Alicia Morgans: Hi, I'm so excited to be here at ESMO 2022 where I have the opportunity to speak with Professor Naomi Haas about the PROSPER Trial. Thank you so much for talking with me today.

Naomi Haas: Thank you for inviting me.

Alicia Morgans:  Of course. Well, you've been involved in the PROSPER Trial for a long time. I remember hearing the initial iterations of this study and really how it was put together and quite a labor of love with you and the ECOG team. This is a cooperative group trial and really an important one. Can you tell me a little bit about the study design and what you and team members actually presented at ESMO?

Naomi Haas: Sure. So the story actually starts when some of us were at the Prostate Cancer Foundation meeting a number of years ago, and Chuck Drake and David McDermott and Lauren Harshman and I started talking about a design for a trial using nivolumab, which had recently been approved in the metastatic setting. And there was data at Johns Hopkins, which in mouse data suggested that priming the intact kidney tumor could lead to better responses, at least in mouse models that had metastatic kidney cancer. And we certainly know a lot more now than we did back then. And in the meantime, there have been neoadjuvant trials in breast cancer and melanoma and lung cancer, which have really shown some very interesting data using either single agent immune checkpoint inhibitors or immune checkpoint inhibitors in combination with CTLA-4 inhibition. And so this trial was designed, we'd hoped to design it as a three arm study, but we were only allowed to design it as a two arm study.

And so the two arms of the study were, the first arm is, and this is 804 patients, were randomized to receive either one dose of nivolumab preoperatively and then go to surgery and then receive nine cycles of nivolumab in the adjuvant setting. And the other arm patients were randomized to surgery alone. And this was really a trial that was designed with a lot of patient advocate input. And so we originally started out with patients requiring a biopsy in both arms, but there was a lot of pushback about that. And so the trial design ended up being a biopsy to confirm basically the presence of renal cancer followed by one dose of nivolumab, followed by surgery in the treatment arm and then going on to adjuvant therapy. And there was a good faith effort. In other words, if the biopsies were not definitive, we still allowed patients to enroll in the trial.

So patients had to have clinical staging in order to enter the trial and they had to have clinical stage T2 cancer, which means that their cancer had to be bigger than seven centimeters. And so by design, we allowed both clear cell and non-clear cell to enroll in the trial. And the primary endpoint was recurrence free survival and we did as a secondary endpoint power for clear cell as well as for overall survival and a number of correlatives. But the primary design was recurrent free survival for everybody that registered. So that was both clear cell and non-clear cell.

Alicia Morgans: Very important, I think. And these are important nuances of the trial design that help us to understand how things turned out. And of course, this sort of perioperative peri-nephrectomy approach is I think one that's been really appealing. But what were the study results reported at ESMO?

Naomi Haas: So unfortunately, the primary endpoint recurrence free survival did not show a statistically significant benefit. And the data safety monitoring committee actually stopped the trial early because based on some of the statistics, which as to assess for futility, they saw that even if the trial were to continue, it was not going to show a benefit to receiving nivolumab in the study arm. And so we are a busy looking at a lot of the different parts of the trial, and it's a very interesting trial. And what we know so far that was presented at the meeting this morning were several things. We know that patients, some patients did drop out before they actually had surgery or even nivolumab treatment, and that was in both arms of the study. So one of the things that I guess we're wondering is that maybe there were patients who were randomized to the nivolumab arm that got nervous and didn't want to be on that arm, and maybe there were patients on the surgery arm alone that wanted nivolumab, and so they dropped off.

And then additionally, we noticed that there were also some patients who received a dose of nivolumab and didn't go on to surgery. And so we wonder whether those patients, perhaps some of those patients, maybe they had a scan review that showed that they had a benefit and they decided they didn't want anymore. Or maybe they had side effects or maybe they just didn't know why. There's a lot of different reasons and we're in the process of, those are things that we don't have answers to yet, but we're in the process of analyzing all of that. And then there were also patients who had both nivolumab and surgery and didn't go on to receive adjuvant nivolumab. So there were some drop offs in the trial.

But if you look at the overall recurrence-free survival, the curves are really right on top of each other. And so it's really hard to think that even without that it would've been a positive trial. And we're very interested in looking at the correlatives, which I think are going to be really, really useful in informing future perioperative trials. I don't think by any stretch that we shouldn't do another neoadjuvant trial, but I think that there's a lot of tools that will help us in the future. So I'll give a few examples of those. We're collecting CDs of all of the imaging, and I think that'll be useful for radiomics. We are currently still finishing the central path results, and I think that that will be informative as far as the degree of response that patients had on the study, as well as the actual composition of the trial.

We are looking at PD-L1 expression. We are also doing a whole slew of correlatives. Sabina Signoretti in Boston is going to be doing some of the immune profiling that has previously been reported with things such as the IMmotion trial. And in the paper cancer cell that came out a couple of years ago, that looks at the seven clusters of renal cancer that are based on some of the T-cell signature as well as some of the bioinformatics that, or heat maps that we use and seem to suggest, at least for immune mediated therapy, that you could almost predict who would respond to immune therapy or not. And I think that'll be really, really important.

And in addition, we did collect blood for cytokines, we collected blood for methylated DNA, and we're also measuring some other things like KIM-1, which was previously validated in the ASSURE trial. And it does seem to correlate, at least in that trial, with recurrence free survival. So I think that all of those things are things that we didn't have to begin with. And as we get smarter, as the world gets smarter, I think that there are a whole bunch of tools that we can use to better select patients who might be good candidates for both neoadjuvant and adjuvant therapy in kidney cancer.

Alicia Morgans: So that is exactly where we need to go. And it's so interesting to hear about your thoughts on where we might be continuing to investigate. I wonder, as I think about this PROSPER data, which is unfortunately disappointing, but has so many aspects that still need to be investigated to understand why. How do we put this in context with the pembrolizumab data suggesting a benefit for adjuvant treatment in the high and very high risk patient population with kidney cancer?

Naomi Haas: So, I think there's some important things. There's basically four adjuvant trials that we're going to have information for. There's IMmotion-010, there's CheckMate, which is going to be presented tomorrow. There's the KEYNOTE-564, and then there's the PROSPER trial. And the PROSPER trial is really by design, very unlike the other three adjuvant trials because it's the only one looking at the neoadjuvant approach. But the patient populations are going to be quite heterogeneous between all of these. So although we did allow later on in the trial some enrollment in patients who had metastasectomy, they had to have their metastasectomy within that 12 week period. So that's very different, for example, than KEYNOTE-564 where they were allowed to have metastasectomy up to a year and very different from IMmotion where they kind of didn't want to have any of the mastectomies close to resection of the primary.

And then CheckMate, I believe, I could be wrong about this, but I believe did not allow metastasectomy in their patient population. So that's one thing. The other is that we had a mixture of both clear cell and non-clear cell. We're not sure at this point. The pathology I will say when we looked at tumor pathology, at least by investigator review, more than 60% of our population did have very high risk disease. So at least pT3, which means tumor, that's invading either the renal pelvis or the renal fat. So we did have that population in more than 60%. But what we don't know is were some of the patients that were enrolled and got neoadjuvant therapy, did we actually downstage some of those patients? Because we did have some patients that had what we call TX disease and T1. So were they downstaged or were they misrepresented on their scans?

So there is quite a bit of heterogeneity in the populations. There's also the differences between nivolumab and pembrolizumab. And I think that it's going to be difficult even with the results of CheckMate to know that for sure, because the CheckMate presentation is only including the Ipi/Nivo arm. And it was number one, a shorter six month duration. The combination of the ipilimumab and the nivolumab was given in an every six week dosing, which is not the standard dosing that has previously been used for renal cell at least, but seems to be a more tolerable dose. And we won't have the data from their single arm nivo because they're still waiting for that data to get a little bit more mature.

Also, our trial is fairly young in its follow up. The follow up was about 18 months. And what we see for IMmotion, for example, is about 38 months. And what we see for pembro is so far the 30 months of, and by disclosure, I'm on the paper for the KEYNOTE-564 trial as well. So I did enroll patients to that trial. So I think there's a number of differences, but I think the provocative things are wondering about the activity of pembrolizumab versus nivolumab. I think we don't know the answer. It's certainly easier to give pembrolizumab because you can give in every six week, although everybody on the trial was on a three week dosing. So that probably didn't have much to do. Patients still had to come in.

And then the other question that I've been asked previously is, did COVID interfere with follow up? And I think it did. I think it interfered with all of the adjuvant trials. And so you must wonder, patients couldn't come in and be seen physically for exams. They had to be evaluated by sometimes by video. Sometimes they couldn't come in and get their dosing and stuff. And so I think that was very challenging for all of the trials in this era.

Alicia Morgans: I could not agree more on that. But I really appreciate you walking us through some of those aspects of the study that make it difficult to compare all these trials to each other and of course, the challenges that were faced and the nuances within the PROSPER Trial. So thank you so much for shedding light on this study. Congratulations to you and the team for doing the work, getting the patients on study despite COVID. And again, congratulations on the team's excellent presentation at ESMO 2022. I appreciate your time.

Naomi Haas: Thank you. And I just wanted to add one other thing, which is that I certainly hope that all of the different phase III trials that we can get together and compare our correlatives because I think that will ultimately really help to flesh out what the ideal patient population is for adjuvant therapy.

Alicia Morgans: That's an excellent point to end on. I absolutely agree, and I hope that that happens too. Thank you so much, Dr. Haas.