Classification of Muscle-Invasive Bladder Cancer Based on Immunogenomic Profiling

Abstract

There is a significant heterogeneity in the immunotherapeutic responsiveness of each muscle-invasive bladder cancer (MIBC) patient. In our research, we aimed to identify a novel classification of MIBC based on immunogenomic profiling that may facilitate the reasonable stratification of prognosis and response to immunotherapy. The single-sample gene-set enrichment analysis (ssGSEA) was used to analyze the RNA-seq data of 29 important immune signatures from TCGA. Unsupervised hierarchical clustering was performed to identify an immunogenomic classification of MIBC. Then, we assessed the features of the classification in prognosis, immune infiltration, tumor-infiltration lymphocytes, HLA genes, and PD-L1 expression level. A total of 399 MIBC samples were included and three subtypes named Immunity_High, Immunity_Medium, and Immunity_Low were identified. The Immunity_High had a significant advantage in overall survival over the Immunity_Medium and Immunity_Low (p = 0.046 and p = 0.024). From Immunity_Low to Immunity_High, immune cell infiltration and stromal content showed an upward trend (p < 0.001). Meanwhile, Immunity_High was associated with a significantly higher proportion of TILs including dendritic cells resting, macrophages M1, mast cells resting, T cells CD4 memory activated, and T cells CD8+. And the expression levels of all HLA genes and PD-L1 of Immunity_High were the highest, consistent (p < 0.001). Two hundred ninety eight MIBC patients treated with immunotherapy from the IMvigor210 were included to form an independent validation cohort to verify the robustness of immunogenomic classification and the ability to predict the response to immunotherapy. This classification had potential clinical implications for predicting prognosis and immunotherapeutic responsiveness of MIBC patients.