Circulating tumor DNA (ctDNA) monitoring in patients (pts) with advanced urothelial carcinoma (aUC) treated with enfortumab vedotin +/- pembrolizumab (EVP).

4560 Background: CtDNA is an emerging biomarker in aUC, but its role for pts receiving EVP is unclear. Methods: We undertook a retrospective analysis of pts with aUC who were longitudinally tested for ctDNA (MTM/mL) with a tumor-informed assay (Signatera, Natera, Inc) while on treatment (tx) with EV+/-P. Pt data were abstracted from electronic medical record. Outcomes were compared based on changes in ctDNA from pre-tx baseline, using Kaplan-Meier method and Cox proportional hazards test to assess progression-free survival (PFS) and overall survival (OS). Results: Longitudinal ctDNA data were available for 36 pts (tx: 28 EV+P; 8 EV). Pts had a median (med) of 3 ctDNA tests (range: 2-14) over 8 months (mos) of med follow-up. At pre-tx baseline, 33/36 (92%) pts had detectable ctDNA. Pt characteristics and outcomes are shown in the Table. After tx start, 26 pts (79%) had a decrease in ctDNA (med time to decrease 50 days), and 11 pts (33%) achieved negative ctDNA (-) after a med of 54 days. The 11 pts with ctDNA (-) had a response rate of 91% (CR: 6, PR: 4, SD: 1). Among 14 pts with ctDNA nadir followed by a rise, 8 pts (57%) had PD on next scan, and med time from initial ctDNA rise to PD was 124 days. Pts who achieved ctDNA (-) within 2 mos of tx (n = 8) had improved PFS relative to pts with ctDNA data (n = 21) who did not (HR: 0.08, 95% CI: 0.01 – 0.59, p = 0.01). Pts with no decrease in ctDNA within 2 mos of tx start (n = 7) had inferior PFS (HR: 5.9, 95% CI: 1.9 – 19.2, p < 0.01) and OS (HR: 20.8, 95% CI: 2.3 – 192, p < 0.01) vs pts with a ctDNA decrease (n = 22). In the EV+P group, pts with no ctDNA decrease within 2 mos (n = 4) had inferior PFS (HR: 15.9, 95% CI: 1.6 - 154, p = 0.01) vs pts with a decrease (n = 18). Conclusions: In this retrospective analysis of pts treated with EV+/-P, 92% had detectable ctDNA and 79% had a ctDNA decrease after tx start. Within 2 mos of tx start, pts who achieved ctDNA (-) had improved PFS, while pts who did not have ctDNA decrease had inferior PFS and OS. These hypothesis-generating results warrant validation in larger prospective cohorts and can inform clinical decision-making. Total cohort (n= 33) EV+P (n= 25) Sex – n (%) Male Female 7 (21%)26 (79%) 6 (24%)19 (76%) Primary Tumor Site – n (%) Lower Tract Upper Tract 25 (76%)8 (24%) 19 (76%)6 (24%) Histology – n (%) Pure Urothelial Majority Urothelial Majority Variant 25 (76%)7 (21%)1 (3%) 19 (76%)6 (24%)0 Median baseline ctDNA (MTM/mL) Responders Non-Responders 23.139.8 30.169.6 Outcomes by ctDNA status within 2 mos of EV+/-P tx start, mos (95% CI), p mOS: ctDNA (-) vs detectable NR (NR – NR) vs NR (11.7 – NR), p=0.32 NR (NR – NR) vs NR (NR – NR), p= 0.99 mPFS: ctDNA (-) vs detectable 18.7 (NR – NR) vs 7.2 (5.6 – NR), p=0.01 NR (NR – NR) vs 7.2 (5.6 – NR), p=0.99 mOS: No ctDNA decrease vs ctDNA decrease 11.7 (3.9 – NR) vs NR (NR – NR), p<0.01 NR (3.9 – NR) vs NR (NR – NR), p= 1.0 mPFS: No ctDNA decrease vs ctDNA decrease 3.9 (1.9 – NR) vs 13.2 (7.9 – NR), p<0.01 4.8 (1.9 – NR) vs 13.2 (8.4 – NR), p=0.01