A novel gene signature combination improves the prediction of overall survival in urinary bladder cancer.

摘要: Objectives: Bladder carcinoma is a clinical heterogeneous disease, which with significant variability of the prognosis and high risk death. This revealed prominently need to identify high-efficiency cancer characteristics predict prognosis. Methods: Gene expression profiles 93 bladder tumor patients from Expression Omnibus data sets was performed in this study, along 408 retrieved The Cancer Genome Atlas database. relationship gene signature overall survival analyzed training cohort (n = 46). validation for that an internal 47) external 408). Results: Four genes (TMPRSS11E, SCEL, KRT78, TMEM185A) were identified by univariable multivariable Cox regression analysis. According score on bases four-gene signature, we grouped these high-risk group low-risk significantly different series successfully validated it both cohorts. Subsequent studies demonstrated independent radical cystectomy stage, chemotherapy lymph node status. Higher rates FAT4 mutation MACF1 tumors found compared low score. set enrichment analysis associated some progression recurrence pathways. Conclusions: might have potential implications selection urinary aggressive therapy. selected four become therapeutic targets diagnostic markers cancer.