Real-world (rw) first-line (1L) treatment (tx) patterns and overall survival (OS) in patients (pts) with locally advanced/metastatic urothelial carcinoma (la/mUC) and frailty or comorbidities.

Abstract

666 Background: Enfortumab vedotin + pembrolizumab (EV+P) is now the preferred 1L tx for la/mUC, showing superior OS vs platinum-based chemotherapy (PBC) in the pivotal EV-302 trial. EV is associated with distinct adverse events, and clinical factors have been proposed to identify pts who are potentially EV unsuitable (EVU) to minimize serious toxicity. This study analyzed rw txs of pts with frailty or certain comorbidities ineligible for EV-302. Methods: This retrospective cohort study analyzed data from pts diagnosed with la/mUC between Jul 1, 2020, and Aug 31, 2024 using the Flatiron Health Research Database. Based on proposed criteria, pts were divided into 2 hypothetical/mutually exclusive cohorts: EVU and EV suitable (EVS), and within each group, by 1L tx status (treated, untreated) and tx choice. EVU was defined as having ≥1 of these factors: hemoglobin A1c ≥11% or blood glucose > 200 mg/dL in 2 consecutive samples within 1 week; creatinine clearance or glomerular filtration rate ≤20 mL/min; liver impairment grade ≥2; ECOG PS ≥3; neuropathy; corneal or retinal abnormality. EVU prevalence, pt characteristics at index, and tx patterns were summarized with descriptive statistics. Kaplan-Meier method was used to estimate rwOS from index date (treated: 1L tx date; untreated: la/mUC diagnosis date) to death. Cox proportional hazards regression was used to compare rwOS between treated EVU and EVS pts. Results: This study included 3500 pts with la/mUC (median age, 74 years; male, 72%; ECOG PS 0-1, 55%; bladder as primary site, 77%; median follow-up, 7 mos). A total of 748 (28%) of pts were identified as EVU, of whom 24% were untreated. The most common EVU factor was neuropathy (14%). Median rwOS was shorter in EVU vs EVS untreated (2.4 vs 7.0 mos) and treated pts (9.3 vs 17.0 mos) regardless of 1L tx. Immunotherapy monotherapy (IO mono) was the most common agent in EVU (38%). In the treated EVU group, rwOS was not significantly different between EV+P, PBC, or IO mono groups (Table). EVU vs EVS pts had 67% higher risk of death. Conclusions: In this study, 28% of pts with la/mUC in rw practice were EVU, representing a population typically excluded from clinical trials, and 24% were untreated. In pts treated with standard therapies, rwOS was shorter in EVU vs EVS. As tx outcomes did not significantly differ in the EVU group, shared tx decisions that incorporate various factors, including patient goals and preferences, tx tolerance, financial issues and quality of life in the context of advanced disease, has become increasingly important for this vulnerable group. Treated EVU n (%) EVU rwOS, median, mos Adjusted HR (95% CI); P value) Overall 748 (28) 9.3 Ref EVS: 171.67 (1.50-1.87); p<.001 EV+P 90 (12) Not reached Ref PBC 257 (34) 10.9 1.14 (0.70-1.86); p=0.60 IO mono 284 (38) 6.4 1.60 (0.99-2.58); p=0.06 Other tx 117 (16) 8.8 1.57 (0.95-2.59); p=0.08