Cuproptosis depicts tumor microenvironment phenotypes and predicts precision immunotherapy and prognosis in bladder carcinoma.

Background: Though immune checkpoint inhibitors (ICIs) exhibit durable efficacy in bladder carcinomas (BLCAs), there are still a large portion of patients insensitive to ICIs treatment. Methods: We systematically evaluated the cuproptosis patterns in BLCA patients based on 46 cuproptosis related genes and correlated these cuproptosis patterns with tumor microenvironment (TME) phenotypes and immunotherapy efficacies. Then, for individual patient's evaluation, we constructed a cuproptosis risk score (CRS) for prognosis and a cuproptosis signature for precise TME phenotypes and immunotherapy efficacies predicting. Results: Two distinct cuproptosis patterns were generated. These two patterns were consistent with inflamed and noninflamed TME phenotypes and had potential role for predicting immunotherapy efficacies. We constructed a CRS for predicting individual patient's prognosis with high accuracy in TCGA-BLCA. Importantly, this CRS could be well validated in external cohorts including GSE32894 and GSE13507. Then, we developed a cuproptosis signature and found it was significantly negative correlated with tumor-infiltrating lymphocytes (TILs) both in TCGA-BLCA and Xiangya cohorts. Moreover, we revealed that patients in the high cuproptosis signature group represented a noninflamed TME phenotype on the single cell level. As expected, patients in the high cuproptosis signature group showed less sensitive to immunotherapy. Finally, we found that the high and low cuproptosis signature groups were consistent with luminal and basal subtypes of BLCA respectively, which validated the role of signature in TME in terms of molecular subtypes. Conclusions: Cuproptosis patterns depict different TME phenotypes in BLCA. Our CRS and cuproptosis signature have potential role for predicting prognosis and immunotherapy efficacy, which might guide precise medicine.