Two drugs around surgery sharply cut relapse in muscle-invasive bladder cancer

Giving the antibody-drug conjugate enfortumab vedotin plus the immunotherapy pembrolizumab before and after bladder-removal surgery, rather than surgery alone, sharply reduced the chance that muscle-invasive bladder cancer came back: in the phase 3 KEYNOTE-905/EV-303 trial, estimated event-free survival at two years was 74.7% with the drug pair against 39.4% with surgery alone [s1]. The regimen also improved survival and cleared all detectable tumour in the bladder far more often, at the cost of frequent serious side effects [s1].

The trial addresses a hard gap in treatment. Patients with muscle-invasive bladder cancer who cannot tolerate cisplatin-based chemotherapy — often because of impaired kidney function or age — have until now usually gone straight to radical cystectomy, the surgical removal of the bladder and surrounding lymph nodes, with no proven drug therapy to add [s1]. Most patients in this trial fell into that cisplatin-ineligible group, which makes the result directly relevant to the people who have had the fewest options.

What they did

KEYNOTE-905/EV-303 randomly assigned 344 participants with muscle-invasive bladder cancer who were ineligible for or declined cisplatin-based chemotherapy: 170 to perioperative enfortumab vedotin plus pembrolizumab and surgery, and 174 to surgery alone [s1]. The drug schedule ran nine total cycles of enfortumab vedotin (1.25 mg per kilogram on days 1 and 8) and 17 total cycles of pembrolizumab (200 mg every three weeks), with surgery after the first three cycles [s1]. The primary endpoint was event-free survival; overall survival and pathological complete response — no viable tumour left in the removed tissue — were key secondary endpoints [s1]. Surgery went ahead in 87.6% of the drug group and 89.7% of the control group, so the treatment did not, on the whole, cost patients their operation [s1].

What it showed

At a median follow-up of 25.6 months, the two-year event-free survival was 74.7% in the drug group and 39.4% with surgery alone, a hazard ratio for an event or death of 0.40 (95% confidence interval, 0.28 to 0.57; P<0.001) [s1]. Overall survival followed: estimated two-year survival was 79.7% against 63.1%, a hazard ratio for death of 0.50 (95% confidence interval, 0.33 to 0.74; P<0.001) [s1]. A pathological complete response occurred in 57.1% of the drug group versus 8.6% of the control group, a difference of 48.3 percentage points (95% confidence interval, 39.5 to 56.5; P<0.001) [s1].

That an overall-survival benefit was already significant at this analysis is notable. Trials that clear a drug on how often tumours shrink, or on relapse alone, often leave the survival question open for years — a recurring tension in adjuvant cancer therapy, and one that has kept some perioperative immunotherapy regimens from settling their case. Here the harder endpoint moved with the softer ones.

The cost side

The benefit came with real toxicity. Every participant in the drug group had at least one adverse event; grade 3 or higher events occurred in 71.3%, and grade 3 or higher events judged related to the drugs in 45.5% [s1]. In the surgery-alone group, 64.8% had any adverse event and 45.9% a grade 3 or higher one [s1]. Enfortumab vedotin and pembrolizumab both carry recognised risks — skin, nerve and blood-sugar effects for the conjugate, immune-mediated inflammation for the checkpoint inhibitor — and adding them to major surgery is not a light undertaking.

What to watch

The trial was funded by the drug's manufacturer, and it is open-label rather than blinded, so patients and doctors knew the assignment [s1]. Even so, the size of the event-free and survival differences, and the matching jump in complete responses, make this among the strongest perioperative results reported in bladder cancer, and specifically for the cisplatin-ineligible patients who most needed one. A companion analysis in cisplatin-eligible patients is being reported separately; whether regulators extend a perioperative indication, and whether the survival curves hold with longer follow-up, are the open questions. For now the finding sits alongside the older lesson that most bladder cancers are caught through symptoms such as blood in the urine, where the disease is still often muscle-invasive by the time it is found.

Sources

• [s1] Vulsteke C, Adra N, Danchaivijitr P, et al. Perioperative Enfortumab Vedotin and Pembrolizumab in Bladder Cancer (KEYNOTE-905/EV-303). New England Journal of Medicine. 18 February 2026.
• [s2] ClinicalTrials.gov. Perioperative Enfortumab Vedotin Plus Pembrolizumab Versus Surgery in Muscle-Invasive Bladder Cancer (NCT03924895). U.S. National Library of Medicine.