Abstract
Muscle-invasive bladder urothelial carcinoma (MIBC) is characteristic of high mortality and high recurrence. Distinguishing the prognostic risk of MIBC at the molecular level of miRNA expression is rarely performed and thus of great significance for the management and treatment of MIBC in clinics. Adaptive lasso Cox's proportional hazards model was used to explore the relationship between differential expression miRNAs (DEmiRNAs) and MIBC survival. Furthermore, we evaluated the epithelial-mesenchymal transition (EMT) score and immune infiltration abundance by exploring EMT signature genes and TIMER database, respectively. A total of 8 DEmiRNAs were detected to be associated with the survival rate of MIBC by using the lasso Cox algorithm. Through the linear combination of these 8 DEmiRNAs, we constructed a calculated marker, which could be used to distinguish the prognosis risk in both TCGA dataset (HR = 2.03, 95% CI = (1.47, 2.83)) and independent validation dataset (HR = 7.74, 95% CI = (1.05, 56.93)). Meanwhile, the constructed marker had reasonably high predictive values of the AUC (area under the curve) in the TCGA dataset and validation dataset being 0.73 and 0.63, respectively. In addition, we observed that the expression values of let-7c, miR-100, and miR-145 were associated with EMT score and the abundance of macrophage in tumor tissue as well. This newly identified risk score signature based on the combination of 8 miRNAs could significantly predict the prognostic risk of MIBC and might provide insight into immunotherapy and targeted therapy of MIBC.