Characterization of Ligand-Receptor Pair in Bladder Cancer Develops a Validated Scoring Model for Prognosis and Treatment Response.

Abstract

The role of ligand-receptor (LR) pairs in disease progression has been explored in bladder cancer. However, the relationship of LR pairs with cancer prognosis and treatment response remains poorly understood. We characterized the LR pair network and identified three distinct molecular subtypes with distinct biologic features based on the TCGA database (n = 406) and validated in GSE13507 (n = 165) and GSE32894 (n = 224). Three subtypes were compared for differences in patient clinical characteristics, genomic, and transcriptomic features. A multivariate Lasso Cox regression model was applied to construct an LR pairs-based scoring model to stratify the prognostic risk of patients. We demonstrated the high LR. score patients had better responses in chemotherapy, while low LR. score patients may benefit from immune checkpoint blockade (ICB). Collectively, we identified three LR pair-related subtypes associated with prognosis. We constructed and validated a LR pairs-based gene signature, which helps to predict prognosis and differentiate the susceptible population to chemotherapy and immunotherapy in patients with bladder cancer. Among the LR pairs significantly related to prognosis, ANAX1−EGFR axis was found to be potential therapeutic target for treatment of bladder cancer.