Integrated Genomic and Single-Cell Analysis Reveals Heterogeneity, Prognosis, and Treatment Vulnerability in Urothelial Carcinoma

At the transcriptomic level, several molecular subtyping schemes have been established to elucidate the intrinsic heterogeneity of urothelial carcinoma and to inform prognostic assessment and therapeutic guidance. However, a unified molecular classification scheme characterizing genomic alterations is lacking. Unsupervised and supervised clustering identified two distinct mutational signature subtypes. Kaplan-Meier analysis demonstrated that patients with the MUT2 subtype had a higher risk of death than those with the MUT1 subtype across multiple cohorts, including IMvigor210 (hazard ratio [HR], 1.74; 95% confidence interval [CI], 1.27-2.37; <i>p</i> < 0.001), UC-GENOME (HR, 1.54; 95% CI, 0.93-2.54; <i>p</i> = 0.091), The Cancer Genome Atlas (TCGA; HR, 1.45; 95% CI, 1.06-1.98; <i>p</i> = 0.020), MSK2022 (HR, 1.34; 95% CI, 1.10-1.64; <i>p</i> = 0.004), MSK2015 (HR, 3.43; 95% CI, 1.36-8.64; <i>p</i> = 0.005), and the Tongji cohort (HR, 4.99; 95% CI, 0.57-43.69; <i>p</i> = 0.11). Immunotherapy response rates were significantly higher in the MUT1 subtype than in the MUT2 subtype in IMvigor210 (31.8% vs. 13.1%; <i>p</i> = 0.003) and UC-GENOME (42.3% vs. 29.0%; <i>p</i> = 0.022). Consistent with these findings, single-cell analysis showed that MUT2 tumors were enriched in tumor-associated fibroblast subpopulations and had a lower abundance of immune effector cells. Overall, this genomic analysis identified two mutation-based subtypes of urothelial carcinoma associated with patient prognosis and immunotherapy response.