A novel risk score based on a combined signature of 10 immune system genes to predict bladder cancer prognosis.

Bladder cancer (BC) is a common internal malignant tumor with a poor prognosis worldwide. There is an urgent need to better understand the pathogenesis and progression of BC and to find useful biomarkers for diagnosis and prognosis. This study was aimed at developing a potential immunogenomic prognostic signature for BC patients. To identify possible immune-system-related genes (IRGs) whose parameters predict the survival of BC patients, we chose 371 BC patients and analyzed differentially expressed IRGs from The Cancer Genome Atlas (TCGA) datasets. We then derived a 10-IRG formula, including MMP9, RBP7, PDGFRA, AHNAK, OAS1, OLR1, RAC3, SLIT2, IGF1, and AGTR1, to estimate BC prognosis. To validate the mRNA levels of these IRGs, we performed quantitative PCR and found that the expression of these genes almost matched the corresponding mRNA expression levels in TCGA. Furthermore, we validated the prognostic value of the new risk model using two external datasets from Gene Expression Omnibus: GSE13507 (n = 165) and GSE32894 (n = 224). Our data pointed to a significant correlation between the risk model and patients' prognosis. Bioinformatic analysis revealed that products of the IRGs have possible effects on tumor immune processes such as an inflammatory response and cytokine-cytokine receptor interaction. Finally, assessment of the clinical value of the immune-system-based risk signature showed that several of these IRGs were differentially expressed between patients with different clinical characteristics: a high risk score positively correlated with female sex, advanced tumor stage, more advanced T stage, and lymph node metastasis. This immunogenomic signature may represents a reliable prognostic tool for BC and can help to design an individualized immunotherapy.