Clinical utility of circulating tumor DNA profiling in patients with metastatic urothelial carcinoma treated with enfortumab vedotin.

798 Background: Enfortumab vedotin (EV) is a key drug for metastatic urothelial carcinoma (mUC). However, no established biomarkers have been identified to predict treatment response or prognosis in EV therapy. Recently, circulating tumor DNA (ctDNA) has attracted attention as a promising biomarker. The aim of this study was to evaluate the clinical utility of ctDNA in patients with mUC treated with EV. Methods: This was a multicenter, retrospective study conducted as part of the tumor-agnostic genomic profiling project MONSTAR-SCREEN2 . Among 151 patients with mUC enrolled, 81 received EV monotherapy, and 60 patients who underwent ctDNA analysis just prior to EV administration were included in this study. All patients underwent both tumor tissue sequencing and ctDNA analysis prior to EV treatment. Tumor tissue sequencing was performed using the MI Profile (Caris Life Sciences), and ctDNA was analyzed with the Caris Assure (Caris Life Sciences). Progression-free survival (PFS) and overall survival (OS) were evaluated using the log-rank test. Results: The primary site was the bladder in 36 patients and the upper urinary tract in 22 patients. A total of 139 variants were identified through ctDNA analysis in 46 patients (76.7%). Of these, 31 variants were derived from clonal hematopoiesis (CH), with a median allele frequency (AF) of 1.5% (IQR 1–5%). Excluding CH, a total of 108 variants were identified in 43 patients (71.7%). The most frequent substitutions were TP53 (36.7%), KMT2D (23.3%), ARID1A (10%), and RB1 (6.7%). The median interval between tumor for sequencing and ctDNA sampling was 388 days (IQR 211–769). Tumor based sequencing identified 243 variants (239 substitution mutations and 4 fusion genes), and 61 mutations were shared between DNA derived from tumor tissue and ctDNA. Patients harboring ≥3 ctDNA mutations had a significantly shorter PFS compared with those with ≤2 mutations (p = 0.040, hazard ratio HR: 2.14, 95% CI: 1.02-4.48). A trend toward shorter overall survival was also observed in patients with ≥3 mutations compared with those with ≤2 mutations (p = 0.062, HR: 2.27, 95% CI: 0.94-5.49). In addition, patients achieving disease control (DC: CR + PR + SD) had significantly lower ctDNA AF than those without DC (median 1.3% vs. 7.45%, p = 0.029). Conclusions: The ctDNA profile prior to EV treatment was associated with treatment outcomes and prognosis in mUC, suggesting its potential as a clinically useful biomarker. Association between number of ctDNA substitution variants at pre-EV administration and prognosis. Interquartile Range Progression-free survival (days) Strata Median Lower Upper p-value Number of ctDNA variants =3 (n=20) 78 50 Not available Interquartile Range Overall survival (days) Strata Median Lower Upper p-value Number of ctDNA variants =3 (n=20) 205 113 Not available