Results from the randomized phase 3 KEYNOTE-B15/EV-304 trial, presented as a late-breaking abstract at the 2026 ASCO Genitourinary Cancers Symposium, showed that perioperative enfortumab vedotin (Padcev) plus pembrolizumab (Keytruda) reduced the risk of disease-related events by 47% and the risk of death by 35% compared with neoadjuvant gemcitabine plus cisplatin in patients with muscle-invasive bladder cancer eligible for platinum-based therapy.
Lead author Matthew Galsky, MD, of the Mount Sinai Tisch Cancer Center, reported a 24-month estimated event-free survival rate of 79.4% with the combination versus 66.2% with cisplatin-based chemotherapy (HR 0.53; P<.0001), along with an overall survival rate of 86.9% versus 81.3% (HR 0.65; P=.0029). The pathologic complete response rate was 55.8% in the combination arm compared with 32.5% in the chemotherapy arm. The study enrolled 808 cisplatin-eligible patients randomized 1:1 between the two regimens, each followed by surgery.
The findings extend prior results from KEYNOTE-905, which established this combination in patients ineligible for cisplatin and led to an FDA approval in November 2025. Together, the two trials support enfortumab vedotin plus pembrolizumab as a perioperative option regardless of cisplatin eligibility, with significant implications for coverage policy.
For managed care organizations, the shift away from cisplatin-based neoadjuvant chemotherapy involves both cost and site-of-care considerations. The enfortumab vedotin plus pembrolizumab regimen involves intravenous administration and a more complex dosing schedule. Grade 3 or higher adverse events occurred in 75.7% of patients in the combination arm versus 67.2% with chemotherapy, with skin reactions as the most common high-grade drug-related concern.
Regulators are expected to evaluate supplemental applications based on KEYNOTE-B15 data. Prior authorization and formulary policies will likely need to be updated given the potential for broad adoption across cisplatin-eligible and ineligible patient populations.