Alicia Morgans: Hi, I'm so excited to be here with Professor Tom Powles, who is joining me to talk about the COSMIC-313 Phase III trial. And of course, Dr. Powles, Professor Powles, is a professor of GU Oncology and the Director of the Barts Cancer Centre at St. Bartholomew's Hospital. Thank you so much for being here with me today, Dr. Powles.

Thomas Powles: It's lovely to see you. How are you?

Alicia Morgans: I am well and it's lovely to see you too. And thank you. You presented this update at ASCO GU 2023 and I'd love to hear some of the data. Can you set us up by telling us what this study involved? This was a first-line clear cell RCC trial, right?

Thomas Powles: Yes, it was. So COSMIC-313 explores the triplet of cabozantinib plus ipilimumab and nivolumab versus the standard of care, which is ipilimumab and nivolumab in clear cell renal cancer patients with intermediate and poor risk disease.

It's the first study looking at a second-generation control arm, a doublet ipi/nivo as the control arm, which as you know, is now superseded sunitinib. So it's really contemporary and super cool. And you'll be aware that Toni Choueiri presented that of course at this plenary session and at ESMO. And what he showed in a population of 550 patients was about a 25% reduction in the risk of progression, which was statistically significant and positive. So it's a positive randomized Phase III, with a hazard ratio of 0.73 in the PITT population, the initial population.

What the data also showed was response rates for the triplet of about 45% versus 35% for the doublet. CR rates were lower than we expected at about 3% and we didn't see an overall survival signal.

And at that point we also then looked at the toxicity. There was a bit more toxicity with the triplet. It was a bit harder to give and it wasn't impossible. And we were able, obviously, a trial was positive, but that is a consideration. If we're going to give more and more drugs, we have to be hitting really important endpoints. And that's why I think overall survival is super important in this study. We haven't yet seen a survival signal and we're looking forward to that in the future.

What we showed at ASCO GU was an update of this. Now, two parts of the update. Number one, five further months of follow-up, but also there's a larger population, 850 patients for the ITT population, which we're looking at for OS. And we showed the progression-free survival signal there and it's still 0.74 with a larger number and more robust follow-up.

So there's this consistent 25% reduction in the risk of progression. The response rates didn't change and the toxicity didn't change much as you would expect.

There were two or three intriguing parts, or one or two intriguing parts. One of those intriguing parts is that we showed in the subgroup of patients, the poor risk subgroup of patients by IMDC classification. We showed that the benefits seemed less in that subgroup with a hazard ratio of 0.93 than in the intermediate risk group where the hazard ratio was 0.68. That was in Toni's presentation, but the numbers were super low and it was really exploratory.

So we wanted to look at that further because it was a bit unexpected, because you might think it's the poor-risk patients that need the triplet and give the poor-risk patients the strongest therapy. What we actually showed in the update with an additional 350 patients was the same trend. So the benefit is confined mainly to the intermediate risk subgroup and less benefit in the poor risk group.

And so what we did next is we then looked at the toxicity profile in those two subgroups. Because you might say it might be the poor-risk patients, less poor performance data, comorbidities, not able to get the triplet in. But it wasn't that at all. Actually, the exposure to the drugs in the two groups was almost identical.

And so if it's not decreased exposure, is it actually decreased efficacy? Well, what is there a biological rationale for that? And there's a little bit of work going on and Dave McDermott started this and others have, Brian and others have pursued it further, where we're showing the good risk patients seem to have a stronger VEGF signature associated with their disease and therefore they're responding better to drugs like VEGF, single agent VEGF targeted therapies. And the poor-risk patients have a more complex tumor and are more likely to respond to immune therapy. So there is a biological rationale why the addition of VEGF-targeted therapy might not add very much.

And then you've got two other groups of people who have their own hypotheses. And Hans Hammers, who's a great friend of mine and a real guru of immune therapy. His take on this is actually the dose of cabo was too low, too many dose reductions, too difficult to give. And so you never gave the VEGF targeted therapy properly. I'm not totally convinced by that, but Hans is normally right in the end. So we'll see.

And then of course there is the statistical purists who come to me. They'll say, "Hold on a second, Tom. This is an exploratory subgroup of a handful of patients, a hundred patients. It's not statistically robust or indeed statistically significant. And this is hypothesis generating and let's wait and see other signals like survival before we get carried away by retrospective exploratory analysis."

Alicia Morgans: It is so interesting though, and as you said, in those poorest-risk patients, certainly you would hope that more intensive therapy would result in an improved response there. Certainly, they have the poorest prognosis. And to your point, maybe we just are not choosing the right thing to target the proper pathway.

And it's interesting Hans' comment, especially since you said that exposure to the drug was similar between the two groups. Do you think it's possible to safely give a higher dose of something in this setting, cabozantinib, that does have toxicities and did add to the toxicity burden here? What are your thoughts?

Thomas Powles: So Hans' comment was around the intermediate and the poor risk comparison. I think when you look from the back of the room and say, I know you obviously are a prostate cancer guru, and the triplet therapies there, I think they're a bit easier to give. I think LHRH agonists, I think abiraterone, I think that's relatively straightforward.

VEGF targeted therapy, you come in sore hands, sore feet, sore mouth transaminitis, fatigue, hypertension, and then you put on immune therapy on top of that. And then you've got the complexity of immune-related toxicity. You've got diarrhea which goes up. The transaminitis is a lot worse and suddenly you're in the realms. It's not lymphoma, but you're in the realms of sort of bladder cancer treatment, not prostate cancer treatment, where you're coming in all the time and there's delays and there's dose reductions.

I think it's also worthwhile saying, I don't talk about this very much, but the trial, COSMIC trial 313, had a blind in it, which is great. Blinded trials, terrific. Particularly if progression-free survival is the endpoint. The problem with a blind is with immune therapy. If you've got immune therapy, immune therapy, immune therapy, cabozantinib, and you don't know if the patient's getting placebo or cabozantinib.

If they come in with grade two diarrhea, which is probably the cabozantinib speaking, frankly, you can't be sure about that. They could be on placebo and ipi/nivo. So that makes it complicated, because if you've got diarrhea and we don't know if the patient's on cabozantinib or placebo, you'd have to assume they were on placebo because you obviously wouldn't just stop a tablet with grade two diarrhea with immune-related toxin. So you then have to give those patients steroids or unblind them and it makes it more complicated.

And I think therapies with blinds in them make it hard for investigators. And of course, investigators will always err on the side of caution. But there are a new generation of trials without blinds in them. Now that's going to make the PFS signal more complicated. We'll have to have a central review and all the bits that go with that.

But in the end, we're looking for overall survival in these trials anyway. And it looks like we're not going to use them unless we show CR rates of 25% or something out of the park. We're going to have to get OS in kidney cancers, in my opinion. So actually, I think we are better off doing open-label trials, having central radiology reviews than trying to give challenging regimes with a blind.

Alicia Morgans: I think that's a great point. And I think certainly from a patient-centric perspective, it's really important when you're managing that diarrhea to know if the patient's on cabozantinib or not. I think it makes a big difference and it may make a difference even in terms of certainly hospitalization, high dose steroids and all of those things that would come if we really were only maybe focusing on the checkpoint approach versus the cabozantinib approach where we may be doing something a little bit differently.

So this is all something that I guess we'll have to see if this happens. If we're able, and you in particular, are able to influence the teams, the companies to design the trials in that way. And the central read I think would help PFS, but to your point, it is challenging when we have PFS as an endpoint.

So as we start to wrap up, I'm wondering does this influence your practice today, this new update on the data and certainly the original data from COSMIC-313 and where do we go from here?

Thomas Powles: So, I'm not going to give the triplet until we've seen that OS signal, and I don't think it's the right thing to do. I think we're currently still sequencing the doublet followed by single agent therapy.

We saw the results of a trial called CONTACT-03, the results, we got a press release showing that the cabo plus atezo didn't have a progression-free survival advantage compared to cabo alone in the second-line setting after immune therapy. And so I'm beginning to sort of feel that this field is evolving into choosing first-line therapy correctly and giving it well is really important, that the immune therapy got an incredibly important role to play. We don't quite know what it is. I think we're still in the era of equipoise between ipi/nivo and VEGF-TKI PD-1 combinations.

I don't think the triplet has been practice-changing without this survival signal. We need to wait for that. And I don't think that PD-1 therapy forever is the right answer to this either. I think once you've had one go at it with PD-1 therapy and progression has occurred, it doesn't look like sequencing these immune checkpoint inhibitors, unlike the VEGF-TKIs, is associated with significant benefit.

Alicia Morgans: Well, that is certainly a wonderful way for us to put it all into context, and I really appreciate your time, your expertise, and your thoughts about this data. We look forward to talking to you about the next update in the future. Thank you so much, Dr. Powles.

Thomas Powles: Thank you very much. It's lovely to see you.