End-of-life treatment patterns in patients with advanced urothelial carcinoma.

Abstract

687 Background: Immune checkpoint inhibitors and antibody-drug conjugates have expanded therapeutic options for advanced urothelial carcinoma (aUC); however, prognosis remains poor. Data on end-of-life treatment patterns in this setting are limited but are critical to inform clinical decision-making and optimize quality of care. Herein, we aimed to evaluate real-world treatment patterns in relation to the time of death in patients (pts) with aUC in a large US-based database. Methods: This study used the US-based, electronic health record-derived deidentified Flatiron Health Research Database. Eligibility: pts who were diagnosed with aUC, initiated first-line (1L) therapy, and had a recorded date of death. The data cut-off date was 6/30/2025. Pts were categorized into two groups: treatment-free, if no systemic therapy was administered during the last 3 months of life, or on-treatment, if systemic therapy was received within 3 months preceding death. Demographic and clinical variables at their last line of therapy (LOT) initiation, including age, race, region, socioeconomic status, practice type, and insurance, were summarized using medians (IQR) or proportions. Comparisons were performed using Wilcoxon rank-sum or chi-squared tests. Results: Among 15,236 pts with aUC diagnosed between 1/1/2011 and 5/8/2025 in the dataset, 7,815 were eligible and included in the final analysis. Of these, 5,221 pts (66.8%) received systemic therapy within 3 months of death (on-treatment), while 2,594 pts (33.2%) did not (treatment-free). At the time of last LOT initiation, treatment-free pts were older than those on-treatment (median age 75 years [IQR 68–81] vs. 74 [IQR 67–80], p = 0.042). Pts in the on-treatment group were more often treated in community practice (85% vs 82%, p = 0.005) and more likely to have commercial health plan (60% vs 57%, p < 0.001). Among on-treatment pts at end of life, the most common last regimen was single agent programmed cell death protein 1 (PD-1) inhibitors (42%), followed by carboplatin-based (17%), and cisplatin-based therapies (8%). Further baseline characteristics and treatment patterns will be presented at the meeting. Conclusions: Our data show that over two-thirds of pts with aUC continued systemic therapy within 3 months of death. Importantly, ~25% of these were still receiving platinum-based therapies. These findings highlight the importance of realistic treatment goals and the need to balance potential benefit with quality of life in the final months of life. Limitations include retrospective nature of study and possible data missingness.