Immune and Stromal Signature-Based Prognostic Gene of Bladder Cancer
Abstract Background: Immune and stromal cells in the tumor microenvironment have exhibit critical relevance with tumorigenesis and progression. Therefore, our analysis was conducted to explore the potential prognostic factors and the therapeutic targets of bladder cancer (BC) that affiliate to immune and stromal infiltration signature.Methods: Immune and stromal score were calculated for BC patients from TCGA database using ESTIMATE algorithm to predict the level of infiltration. Kaplan-Meier curves were utilized to assess the correlation of immune/stromal infiltration with survival outcomes. Differential expression genes (DEGs) were identified. Enrichment analyses, Kaplan-Meier curves, protein-protein interaction (PPI) network construction were performed to describe and screen the core module genes, whose prognostic value was further validated by an independent GEO dataset. Transcriptional factors (TFs) and ncRNA. correlated with the core module were identified using RAID 2.0 and TRRUST 2.0 database, and drugs potentially regulative for BC were accordingly screened using DrugBank.Results: 393 and 93 BC patients were enrolled from TCGA and GEO respectively. Higher stromal infiltration indicated worse overall survival (P = 0.015), and higher immune infiltration indicated an improvement on overall survival (P = 0.042). 562 DEGs were identified and 123 of them associated with survival outcomes. PPI has identified the core module genes, in which EFNB2 was the only core prognostic gene that was validated by both TCGA (P = 0.042) and GEO dataset (P = 0.036). Four TFs and Five ncRNAs (e.g. HIF1A) were associated with the regulation of the core module genes, and six drugs were screened as potential candidates to regulate BC.Conclusion: Higher immune infiltration in bladder tumor was correlated with improved survival and higher stromal infiltration corelated with worse survival. Furthermore, a higher expression of EFNB2 was tied with poorer prognosis of BC, which was validated by two independent datasets.