Abstract
The current options and recent developments in the field of systemic therapy for advanced urothelial cancer (UC) patients urge the need for selection criteria to identify the most optimal therapeutic option for individual patients. The molecular makeup of tumors, including molecular subtype, tumor microenvironment composition, and gene mutations, fusions, and amplifications, has previously been correlated with a response to immune checkpoint inhibitors, erdafitinib, or enfortumab vedotin (EV) monotherapy, and may withhold potential candidate biomarkers. In this study, we aimed to stratify metastatic UC (mUC) patients based on molecular biomarkers that might be associated with a response to EV, a fibroblast growth factor receptor inhibitor, or anti-PD-(L)1, by using whole-genome DNA-sequencing and paired RNA-sequencing data of fresh-frozen metastatic tumor biopsies of 155 mUC patients. We observed that NECTIN4 amplification, FGFR2/3 mutations, and the RNA expression-based T-cell-to-stroma enrichment (TSE) score were mutually exclusive, and may therefore reflect biologically distinct tumors and sensitivity to treatments. This finding was validated in two independent bladder cohorts: the IMvigor210 study and The Cancer Genome Atlas. Stratification of patients into subgroups based on these molecular features is possible. Our data challenge the concept of a one-treatment-fits-all paradigm and support the rationale for prospective clinical trials with biomarker-guided treatment selection of mUC patients.