Construction and multiple validations of a robust ferroptosis-related prognostic model in bladder cancer: A comprehensive study

Xianyu Dai,Kai Yu,Hongjie Wang,Rong Zhong,Zhongqi Zhang,Yuchuan Hou

DOI: https://doi.org/10.1097/MD.0000000000040133

2024-10-18

Abstract:Ferroptosis is iron-dependent programmed cell death that inhibits tumor growth, particularly in traditional treatment-resistant tumors. Prognostic models constructed from ferroptosis-related genes are lacking; prognostic biomarkers remain insufficient. We acquired gene expression data and corresponding clinical information for bladder cancer (BC) samples from public databases. Ferroptosis-related genes from the ferroptosis database were screened for clinical predictive value. We validated gene expression differences between tumors and normal tissues through polymerase chain reaction and western blotting. Gene ontology and Kyoto encyclopedia of genes and genomes enrichment analyses were conducted to explore signaling pathways affecting the overall survival of patients with BC. CIBERSORT was used to quantify the infiltration of 22 immune cell types. We identified 6 genes (EGFR, FADS1, ISCU, PGRMC1, PTPN6, and TRIM26) to construct the prognostic risk model. The high-risk group had a poorer overall survival than the low-risk group. Receiver operating characteristic curves demonstrated excellent predictive accuracy. The validation cohort and 3 independent datasets confirmed the models' general applicability and stability. BC tissues had elevated FADS1, PTPN6, and TRIM26 mRNA and protein levels and decreased ISCU levels. Enrichment analysis indicated that neurosecretory activity might be the main pathway affecting the overall survival. High- and low-risk groups had significantly different immune cell infiltration. Specific ferroptosis-related gene expression was associated with immune cell infiltration levels. The risk score was significantly correlated with patients' clinical characteristics. A novel, widely applicable risk model with independent predictive value for the prognosis of patients with BC was established; candidate molecules for future BC research were identified.