Pembro monotherapy for participants with BCG–unresponsive HR NMIBC shows anti-tumour activity

Share :
Published: 16 Mar 2023
Views: 115
Prof Andrea Necchi - The Vita-Salute San Raffaele University, Milan, Italy

Prof Andrea Necchi speaks to ecancer about the primary and subgroups analyses by PD-L1 status of cohort B of the phase 2 KEYNOTE-057 trial. The study investigated pembrolizumab monotherapy for participants with bacillus Calmette-Guérin (BCG)–unresponsive high-risk (HR) non–muscle-invasive bladder cancer (NMIBC).

Prof Necchi explains the background and study design. The study found that after 45 months of follow-up, pembrolizumab showed anti-tumour activity in the total population and across PD-L1 subgroups in patients with BCG-unresponsive papillary-only HR NMIBC who declined or were ineligible for radical cystectomy.

Prof Necchi concludes by discussing the impact of these results on clinical practice.


I will present at this congress the update related to the cohort B of the KEYNOTE-057 study. The KEYNOTE-057 study is a multicentre phase II study investigating multiple cohorts of patients with high-risk non-muscle invasive bladder cancer whose disease has failed, has relapsed, after adequate BCG therapy. So, for this patient population there is still an unmet medical need. Cohort A of the study, of the KEYNOTE-057, already provided data on the activity of pembrolizumab monotherapy in patients with carcinoma in situ components with or without papillary tumours and resulted in 40% response, complete response, at three months with about one-third of the responses maintained over time long-term, and resulted in the US FDA approval of pembrolizumab monotherapy in patients with BCG unresponsive CIS population.

The cohort B of the study, which is the focus of this meeting presentation, is related to the patients with BCG unresponsive non-CIS tumours, so papillary tumours, pTa or pT1, unresponsive to BCG. This is a very special patient population for which we don’t have any standard of care, any robust standard of care, and for which there is limited opportunity for inclusion in clinical trials. 

The overall findings from this cohort, which included 132 patients, were related to the fact that the primary endpoint consisting of the twelve-month disease-free survival for high-risk disease recurrence was 43% in the ITT population. If we look at the subgroup analysis by PD-L1 expression, stratified by CPS combined positive score expression of PD-L1 based on the DAKO platform, we realised that patients with PD-L1 high tumours benefitted with a twelve-month disease-free survival rate of 54% while for PD-L1 negative tumours the rate of the twelve-month disease-free survival for high-risk disease recurrence was 39%. 

So a signal for numerically higher maintained responses with PD-L1 positive tumours but, at the same time, the main message providing the efficacy and confirming the efficacy of pembrolizumab regardless, mainly, of PD-L1 expression. 

This is important data and these data are coupled with the confirmed safety outcomes that are in line with the safety outcome already provided in various disease settings with monotherapy with pembrolizumab, including cohort A of the study, with stable quality of life parameters over time, meaning that the combination of safety and efficacy in the all-comers and in the PD-L1 positive and negative subgroups may justify a potential expansion of the indication of pembrolizumab besides the CIS population, also including the BCG unresponsive papillary tumours.

How can the study results impact clinical practice?

It may have a huge impact because the non-CIS patient population is a consistent patient population, pretty well defined by the guidelines. The major point is that the US FDA and the regulators are requiring randomised studies with any newer therapies in order to provide a new standard of care in this patient population. Because we might have some biases related to the quality of the TURBT and some other biases, they could be adjusted just with a randomisation design rather than with monotherapy and single-arm design. The point is that the investigators are struggling against the issue of identifying the right comparator arm, the right standard of care arm, based on robust data because we currently lack data, robust data, related to a standard of care arm to randomise against any newer therapies. 

So there is a grey zone, a grey area, a lot of analysis related to the next step that is required in terms of randomisation studies besides the outcomes that we have already provided. So, I still think that there may be room for re-discussing at the multi-stakeholder tables, including regulators, patient stakeholders, patient advocates and clinicians the way of moving forward in this particular space, also in light of the other findings that we provided in KEYNOTE-057 cohort B.

Anything else to add?

In the same KEYNOTE-057 platform we are already targeting the combination therapies. Again, combination therapies are targeting the CIS population; we are not moving forward with combination therapies in the non-CIS population like in KEYNOTE-057 cohort B, but KEYNOTE-057 cohort C, which is the newly added cohort in this platform which is again targeting the CIS population like in cohort A, randomising patients to receive pembrolizumab together with anti-LAG3 or anti-TIGIT with a favezelimab or vibostolimab combination as a co-formulation with the pembrolizumab. The complete response, again, is the primary endpoint coupled with safety. 

So we will see whether combination therapies and combination immunotherapies may be able to even raise the number of complete responses and durable complete responses in the CIS population. This is actually the next goal for research in this patient population.