A robust gene prognostic index composed of GZMB, IRF1, and TP63 can stratify the risk of two metastatic urothelial carcinoma cohorts based on immune checkpoint blockade therapy

Background Immune checkpoint blockade (ICB) therapy has become a first-line treatment option for metastatic urothelial carcinoma (mUC) patients who do not meet the criteria of cisplatin. Still, only a few people can benefit from it, so useful predictive markers are needed. Methods Download the ICB-based mUC and chemotherapy-based bladder cancer cohorts, and extract the expression data of pyroptosis-related genes (PRG). The LASSO algorithm was used to construct the PRG prognostic index (PRGPI) in the mUC cohort, and we verified the prognostic ability of PRGPI in two mUC and two bladder cancer cohorts. Results Most of the PRG in the mUC cohort were immune-activated genes, and a few were immunosuppressive genes. The PRGPI composed of GZMB, IRF1, and TP63 can stratify the risk of mUC. In IMvigor210 and GSE176307 cohorts, the P -values of Kaplan Meier analysis was < 0.01 and 0.002, respectively. PRGPI could also predict ICB response, and the chi-square test of the two cohorts had P -values of 0.002 and 0.046, respectively. In addition, PRGPI can also predict the prognosis of two bladder cancer cohorts without ICB therapy. The PRGPI and the expression of PDCD1/CD274 had a high degree of synergistic correlation. The Low PRGPI group showed prominent characteristics of immune infiltration and was enriched in the immune signal activation pathway. Conclusion The PRGPI we constructed can effectively predict the treatment response and overall survival rate of mUC patients treated with ICB. The PRGPI can help mUC patients achieve individualized and accurate treatment in the future.