Petros Grivas: Hello, everybody. I'm Dr. Petros Grivas. I'm a medical oncologist in Seattle, Fred Hutch Cancer Center, University of Washington.

I'm very excited today to discuss with a dear friend and colleague, Dr. Michiel van der Heijden, who is a GU genitourinary medical oncologist, and a great expert global leader in the field of GU cancers. He's at the Netherlands Cancer Institute. Michiel, welcome.

Michiel Van Der Heijden: Thank you, Petros. Good to be here.

Petros Grivas: Great to have you. And great to discuss with you. You have been working together over the years, seeing the field being transformed over time. You have been a big part of this transformation.

I want to start asking you about the CheckMate 901 trial, and the data you presented at the ESMO 2023, practice-changing data with gem/cis nivolumab versus gem/cis, just to give us a little bit of background of that randomization as part of the CheckMate 901, before we pivot to the ipilimumab/nivolumab data.

Michiel Van Der Heijden: Yes. So CheckMate 901 is a large study, multi-arm study, that started actually in 2017. So it's already been a while. And this study actually had two parts.

It was initially set up for the ipilimumab/nivolumab comparison, but what we at that point called the substudy, later was expanded. And that's the part that the cis/gem plus or minus nivolumab, that was presented in 2023, and was positive, and has now been registered.

That part of the study randomized patients who could receive cisplatin between cis/gem, with or without nivolumab.

Petros Grivas: And I remember from your presentation, Michiel, at ESMO. Of course, this was a practice-changing trial with gem/cis nivolumab prolonging OS and PFS compared to gem/cis alone.

And this is now approved also in Europe, not only in the US. Do you have access in the Netherlands for gem/cis nivolumab?

Michiel Van Der Heijden: Well, actually, in the Netherlands, we don't have access yet. I'm not sure if it's coming. But we don't. And neither for enfortumab vedotin–pembrolizumab yet.

Petros Grivas: And that shows the global differences in the access and reimbursement of different agents. And I'm sure the same may apply to pembrolizumab, enfortumab.

You were the senior author on that paper with Tom Powles presenting the data. And enfortumab vedotin–pembrolizumab was the preferred standard of care in frontline metastatic urothelial carcinoma, is not reimbursed in many countries. So just showing the talents and the differences across countries.

Michiel Van Der Heijden: Yes, absolutely.

Petros Grivas: Having said that, Michiel, you have done the enfortumab vedotin–pembrolizumab work with Tom Powles and the team. You have worked on the CheckMate 901. You mentioned that the CheckMate 901 had a different randomization, ipilimumab nivolumab versus platinum gemcitabine. I'll let you talk us through about that randomization.

Michiel Van Der Heijden: Yeah. So this is the complicated part. So it has become a really big study of 1,300 patients in total, including the cis/gem nivolumab part. And basically, patients who were cisplatin ineligible were randomized 1 to 1 to receive either ipilimumab/nivolumab for four cycles, followed by nivolumab versus platinum chemotherapy.

And in the case of cisplatin ineligible patients, that was gem/carbo, of course. The patients who were cisplatin eligible were randomized equally between four arms. So cis/gem nivolumab, cis/gem, again, cis/gem and ipilimumab/nivolumab.

So there were some patients who were cisplatin eligible, and who also ended up in the ipilimumab/nivolumab comparison study. However, the main primary endpoints-- there were two primary endpoints. The main primary endpoints were on the cisplatin ineligible population. And that's the data that we present here at ASCO.

Petros Grivas: And in that comparison we have ipilimumab/nivolumab with nivolumab continuation maintenance, and on the control arm we have gem/cis or gem/carbo, specifically gem/carbo for cisplatin ineligible patients who present here at ASCO. And this was without any maintenance, at least as part of the trial design for the gem/carbo?

Michiel Van Der Heijden: Yes. Correct. Remember, this trial started already in 2017. And when it was enrolling most of its patients, maintenance nivolumab was not even published yet.

Petros Grivas: Of course, of course. And again, this was eight years ago. And kudos to getting through a cruel design, of course. And now, reporting the results here at ASCO. Can you summarize the results for the audience, what was presented at ASCO here in Chicago?

Michiel Van Der Heijden: Yes. So, unfortunately, it was a negative study. If we look at the data for the main comparison. So in cisplatin ineligible population, the hazard ratio was 0.79 with a p-value of 0.025, which barely missed the p-value boundary.

So what happened in this study is that the alpha was spent between two primary endpoints. One was the cisplatin ineligible endpoint, allocating 0.02 of the alpha. And the other one was the PD-L1 positives, which allocated 0.03. It was a smaller group, so it was more ambitiously powered, so to say.

So even though the study is negative, I think there's still a lot of things we can learn. What we did observe was that there were really durable responses in quite a few patients who were receiving ipilimumab/nivolumab.

And actually, if you look at the Kaplan-Meier curves, you see that the curves run together for quite a bit. But then after one year, they start to separate, and they remain apart for the remainder of the study with about a 10% difference.

If we look at the responses, we see that they are more or less equal between chemotherapy and ipilimumab/nivolumab. But the duration of responses is very different, with duration of response in ipilimumab/nivolumab being much longer, and especially in the patients who had a complete remission.

So there was a good subset of patients who had complete remission, and remained in remission until the moment of study analysis.

Petros Grivas: Very interesting. And despite being a negative trial, as you mentioned, with no impact on practice, there's still very interest, high interest, I would say, and lessons learned in this subset of patients with durable response.

It seems to me that PD-L1 as a biomarker has not paid many dividends. Any comments on these potential biomarkers? The work you're doing to find, to select those patients with durable benefit with ipilimumab/nivolumab?

Michiel Van Der Heijden: So how much time do we have to talk about the biomarker? So what we've seen so far, and I think that's one of the major lessons from this trial, is that when we do these trials, and we try to beat chemotherapy, if we slice up the subpopulation based on what we think we know, we get into trouble.

So we end up with a population study, populations that are, in the end, too small to show the difference between two treatments. So we end up with underpowered comparisons.

And in this case, in PD-L1, there was good rationale to think that it would work. The single arm study seemed to be better for the combination of PD-L1 positive patients.

And remember that the DANUBE trial was a trial where the primary endpoint was actually reversed. So the intention to treat was the primary endpoint for the durvalumab/tremelimumab. Also, anti-CTLA-4.

And there, actually, the PD-L1 positives look better. But here we see the reverse. Because if we look at the intention to treat population in this study, we actually see a hazard ratio of 0.78 in 700 patients. However, we were not allowed to formally test the comparison because of the statistical design.

So I think the lesson learned here is actually twofold. One is that if we do these trials, especially against the comparator that is in the beginning of treatment, has a very good tumor response and tumor control. We need large groups of patients. We need the larger set of patients to do these comparisons.

The second thing is that we can wonder if, as you know, if a patient sits in front of you, they're mainly interested in am I still going to be around in five years from now?

So if we have treatments that in a smaller subset of patients can have real benefit on the long-term, then perhaps we should look at different endpoints, or we should look differently at these types of trial, or design them differently.

Ideally, we would want to have biomarkers to pick those patients who are going to have that long-term benefit.

Petros Grivas: Great points. And speaking about biomarkers, I know you're working hard with your great team there in the Netherlands on neoadjuvant trials. And the NABUCCO trial comes to mind. And I know you're working on a few others.

Can you tell us a little bit more about how this program is going with the neoadjuvant approach, and how you utilize tissue before and after to inform the dialogue about putative biomarkers?

Michiel Van Der Heijden: So the NABUCCO trial tested ipilimumab/nivolumab in the neoadjuvant setting. We found a couple of things. We found that we can reach very high rates of complete response, even in patients with still operable, but relatively advanced disease. So up to N3. So Stage III urothelial cancer.

We also saw, in a smaller cohort, we tested a lower dose of ipilimumab/nivolumab, because there were some preoperative studies that seemed to suggest that would be sufficient in that setting. But that didn't seem to work.

So, as we have seen in the metastatic setting, a lower dose of anti-CTLA-4 seems to not give IO response bump, as does the higher dose.

I think, for the biomarkers, it's been a bit complicated. So what we found is that pre-existing T-cell immunity did not really predict for response.

Biologically, we think that that means that where monotherapy, you already need to have some of pre-existing immunity that's not needed maybe for combination therapy.

However, we did see a bit better response, again, in the PD-L1 biomarker, which didn't turn out to work in this trial.

So I think it still remains complicated. One of the other things we see quite consistently across trials, that stroma TGF-beta type signature predicts for non-response.

So I think that's still one of the points that we should perhaps get better therapies for.

Petros Grivas: It's very interesting to see this program that you are doing with the neoadjuvant approach. Across different tumor types. And I think that can inform hypotheses, as you said.

And maybe the tumor microenvironment may be different across tumor types, and maybe some common lessons learned, maybe some differences. But I think it's so important to do those trials.

And even from negative trials, as you pointed out, that ipilimumab/nivolumab versus platinum-based chemotherapy, you can still tease out lessons about biology.

So congratulations for doing this work. It's hard to do those trials. And of course, as you mentioned, kudos to the patients and families who participated in those trials.

Any last few minutes words here from your experience about this trial, and overall in the field?

Michiel Van Der Heijden: I think one of the things that are going to be very important for the near future is, we have, of course, now also enfortumab vedotin–pembrolizumab, which is a great new therapy for urothelial cancer. And I'm quite confident that it will also show benefit in the neoadjuvant setting.

We also have the Niagara regimen. And all of these new therapies give a much higher tumor response in the neoadjuvant or induction setting.

And I think we really have to start looking for what patients want. And that is very often to keep their bladder. And I think with these therapies, it's often possible.

So with the ipilimumab/nivolumab combination in the preoperative setting, we already took a bit of a leap, and started giving that in a trial called the INDIBLADE trial, as an induction therapy, which was then followed by chemoradiotherapy.

And I think those types of trials, there's also other people, for example, Magalski, who experiments in his studies, with not giving any consolidation therapy.

And I think we really have to tease out what is best for which patient. Do we always need to give consolidation therapy? Some patients will still need, for example, chemoradiotherapy.

But I think it's possible to save the bladder for many more patients, and still have good outcome.

Petros Grivas: That's a great point. And I think the patient advocacy groups really ask us to do those trials to try to tease out, based on biomarkers, who are the patients, which are the patients who may benefit from a bladder preservation approach?

Could we potentially select out patients with long-term disease control after systemic therapy? These are very important questions. And we're all excited in the field to do these trials, and answer those important questions.

It's very exciting to see you, my friend. Great job. Congratulations for the presentation and all the work you're doing in the field. And thanks to the audience for attending this wonderful discussion.

And I always say that the ASCO Annual Meeting is high yield, very important presentations. I think it's very valuable to attend the sessions, virtually or in person. Thank you so much for your attention and all you are doing.