in stage T1 disease, identify tumor phenotypes associated with differences in recurrence, progression, and outcomes after BCG. BRS3 is characterized by basal-squamous and immunosuppressive features and is associated with shorter recurrence-free and progression-free survival. Genomic alterations, including FGFR3 and TERT promoter mutations and APOBEC signatures; DNA methylation markers such as GATA2, TBX3, and GPR158; and tumor microenvironment features, including regulatory T cells, exhausted CD8+ T cells, and PD-L1 expression, have been associated with outcomes in individual cohorts but remain unvalidated as treatment-specific predictors. Bladder-preserving options for BCG-unresponsive disease include pembrolizumab, nadofaragene firadenovec, nogapendekin alfa inbakicept-pmln with BCG, and the gemcitabine intravesical system. Randomized trials adding checkpoint inhibition to BCG in BCG-naive disease have reported inconsistent results. No molecular biomarker is currently validated for treatment selection. Molecular profiling remains investigational and may complement, but not replace, established clinical risk assessment. Clinical implementation requires prospective biomarker-stratified trials with standardized definitions, assays, and endpoints.