Pedro Barata: Hi there and welcome. I'm Pedro Barata. I'm a GU medical oncologist at Seidman Cancer Center, Case Western University in Cleveland, Ohio. It's my privilege to welcome Dr. Michael Atkins. Dr. Atkins is obviously known from everybody in the kidney cancer world and beyond, melanoma and beyond. He's the deputy director of the Georgetown Lombardi Comprehensive Cancer Center, also vice-chair of Department of Oncology, is at Washington DC at Georgetown University Hospital. Dr. Atkins, it's a true privilege to have you here today, and thank you for taking the time.

Michael Atkins: It's a pleasure to be here, Pedro. Always nice to talk to you.

Pedro Barata: Thank you for that. First, congrats. We'll be talking a little bit, we're going to be shedding this morning a little bit about TIVO-3 trial. Very important phase III trial led to the approval of tivozanib and it's really, really important, as we see these trials reading out, to see what happens after more or longer follow up. So you did present results for long-term progression-free survival, which, I believe, was the primary endpoint of this trial. I think you presented quite interesting findings for tivozanib that was compared with sorafenib in that trial. Can you briefly comment or tell us, briefly, the findings, as far as efficacy, and also from the safety profile, that I think is quite remarkably for tivozanib in this particular trial with longer follow up?

Michael Atkins: Sure. The TIVO-3 trial was a trial that really was in patients with third-line, fourth line treatment, and it compared tivozanib, which is a relatively new selective TKI, with sorafenib, which is a drug that we used a decade ago but is really gone out of favor because the other TKIs are better. But when you get to the third and fourth-line, sorafenib was one of the few drugs that was left.

In this randomized phase III trial, tivozanib showed, early on, a better progression-free survival, median progression-free survival, than sorafenib, and because it's a more selective drug, was also better tolerated. But it was confusing because the overall survival hazard ratios were not different. In fact, the hazard ratio for tivozanib versus sorafenib was above 1. This was of concern, and we watched and watched and watched the data for a while. And then over time, the curves cross a hazard ratio. Became less than 1. This poster presentation was an opportunity to look at what was actually happening later on in these patients who were being treated on these two arms.

What we saw was that patients on the sorafenib arm were continuing to experience progression and continuing to die, while there was a subset of patients that was out at three and four years who were continuing to remain on tivozanib and remain progression free. I think it was three or four times the number of patients, even though the number was small, at three years, and at four years, nobody was progression free on sorafenib, while there was around 8-10% of patients who were still progression free on tivozanib. This was just showing that patients were getting long-term benefit, a small subset of patients were getting long-term benefit, from tivozanib, while that wasn't happening with sorafenib. And that's why the overall survival numbers, as we followed further out, that hazard ratio was continuing to decline.

So what was happening early on that made that hazard ratio for overall survival worse with tivozanib? I think we saw something like this TIVO-1 as well. And I think when you do studies in situations where you're in countries where there's an imbalance in what people can get after a particular therapy, it may influence what the overall survival data is.

Pedro Barata: Actually, that's a perfect segue, Dr. Atkins, because you touched on very important points, in my opinion, on TIVO-3 and goes beyond TIVO-3. Maybe I'll ask you that because it came to mind. You're really highlighting, first of all, what you saw in TIVO-3, this long PFS, and to your point, I think you got like 12-7% at three or four years on patients who remain without progression on Tivo versus 0% on sorafenib, which is quite interesting, because we always see those patients who stay on and on a TKI, they're definitely responders to TKI long-term, and in my opinion, that highlights the value of tivozanib as an active TKI out there.

But then you touch base also on a very important point, which has to do with TIVO-1, why TIVO-1 was not considered positive. What, it got approved in someplace in Europe, not in US. But then we got TIVO-3 and we actually took rPFS. So I guess my question is, as a broader question, I'll get to back to TIVO-3 in a second, is, we do care about rPFS. Actually, that has been an acceptable endpoint for the regulatory agencies to approve these agents. I guess how in the context of so many different therapies available out there, and I guess TIVO-1 is an example of that, where subsequent lines of therapy can definitely impact the rest, what are your insights about rPFS as an endpoint? There's one question. Now, the question is, is it fair that we continue to believe that rPFS is probably the best endpoint out there beyond overall survival? What are your thoughts about that?

Michael Atkins: Yeah, well, I think when you're dealing with palliative therapies, and I think the VEGF receptor TKIs, by and large, are not curative, they're palliative, they're great drugs, but they're not getting rid of the last tumor cells. I think PFS and the median PFS is probably a good surrogate for showing efficacy and probably correlates with overall survival reasonably well if all other things are balanced.

You also have to take into account toxicity when you're giving palliative therapies. Tivozanib is probably the TKI with the best therapeutic index, because it focuses directly on the VEGF pathway and has a lot less off-target toxicities. So it's a good drug to use for palliation. But a lot of our drugs right now in kidney cancer, particularly first-line treatments, involve immunotherapies, and immunotherapies, PFS is a really bad indicator of long-term efficacy because immunotherapies help a smaller percentage of patients than the VEGF-targeted therapies. But when a patient responds, they respond more durably and you can even stop the treatment in many of those patients.

And so, I think when you're dealing with immunotherapies, looking at landmark PFS and ability to stop treatment, overall survival is a better endpoint. I don't think the FDA has caught on to that yet, but hopefully they'll understand that, because there's a lot of developments in immunotherapy, not just in kidney cancer, but across many diseases, and you need to think about approval endpoints differently when you're trying to give curative therapies.

Pedro Barata: Yeah, no, those are definitely a great points. I couldn't agree with you more and you are right. We've been using the classical examples of endpoints, even though we've changed the MOAs and immunotherapies behave differently, as you pointed out. With that note, I'm aware at least for one study with Tivo exploring this potential synergistic effect With IO. There's this trial called TiNivo that's basically doing second-line or third-line progression on a prior checkpoint of Tivo or Tivo plus/minus Nivo. So maybe if we talk a little bit about immunotherapy and as an expert, let me pick your brain on that.

What are your thoughts and insights about, do you think there's a role for an IO-based approach upon progression on a prior checkpoint inhibitor? Sometimes, always, never? We have definitely interesting data from the Sloan group. They did it with Lenv-Pembro in the clear cell. We had a little bit of data in non-clear cell as well. But that was provocative, in my opinion, and you have at least two phase III trials ongoing in the refractory space upon progression on prior checkpoint. So I guess my question is, when you bring a TKI and IO together, what are your insights about a potential synergistic effect upon progression of checkpoint? Is that real, not so? What are your thoughts?

Michael Atkins: Yeah, this is an area of basic science research that I was involved in the 200s where we looked at mechanisms of resistance to antigenic therapy, which what happens when you give a VEGF receptor TKI to a patient who has not seen one before and has a VHL mutation and a tumor driven by VEGF in the absence of hypoxia. You see acute hypoxia and an infarction of the blood vessels, and that's a really bad time to try to stimulate an immune response and the resistance to TKIs is upregulation of a lot of immunosuppressive factors, including myeloid-derived suppressor cells and Tregs coming into the tumor.

And so, I view the IO/TKI combinations as sub additive and I look at the PFS curves for those combinations, and they look like TKI curves shifted to the right. So, in my view, it remains to be seen if those regimens are curative. I certainly have never cured a patient on an IO/TKI combination. I've never been able to successfully stop treatment and not have the disease come back. I think that's true in the front-line. I can't imagine why that would be different or better in the second-line in someone who's already progressed on an anti-PD-1 containing regimen. So I predict all of those regimens in the second-line. Any trial there will be negative.

Pedro Barata: Well, that's amazing insights and it's quite timely, because I think we will have news, at least of one of those trials, coming soon. I think we're going to get news probably ASCO GU, I believe. We'll see what happens. So it's great insights for that.

Final question on TIVO-3. You highlight the safety profile, and maybe just for the audience out there, because I just want to make it clear because a lot of times we think half-life and the importance of a short half-life and then the safety profile, but there we have it. We have actually the longest health life. I think Tivo is around 5 days or so. And with that, to your point, the safety profile seems really, really benign and it's really well tolerated in clinic. You have high blood pressure, which is a on-target effect. Since I have you here, can you comment a little bit about the difference between half-life and then the safety profile of the different TKIs?

Michael Atkins: Sure. Well, I think Cabo has a half-life that's similar or longer than Tivo, but what's different about Tivo is it was given as three weeks on one week off, and so there's a chance for the normal tissues, which, like VEGF pretty much as well, to have a chance to recover without giving too much opportunity for the tumor to recover. And so, I think that makes it a more tolerable regimen that is easier to give and less modifications that are necessary in the dosing and less calls from the patients because of side effects. The side effects are primarily hypertension and hoarse voice or dysphonia. While other things are much less common, there's less of a hand/foot reactions, there's less of, certainly, a rash than we see with sorafenib, there's less GI toxicities.

But I think there is some patients wear axitinib, because of its really short half-life, becomes a better drug, because if you're worried about toxicities in a patient or you're worried about issues related to bleeding or that they might need surgery or something like that, then having a drug that you can stop right away and goes away offers a margin of safety.

Pedro Barata: For sure.

Michael Atkins: It's also a little bit easier to adjust the dosing of axitinib than it is with tivozanib. But if you're have a patient who's pretty stable, they're no longer a candidate for immunotherapy, they have relatively slow-growing disease, so having the most efficacious regimen is not necessary, where your focus is on keeping the disease under control without interfering with the quality of life of the patient, then tivozanib is a good choice.

Pedro Barata: Wonderful. Dr. Atkins, I feel like we could stay here all day chatting. This is wonderful. Thank you so much for being generous with us, taking the time, and again, congrats. Very important data out there. You really helped establish Tivo as another very valid TKI for patients with advanced RCC. Congrats on your presentation. I'm looking forward for more data out of TIVO-3 and I hope to see you soon.

Michael Atkins: Yes, likewise. Bye-bye.