Pedro Barata: Hello and welcome. My name is Pedro Barata. I'm a GU medical oncologist. Today I have the pleasure to have Dr. Hans Hammers of UT Southwestern GU Medical Oncology. He's truly a legend and a giant in the kidney cancer world. And today we're going to be chatting a little bit about his important paper that came out at JCO on these phase II trial with nivolumab in salvage ipi/nivo for patients with treatment naive advanced clear cell RCC. So Dr. Hammers welcome.

Hans Hammers: So thank you so much. I think it's a great, great work. First of all, the person who conceived this and gets really the credit for this investigation trials actually Mike Atkins. And I remember the time that we both fought for BMS supporting this because it was a real lack of data of single agent activity of PD-1 inhibition. And so I think that's one of the major data sets, quite frankly. When it comes up to the question, what does PD-1 monotherapy in particular nivolumab do in previously untreated patients across the different risk groups? And nicely, I would say, compliments the data set that we have with the Keynote 427 studies. So this was 123 patients, clear cell carcinoma across all risk groups and had a sizable portion of good risk. The larger proportion as usual was intermediate risk and then some poor risk patients, some patients with sarcomatoid.

There was also a biomarker component looking at PDL-1 expression levels that was done at Harvard. So there was the basic design and then we built in at the time, and that was probably me pushing a little bit for that, was hey, let's also see if we can have this risk adaptive approach by adding nivo/ipi in patients who have either stable disease for prolonged period of time that was defined as roughly one year stable diseases best response, or patients who have progressed to then exposed to nivo/ipi as a salvage regimen because of the known increased autoimmune toxicity profile with nivo/ipi, whereas PD-1 monotherapy. So there was a basic design, 120 patients started and then depending on response, patients could go on to nivo/ipi by itself. And we would learn a lot about single agent activity and the combination.

Pedro Barata: So I mean there's a beautiful summary. I mean this is perfect. I actually love this proof of concept study, right? Because as you said, it confirms the value of PD-1 in the advanced setting and kind of builds upon the data we had from pembrolizumab with Dr. McDermot and the group. So that's great. That's number one. And number two, I remember we were very excited when we saw this data the first time I think was at ASCO, I believe because really excited because we're starting to talk about, we have a few setups or data sets that are set with ipilimumab as salvage at any point. And so maybe that's the question here. You know, start everybody PD-1 monotherapy and then you bring ipilimumab later on for those who seem to need it. Right. Where nivo was not helpful, I'll let you comment. I mean, just tell us what's your sense from bringing ipi later on versus upfront?

Hans Hammers: Yeah, no, so it's still something that we discussed with some degree, but I would say first of all, I'm going to put it this way. There is activity, albeit not satisfyingly high, but certainly I would say measurable and I think consistent activity of nivo/ipi or PD-1 CDK4 inhibition after prior of prior PD-1 progression. So in this trial was around 11%. We know that ipilimumab single agent activity in renal cell carcinoma around 10, 15%. And there were other trials like OMNIVORE, for example, TITAN RCC, the fraction trial. So I would say in aggregate, it nicely supports the thinking that yes, in select patients you may be able to rechallenge with this particular combination. So there is activity, albeit 11% is not very high, and you may want to be careful with how you select patients. The other experience was there was less toxicity if you gave nivo/ipi after prior PD-1 exposure.

And I think that wasn't too surprising because if you were meant to have had a bad side effect, some of it was sort of say already encountered with nivo monotherapy or PD-1 monotherapy. And so there was certainly less than if you were completely treatment naive. But I think the truth is not everybody in the trial made it to the combination, not because they died or anything, but the disease progression was maybe felt to be too aggressive. Patients couldn't get re-biopsied. Some patients because of the autoimmune toxicities there weren't really illegible to be re-challenged with the combination. So there's some attrition that was concerning that if you really wanted to go after the immunotherapy effect and that you might lose some of them. The surprising data or the surprising dataset was also that quite frankly there was a fairly high response rate in good risk patients was around 50% or so.

Again, good risk is one group of patients that we have we're struggling with, I think to really identify well used to be or is generally thought of a population that might be a little bit less immuno responsive. In this particular study response was fairly high. But I think what's concerning a little bit was that for the classic candidates for a nivo/ipi intermediate and poor risk disease sarcomatoid component, the response to nivo monotherapy was less than what we had hoped. And so we do feel that if you think about combination therapy for those patients, you might be better served with just using it upfront. I think that's the general gist of the adaptive design and experience with this treatment approach.

Pedro Barata: Right. That's a great, great points. And actually I guess that's a great segue, but my next question, which as I was reading the paper and going through the data, I was thinking to myself going beyond PDL-1, and that's what you and others with a lot of work have been doing, right? And so it's interesting to see that immunotherapy seems to work regardless or irrespective of the risk groups as to your point with DC activity in good risk, we actually saw that in with ipi/nivo as well and wondering, how can we explore that further? Is that genomic signature the way to go?

Do you have any further planning as far as correlatives for this study? And if so, how can we bring that biomarker study and work in the future in the next generation of studies to prove that point? Right? Because it seems like not one size does not fit all, and irrespective of the risk, it sounds like we should move on to maybe molecular selection and offer to those patients based on that molecular phenotype, what they should get. Do you agree with that? What are your thoughts?

Hans Hammers: Yeah, I wish I had the answer. Let me start like that because I think it's one of the most complicated topics that we have. Just remember, we've been doing tyrosine kinase inhibitors for forever and we still don't have a biomarker for that. And the immune system is, the complexity of the immune response is just such a difficult topic that I do think a single marker will not work. I think it's good enough to enrich for some population that we've seen that in other indications, for example, where PDL-1 expression has been used to identify the cords of patients who benefit the most. But even for PDL-1 expressing patients, you still have sizable response rates in renal cell carcinoma, even with complete responses and long-term benefits. So very, very difficult. I think the general feeling is that a single biomarker won't work. You need to look at different factors, maybe even factor in tumor heterogeneity across the tumor body.

Remember we just take one piece of a tissue, we may take the old nephrectomy tissue. And the truth is the patient has 10 or 20 different tumors if they have overt metastatic disease. And so it will need to be an aggregate of many different candidates. Even I would say the currently used signatures have difficulty performing well. So right in particular, if you use a signature developed in one combination may not work in the other combination. And so we are struggling with this. I think that's the fair answer, not a reason not to try and continue, but I think we need to take a step back and say, Hey, how do we best represent the disease in the patient? And maybe even combining imaging studies with biomarker correlatives, et cetera. So it's difficult.

Pedro Barata: Right. No, that's really well said. I mean, you raise very, very important points that your trial actually opens the door to a lot of unanswered questions. And with that said, I think was a beautiful effort. It was really, really smart. The design you put together able to enroll well, and to get that data out there I think is really important, as you said, because it builds data on everything else that we know about salvage for inhibition. So I think that's very, very helpful. And this is a very interesting conversation. I feel like we could stay here all day talking about it. Do you have any final points you'd like to make or take home points from your paper? By the way, I invite everybody who's listening to this conversation to go and read the papers. It's really well done. The discussion is quite good.

Hans Hammers: Yeah, no, I mean, I think it's beautiful work. Again, Dr. Atkins was the brain of this, and this was an investigator initiated trial. So kudos to all the people who have participated. Worked hard on this. This was performed to the Hoosier Network, which I think was a fantastic organization to work with. So I think it's worth a read, and I hope that we can learn something from it.

Pedro Barata: Absolutely. Well, Dr. Hammers was a pleasure. Thank you for joining us, taking the time to sit down with us and give us kind of the breakthrough of the paper. Congratulations, very important data to be out there, and I hope to see you soon.

Hans Hammers: My pleasure. Bye-bye.