Assessing the value of perioperative treatment for muscle-invasive bladder cancer (MIBC) requires attention to both pathological response and survival outcomes. Delivering that treatment throughout the patient's journey also calls for close coordination between medical oncology and urology. At the 2026 European Society for Medical Oncology (ESMO) Congress, the phase 3 KEYNOTE-866 trial reported results for perioperative pembrolizumab combined with cisplatin-based chemotherapy in patients with cisplatin-eligible MIBC. In this interview, the trial's presenter, Professor Jens Bedke of Stuttgart, Germany, discusses the study design and key findings, the interpretation of pathological complete response (pCR) and survival endpoints, and multidisciplinary coordination across neoadjuvant therapy, radical cystectomy and adjuvant therapy.

Q1. What are the key findings from KEYNOTE-866, and what new information do they provide for perioperative treatment decisions in patients with cisplatin-eligible MIBC?

Professor Jens Bedke:

KEYNOTE-866 is a phase 3 trial in the perioperative treatment of muscle-invasive bladder cancer. Perioperative treatment here means a three-step approach: neoadjuvant therapy, radical cystectomy, and adjuvant therapy. Much has been explored in this setting in recent years, and several trials have been reported; what distinguishes KEYNOTE-866 is that immunotherapy is embedded in both the first and the third step.

All enrolled patients had cisplatin-eligible disease, mainly a GFR above 60 and no Galsky criterion precluding cisplatin. Patients were randomised 1:1 to standard gemcitabine plus cisplatin with pembrolizumab, or the same regimen with placebo, given as four cycles of neoadjuvant therapy. Radical cystectomy with pelvic lymph node dissection followed. In the third step, the investigational arm continued with 13 cycles of adjuvant pembrolizumab and the control arm with placebo. The primary endpoint was EFS; key secondary endpoints included overall survival (OS) and the pCR rate.

EFS was analysed at the second interim analysis, when approximately 330 events had occurred and median follow-up exceeded 16 months. The trial recruited between 2019 and 2022, and baseline characteristics were well balanced between arms. The primary endpoint was not met: the hazard ratio was 0.88 with a P value of 0.13, short of the prespecified one-sided threshold of 0.025, and the EFS curve showed only a small numerical improvement. Because the primary endpoint was not significant, OS was not formally tested; the event rate was identical in the two arms, with a hazard ratio of 1.00. Nevertheless, both EFS and OS in this trial exceeded historical benchmarks for the setting, including in the control arm. The second key secondary endpoint, pCR, was also not formally tested: in patients who received four cycles of gemcitabine and cisplatin with pembrolizumab, the pCR rate in the radical cystectomy specimen was 44% versus 35% in the control arm, an absolute difference of 9 percentage points.

The regimen was well tolerated. Safety profiles did not differ between the two arms, and no new safety signals emerged for either chemotherapy or pembrolizumab. In particular, surgery proved feasible and safe in this treatment setting.

Q2. When evaluating perioperative treatments, how do you interpret pCR, EFS and OS together? Which findings in this presentation are most informative about the long-term value of the regimen?

Professor Jens Bedke:

In the perioperative setting we work with three important efficacy parameters. pCR is a relatively new endpoint introduced by these trials; EFS, disease-free survival or PFS are widely used to capture disease progression; and the clinical meaning of OS hardly needs explanation. The question is how to rank them and how to place them in a clinical context.

Ultimately, OS remains one of the most important endpoints, because it tells us directly whether a treatment can truly prolong life. EFS is an important surrogate: an event affects patients not only physically but also emotionally. As for pCR, it matters greatly to me as a urologist: surgery on a T0 tumour is far easier than on T3 or T4 disease, and patients who achieve pCR have better EFS, which usually translates into longer OS as well.

Back to KEYNOTE-866: with pembrolizumab, EFS improved only slightly in numerical terms and did not reach statistical significance, while OS showed no numerical difference and was not formally tested. The central fact, therefore, is that this regimen did not prolong OS. That said, I consider the pCR rate of 44% to be valuable information: it shows that the combination has antitumour activity in improving pathological response.

Q3. When pembrolizumab is incorporated into perioperative treatment, how should medical oncologists and urologists coordinate systemic therapy and surgery? Which safety and treatment-completion measures are most important for ensuring that patients proceed to radical cystectomy as planned?

Professor Jens Bedke:

This question goes straight to clinical practice: how to put these data to use. Throughout the perioperative period, the first thing to remember is that this is a three-step journey: the neoadjuvant phase, the surgical phase of cystectomy, and the adjuvant phase. Collaboration between medical oncologists and urologists at each of these steps is essential.

The starting point is the patient’s first visit: how to refer patients appropriately between oncology and urology, or whether urologists themselves deliver systemic therapy. This is very much country-specific. I have learned that in China, Germany and Japan, urologists also take charge of systemic treatment for patients with genitourinary cancer. The critical step is the first one: identifying these patients, offering them the perioperative concept, and starting neoadjuvant treatment.

Trials also show that a proportion of patients receiving neoadjuvant therapy will not proceed to cystectomy as planned, for various reasons. At the interface between the first and the second step, the patient’s status must therefore be assessed: if it allows, the priority is timing — schedule surgery early and do not lose the window; if it does not, we must determine whether the cause is a treatment-related adverse event or deterioration from tumour progression, and reassess.

The same applies after surgery, when starting the adjuvant phase — and this is often more critical. Radical cystectomy is major surgery, so patients must recover; at the same time, patients who have achieved pCR in particular often ask whether adjuvant therapy is still necessary. Here urologists and oncologists must speak with one voice. Conflicting messages undermine the patient’s commitment to adjuvant treatment.

Taken together, the patient journey begins at diagnosis and contains several critical decision points, each requiring ongoing multidisciplinary reassessment — collaboration between oncology and urology, and guidance for the patient in the right direction. Put briefly, the urologist is the friend of the oncologist, and the oncologist is the friend of the urologist, working side by side for the good of the patient. Only in this way can we guide patients safely through the entire treatment journey.

Summary

KEYNOTE-866 informs perioperative treatment for cisplatin-eligible muscle-invasive bladder cancer. Pembrolizumab plus chemotherapy did not produce a statistically significant improvement in the primary endpoint of event-free survival. Overall survival showed no numerical difference and was not formally tested. The numerical increase in pathological complete response, also not formally tested, suggests antitumour activity but does not establish a long-term survival benefit. Interpretation therefore requires attention to the distinct roles and limitations of each endpoint. In practice, neoadjuvant therapy, radical cystectomy and adjuvant therapy should be managed as a continuous journey, with close collaboration between urologists and oncologists, repeated assessment and consistent communication to support timely surgery and coordinated care throughout.