Background: Bladder cancer is one of the most common malignancies. So far, no effective biomarker for bladder cancer prognosis has been identified. Aberrant DNA methylation is frequently observed in the bladder cancer and holds considerable promise as a biomarker for predicting the overall survival (OS) of patients. Materials and methods: We downloaded the DNA methylation and transcriptome data for bladder cancer from The Cancer Genome Atlas (TCGA), a public database, screened hypo-methylated and upregulated genes, similarly, hyper-methylation with low expression genes, then retrieved the relevant methylation sites. Cox regression analysis was used to identify a nine-methylation site signature of a training group. Predictive ability was validated in a test dataset by receiver operating characteristic (ROC) analysis. Results: We identified nine bladder cancer-specific methylation sites as potential prognostic biomarkers and developed a risk score formula to evaluate the OS. The performance of the signature was accurate, with area under curve was 0.73 in the training group and 0.71 in the test group. Taking clinical features into consideration, we constructed a nomogram consisting of the nine-methylation site signature and patients’ clinical variables, and found that the signature was an independent risk factor. Conclusions: Overall, the significant nine methylation sites could be novel prediction biomarkers, which could aid in treatment and also predict the overall survival likelihoods of bladder cancer patients.
Development of a Novel Prognostic Signature for Predicting the Overall Survival of Bladder Cancer Patients
- Views Icon Views
- PDF LinkPDF
- Share Icon Share
- Cite Icon Cite
Huamei Tang, Lijuan Kan, Tong Ou, Dayang Chen, Xiaowen Dou, Wei Wu, Xiang Ji, Mengmeng Wang, Zengyan Zong, Hongmei Mo, Xiuming Zhang, Dan Xiong; Development of a Novel Prognostic Signature for Predicting the Overall Survival of Bladder Cancer Patients. Biosci Rep BSR20194432. doi: https://doi.org/10.1042/BSR20194432
Download citation file: