746 Background: EV-302/KEYNOTE-A39 (NCT04223856) showed sustained efficacy benefits after long-term follow-up for pts with previously untreated la/mUC treated with EV+P vs chemo, establishing EV+P as the standard of care (SOC) in this setting. We previously reported an exploratory analysis of responders from EV-302 (Gupta S, et al. ASCO 2025, Abs 4502). After approximately 2.5 years of median follow-up, confirmed objective response rate (complete response CR + partial response PR) was 67.5% with EV+P and 44.2% with chemo; CR was 30.4% and 14.5%, respectively. For pts with CR, median PFS by blinded independent central review (BICR), median OS, and median duration of CR were not estimable (NE) with EV+P. Among pts with CR in the EV+P arm, median treatment (tx) duration was longer than in the overall population, but no worsening of safety was seen. Building on this exploratory analysis, we report additional data to further characterize responders in EV-302. Methods: Pts were randomized 1:1 to receive EV (1.25 mg/kg; D1 and D8, IV) + P (200 mg; D1, IV) or chemo (gemcitabine + cisplatin (cis)/carboplatin). Primary endpoints were PFS by BICR and OS. Secondary endpoints included ORR, duration of response (DOR), and safety. An exploratory subgroup analysis evaluated outcomes in responders. Results: Median follow-up (cutoff: Aug 8, 2024) was 29.1 mo (95% CI, 28.5-29.9). 886 pts were randomized to EV+P (n = 442) vs chemo (n = 444). Median time to objective response was 2.1 mo with both EV+P and chemo. In the EV+P and chemo arms, respectively, median DOR was 23.3 (17.8-NE) mo and 7.0 (6.2-9.0) mo in responders, 24.4 (17.8-NE) mo and 8.3 (5.9-10.8) mo in cis-eligible pts, and 21.9 (15.7-NE) mo and 6.6 (5.5-9.3) mo in cis-ineligible pts. Median time to CR was 4.3 mo with EV+P and 4.2 mo with chemo. In the EV+P arm, 88/437 (20.1% of the response evaluable set) achieved PR then converted to CR; 38/441 (8.6%) did in the chemo arm. With most pts achieving CR initially experiencing PR (88/133 with EV+P and 38/64 with chemo), median time from first PR to CR was 4.5 mo with EV+P and 6.1 mo with chemo. Among responders in the EV+P arm, the median number (range) of cycles was 12 (1-54) for EV and 17 (1-35) for P. Conclusions: Among pts achieving objective response, PFS and OS were longer in the EV+P arm than in the chemo arm. Durable responses were seen regardless of cis eligibility. Most pts in CR achieved PR before CR, with pts in the EV+P arm converting to CR more rapidly than those in the chemo arm. The safety profile of EV+P in responders was consistent with that of the overall population, and no new safety signals were identified. These data, together with long-term outcomes for pts with CR and PR, reinforce EV+P as the SOC for 1L tx of pts with la/mUC. Clinical trial information: NCT04223856 .
Gupta et al. (Sun,) studied this question.