Comprehensive analysis and prognostic assessment of senescence-associated genes in bladder cancer.

ABSTRACT

BACKGROUND: The prevalence and mortality of bladder cancer (BLCA) present a significant medical challenge. While the function of senescence-related genes in tumor development is recognized, their prognostic significance in BLCA has not been thoroughly explored. METHODS: BLCA transcriptome datasets were sourced from the TCGA and GEO repositories. Gene groupings were determined through differential gene expression and non-negative matrix factorization (NMF) methodologies. Key senescence-linked genes were isolated using singular and multivariate Cox regression analyses, combined with lasso regression. Validation was undertaken with GEO database information. Predictive models, or nomograms, were developed by merging risk metrics with clinical records, and their efficacy was gauged using ROC curve methodologies. The immune response's dependency on the risk metric was assessed through the immune phenomenon score (IPS). Additionally, we estimated IC50 metrics for potential chemotherapeutic agents. RESULTS: Reviewing 406 neoplastic and 19 standard tissue specimens from the TCGA repository facilitated the bifurcation of subjects into two unique clusters (C1 and C2) according to senescence-related gene expression. After a stringent statistical evaluation, a set of ten pivotal genes was discerned and applied for risk stratification. Validity tests for the devised nomograms in forecasting 1, 3, and 5-year survival probabilities for BLCA patients were executed via ROC and calibration plots. IC50 estimations highlighted a heightened responsiveness in the low-risk category to agents like cisplatin, cyclopamine, and sorafenib. CONCLUSIONS: In summation, our research emphasizes the prospective utility of risk assessments rooted in senescence-related gene signatures for enhancing BLCA clinical oversight.