Adding perioperative pembrolizumab to chemotherapy does not improve EFS in MIBC

Adding perioperative pembrolizumab (Keytruda) to neoadjuvant cisplatin plus gemcitabine did not significantly improve event-free survival (EFS) in patients with cisplatin-eligible muscle-invasive bladder cancer (MIBC), according to results from the phase 3 KEYNOTE-866 trial (NCT03924856).

The results were presented at the 2026 European Society for Medical Oncology (ESMO) Congress in Madrid, Spain, and concurrently published in Annals of Oncology.^1,2

KEYNOTE-866 represents the first randomized phase 3 trial of a perioperative PD-1 inhibitor added to cisplatin-based neoadjuvant chemotherapy in MIBC. Its results come as the treatment landscape has already moved toward immunotherapy-based perioperative regimens backed by positive phase 3 data.

“Novel systemic regimens have improved the standard of care for patients with muscle invasive bladder carcinoma,” said Jens Bedke, during the presentation at ESMO. “Since 2025, gemcitabine plus perioperative durvalumab [Imfinzi] has been approved, and this year perioperative enfortumab vedotin [Padcev] plus pembrolizumab [was] approved by the FDA and EMA, irrespective of platinum eligibility. Adjuvant nivolumab, based on the CheckMate 274 trial, [was also] approved in 2021 [in the US] and 2022 in Europe.”

Trial design

KEYNOTE-866 randomly assigned 907 adult patients to pembrolizumab plus cisplatin/gemcitabine (n = 453) or to placebo plus cisplatin/gemcitabine (n = 454). Eligible patients had centrally assessed clinical stage T2-T4aN0M0 or T1-T4aN1M0 MIBC with at least 50% urothelial histology. They had to be candidates for radical cystectomy (RC) with pelvic lymph node dissection (PLND), have an ECOG performance status of 0 or 1, and meet none of the Galsky criteria for cisplatin ineligibility, including a glomerular filtration rate of at least 60 mL/min. Randomization was stratified by PD-L1 combined positive score (≥10 vs <10), clinical stage, and geographic region.

Patients received pembrolizumab 200 mg or placebo every 3 weeks for 4 cycles alongside cisplatin 70 mg/m² on day 1 and gemcitabine 1000 mg/m² on days 1 and 8. RC plus PLND followed, and then 13 cycles of adjuvant pembrolizumab or placebo. The primary end point was investigator-assessed EFS. Overall survival (OS) and pathological complete response (pCR) were key secondary end points.

Efficacy findings

At a median follow-up of 60.9 months (range, 40.7 to 78.7), EFS events had occurred in 157 patients (34.7%) in the pembrolizumab arm and 168 patients (37.0%) in the placebo arm. Median EFS was 74.1 months (95% CI, 72.4 to NR) in the pembrolizumab arm and not reached (95% CI, 67.7 to NR) in the placebo arm (HR, 0.88; 95% CI, 0.71 to 1.10; 1-sided P = .1366). Estimated EFS was 76.7% vs 70.0% at 24 months and 63.5% vs 61.9% at 60 months, respectively.

Results were generally consistent across subgroups. The HR was 0.87 (95% CI, 0.65 to 1.19) in patients with a PD-L1 combined positive score of 10 or higher and 0.94 (95% CI, 0.68 to 1.28) in those with a score below 10.

OS was similar between arms, with 135 deaths occurring in each arm. At the time of data report, the median OS was not reached in either cohort (HR, 1.00; 95% CI, 0.79 to 1.27). The estimated 60-month OS was 70.8% with pembrolizumab and 70.9% with placebo.

The pCR rate was 44.4% (95% CI, 39.8 to 49.2) with pembrolizumab and 35.8% (95% CI, 31.3 to 40.4) with placebo, representing an estimated difference of 8.9 percentage points (95% CI, 2.5 to 15.1). Patients who did not undergo cystectomy were counted as nonresponders.

Cystectomy was performed in 415 patients in the pembrolizumab arm and 399 in the placebo arm. Of these, 318 and 315, respectively, started adjuvant treatment. Among patients who had recurrent or metastatic disease after cystectomy, subsequent anti–PD-(L)1 therapy was given to 4.9% in the pembrolizumab arm and 11.2% in the placebo arm.

Bedke noted, “In general, subsequent therapy was quite heterogeneous in patients who progressed after cystectomy, but also in the patient cohort who did not undergo cystectomy.”

Safety

Grade 3 to 5 treatment-emergent adverse events (TEAEs) occurred in 80.2% of patients in the pembrolizumab arm and 75.1% in the placebo arm. Serious TEAEs occurred in 61.8% and 55.2%, and TEAEs led to discontinuation of any study drug in 36.7% and 26.7%, respectively. Each arm had 1 treatment-related death.

Immune-mediated adverse events and infusion reactions of any grade occurred in 28.9% of patients receiving pembrolizumab and 9.4% receiving placebo; 8.9% vs 1.8% were grade 3 to 5. Hypothyroidism (10.4%), hyperthyroidism (5.1%), pneumonitis (4.0%), and severe skin reactions (3.1%) were the most frequent immune-mediated events with pembrolizumab.

Adding pembrolizumab did not appear to compromise surgery: grade 3 to 5 adverse events during the surgical phase occurred in 42.7% of the pembrolizumab arm and 46.5% of the placebo arm.

Based on these results, Bedke concluded, “While perioperative pembrolizumab plus neoadjuvant cisplatin-based chemotherapy is not supported for participants [with] MIBC, we know that the perioperative enfortumab vedotin plus pembrolizumab has become a new standard of care in this setting, irrespective of platinum eligibility.”

REFERENCES

1. Bedke J, Arkosy PF, Nishiyama H, et al. Perioperative pembrolizumab (pembro) or placebo (pbo) + neoadjuvant cisplatin (cis) and gemcitabine (gem) in participants (pts) with muscle-invasive bladder cancer (MIBC): The double-blind, phase 3 KEYNOTE-866 study. Presented at: 2026 European Society for Medical Oncology Congress. October 23-27, 2026. Madrid, Spain. Abstract 47350

2. Bedke J, Arkosy P, Nishiyama H, et al. Perioperative pembrolizumab or placebo plus neoadjuvant cisplatin and gemcitabine in participants with muscle-invasive bladder cancer who are eligible for cisplatin: the double-blind, phase III KEYNOTE-866 study. Ann Oncol. 2026. doi:10.1016/j.annonc.2026.10.106