Abstract
Background: Muscle invasive bladder cancer (MIBC) is an aggressive cancer characterized by therapeutic resistance and poor prognosis, which are possibly due to the existence of cancer stem cells (CSCs). In this study, we aimed to characterize the expression of cancer stemness-related genes and develop a multi-gene risk signature to predict clinical outcome and treatment response in MIBC.Methods: The mRNA expression data and clinical data of MIBC patients were collected from The Cancer Genome Atlas (TCGA) database and Gene Expression Omnibus (GEO) database, which included the TCGA training cohort (n = 333) and three GEO validation cohorts, GSE13507 (n = 165), GSE32548 (n = 127), and GSE48075 (n = 72). A list of 166 stemness-related genes were obtained from the Cancer Single Cell State Atlas (CancerSEA) database and prognostic genes for overall survival (OS) were identified by univariate Cox analysis. Then, the least absolute shrinkage and selection operator (LASSO) regression and stepwise multivariate Cox regression were performed to generate a multi-gene risk signature. Kaplan-Meier curve, time-dependent receiver operating characteristic (ROC) curve, multivariate analysis, and stratification analysis were used to evaluate the performance of the gene signature. We also explored the relationship between risk score and response to chemotherapy and radiotherapy in MIBC patients. Moreover, independent prognostic factors for OS were combined together into a nomogram to improve predictive performance.Results: Firstly, a total of 25 prognostic genes were identified. Then, a seven-gene risk signature (EGFR, FOXA2, HES1, MME, RBM6, SMOC2, and TFRC) was constructed and it could robustly classify MIBC patients into high -risk and low-risk groups with different clinical outcomes. ROC curves showed that the seven-gene signature had a robust predictive accuracy in four cohorts. Besides, high risk score was significantly associated with advanced clinical stage and treatment failure. As an independent risk factor for OS, the stemness-related seven-gene signature could achieve better prognostic accuracy when integrated with clinical factors. Conclusions: We developed and validated a robust stemness-related gene signature which could robustly predicate clinical outcome and shed light on the cancer stemness in bladder cancer.
BCG-Mediated Bladder Cancer Immunotherapy: Identifying Determinants of Treatment Response Using a Calibrated Mathematical Model
