Prevalence of PD-L1-High Tumor Expression and Association with First-Line Treatment Outcomes in Patients with Metastatic Urothelial Carcinoma: Results from a Prospective Observational Cohort Study

BACKGROUND: Programmed death-ligand 1 (PD-L1) expression has been studied extensively in clinical trials of metastatic urothelial carcinoma. OBJECTIVE: We aimed to provide information on the prevalence of tumor PD-L1 expression and treatment outcomes in patients with metastatic urothelial carcinoma in a real-world setting. METHODS: PREVAIL (NCT03788746) enrolled patients with metastatic urothelial carcinoma undergoing systemic first-line treatment at US community oncology practices. The primary endpoint was the prevalence of pretreatment tumor PD-L1-high expression using an immunohistochemistry antibody for PD-L1 (SP263) [Roche Diagnostics]. Secondary endpoints included treatment patterns and association between PD-L1 expression and objective response rate (ORR), progression-free survival (PFS), and overall survival (OS). RESULTS: Among 139 patients, the pretreatment prevalence of PD-L1-high expression was 53.2% (95% confidence interval 44.6-61.7). Forty-seven patients (33.8%) received anti-PD-L1/anti-programmed death 1 agents alone as first-line therapy. In all evaluable patients, ORR was 50.9%, median PFS was 15.1 months, and median OS was 25.4 months from the first dose of first-line therapy. There was no significant difference in PFS (hazard ratio 0.72; 95% confidence interval 0.34-1.52; p = 0.385) or OS (hazard ratio 0.71; 95% confidence interval 0.44-1.15; p = 0.164) in patients with pretreatment PD-L1-high versus PD-L1-low expression. Objective response rate was similar between patients with PD-L1-high (17/33; 51.5%) and PD-L1-low (12/24; 50.0%) expression. CONCLUSIONS: In this observational study, more than half of patients had high pretreatment tumor PD-L1 expression. There was no significant association between PD-L1 status and ORR, PFS, or OS with any first-line therapy, although definitive conclusions cannot be made because of the limited sample size.