SWI/SNF-associated molecular subtypes reshape tumor microenvironmental features and inform precision therapeutic strategies in bladder cancer.

Background: Bladder cancer (BLCA) is a malignant tumor characterized by pronounced molecular and immunological heterogeneity. Despite continuous advances in immunotherapy and targeted treatment, clinical outcomes vary substantially, and effective molecular stratification and predictive biomarkers remain lacking. The SWI/SNF chromatin-remodeling complex regulates chromatin accessibility and gene transcription through ATP-dependent mechanisms and frequently harbors mutations across multiple cancer types. However, whether SWI/SNF-related molecular patterns can be leveraged for clinically meaningful stratification and therapeutic prediction in BLCA remains unclear. Methods: We integrated the TCGA-BLCA dataset, the IMvigor210 immunotherapy cohort, GEO datasets, and an independently sequenced Xiangya RNA cohort to comprehensively analyze SWI/SNF-related genes. Unsupervised clustering was applied to construct SWI/SNF molecular subtypes, and a quantitative scoring system, SWI_SNF_Score, was established based on differentially expressed and prognostically relevant genes. Using transcriptomic profiling across multiple cohorts and bioinformatics approaches, we assessed the relationships between SWI_SNF_Score, molecular subtypes, clinical outcomes, characteristics of the tumor immune microenvironment, and therapeutic responses, followed by validation in independent cohorts. Results: Two distinct SWI/SNF-associated molecular subtypes were identified across integrated transcriptomic datasets. Based on differentially expressed genes, we constructed the SWI_SNF_Score to quantitatively represent SWI/SNF-related molecular and immune features in BLCA. A high SWI_SNF_Score was significantly associated with the basal subtype, enhanced stromal and immune activation, frequent RB1/TP53 co-mutations, and unfavorable prognosis. In contrast, a low SWI_SNF_Score displayed luminal differentiation features, enrichment of FGFR3/KDM6A mutations, and favorable survival outcomes. The SWI_SNF_Score accurately predicted classical molecular subtypes, TME phenotypes, and multiple treatment sensitivities and was independently validated in both the Xiangya and IMvigor210 immunotherapy cohorts. Notably, the score exhibited distinct predictive value across different immune phenotypes, enabling more precise stratification of potentially responsive populations. Conclusions: The SWI_SNF_Score enables quantitative evaluation of molecular and immune heterogeneity in bladder cancer and provides a clinically relevant framework for prognostic stratification and immunotherapy response prediction.