The Uromigos Episode 111: The Randomized Phase 3 CANTATA in Metastatic Renal Cancer

Dr. Nizar Tannir discusses the phase 3 trial of cabozantinib +/- a glutaminase inhibitor in refractory metastatic renal cell carcinoma (RCC): the CANTATA trial.  Despite a sound mechanism and pre-clinical data, no clinical benefit to the combination was observed. Brian and Tom debate how much academic resources should be put into metabolic pathway inhibitors in RCC. View the ASCO abstract.

Episode Transcript

Brian:
Welcome everyone to another Your Amigo’s ASCO podcast. We’re joined today by our good friend and colleague Nizar Tannir from MD Anderson. Nizar, we’re going to talk to you about the CANTATA data that you presented in the oral presentation.

And if you could just introduce yourself briefly, and then why don’t you start by maybe giving us a little background about glutaminase inhibition and the pathway and why it might be relevant in renal cancer?

Nizar Tannir:
Sure. Thanks, Brian and Tom, for the opportunity. I’m Nizar Tannir, I’m a professor at MD Anderson Cancer Center in Houston, Texas in the Department of Genitourinary Medical Oncology. And for about 20 years now, I’ve been here at MD Anderson focused on kidney cancer trials as well as translational research.

So, CANTATA is a large, randomized global placebo controlled and double-blind trial that compared a new first in the clinic, glutaminase inhibitor, which was previously referred to as CB-839, now it’s Telaglenastat plus an approved second line agent for RCC cabozantinib versus placebo plus cabozantinib.

So, I think this is a new area in RCC research. And actually, I would say in tumor oncology, which is tumor metabolism. I think we all know for RCC, obviously, what’s relevant is we have immune checkpoint inhibitors that work, we have interleukin 2 pathway, and we have target therapies directed at VGF pathway and mTOR pathway. So …

Brian:
Nizar, is there something about glutaminase inhibition specifically in renal cancer, or?

Nizar Tannir:
Well, yes. Yes. In general, many tumors, as we all know, use glucose. This is an established tumor Metabolist for tumor cells and cells in general, which is the glucose utilization. And of course, we know that a normal cells, glucose is utilized and shifted through glycolysis to produce energy for several of demands.

That’s normal. Now, in tumor cells glucose, as we all know, there is an abnormal utilization of glucose by tumor cells because of the Warburg Effect where glucose is converted to lactate. And therefore, as a result, there is stress, there is increased demand on tumor cells to utilize other pathways for their energy and for growth and survival. And that’s really what a glutamine comes in.

And it’s been established that in many tumor types, including RCC, clear cell papillary chromophobe as well as other tumor types, for example, triple negative breast cancer, lung cancer, where there is increased expression of glutaminase, which is the enzyme that converts glutamine to glutamate, and glutamate is taken up in the TCA cycle for energy production.

So again, this we think is relevant and it’s been demonstrated in pre-clinical models. The one that’s been used extensively in many tumor models is the colorectal CT26.

Brian:
So, it sounds like blocking a critical enzyme in tumor cell metabolism, which could be applicable across tumors.

Nizar Tannir:
Yeah.

Brian:
How about, I know there was some preliminary data, before you talk about the phase three you just presented, there was some preliminary clinical data that had been presented before with this compound. Can you describe that briefly?

Nizar Tannir:
Yes. Obviously, we all start with monotherapy. Obviously, we have to really establish if this agent, by itself, is effective, and whether it’s safe or not. So, there was a large phase one trial to first establish safety, and this was established at a dose that blocks, pharmacodynamically, the pathway this glutaminase enzyme we’re talking about.

There was an upward of 96% inhibition of glutaminase in platelets used as surrogate for tissue blockade of this enzyme. And so, the phase one monotherapy was safe, and the dose of 800 milligrams twice a day was selected.

And the activity in heavily pre-treated patients with this agent, I’m talking about two cell RCC of course, and other variant histologist RCC, the activity was modest. Similar to what you see with other tumor metabolism agents, such as mTOR inhibitors, and I’m talking about everolimus here and [inaudible 00:05:25], where the response rate, as you know Brian, was in the 2% range, but a larger percentage of patients with stabilization, with stable disease.

So, this was established as monotherapy. That it’s safe and it can lead to prolonged stabilization of disease. Although the response rate by resist was low. So next then we, and again, in the phase one trial with the combinations, two combinations were tested, one with everolimus obviously again blocking glucose with everolimus and this agent Telaglenastat blocking glutaminase. Also, the other combination that was tested was cabozantinib plus Telaglenastat. [00:06:14].

And so again, there was some pre-clinical data that showed synergy when you combine this agent Telaglenastat, the glutaminase inhibitor, with everolimus as well as with cabozantinib. And again …

Brian:
And I know there was a randomized phase two with everolimus.

Nizar Tannir:
Yes.

Brian:
Do you want to talk real briefly about the clinical effects?

Nizar Tannir:
After the phase one showed the 800 milligrams twice a day dose and schedule of this agent was deliverable in combination with full dose everolimus 10 milligrams a day, and with full dose cabozantinib above 60 milligrams a day with, as I said, good, strong pre-clinical data showing synergy in vitro, as well as in vivo, using these two combinations.

Then the next step was to design and conduct clinical trials, randomized trials, with everolimus as well as with cabozantinib. And so, the first …

Tom:
Nizar, what did the everolimus randomized trial show?

Nizar Tannir:
Here is what the everolimus trial showed. That trial preceded Cantata trial, and we called it Imprata, and we actually presented the data at ESMO 2019. Jolie from Memorial presented on behalf of our investigators. And what it showed was doubling of the median PFS compared to everolimus plus placebo.

So Telaglenastat plus everolimus produced a very modest PFS median of 3.8 months, compared to 1.9 months median PFS for everolimus plus placebo. And it was safe. This small, randomized phase two trial was 69 patients, with two to one randomization, in favor of the experimental arm Telaglenastat was …

Brian:
Nizar, you say that those medium PFS were pretty modest. It was a doubling, but the absolute results were modest.

Nizar Tannir:
Yes.

Brian:
I guess my question is, was the current Cabo study that you’re about to describe, was it already in the works and planned before the phase two, or was the phase two felt to be positive enough that it led to the phase three?

Nizar Tannir:
No, actually the Cantata trial, the thinking about doing a trial in the salvage setting was in the works …

Brian:
Okay.

Nizar Tannir:
… as the Cantata trial was ongoing. But obviously a small biotech company [Califera 00:08:44], as you well know, Brian and Tom, they have constrained resources, they have to use their resources in an efficient way.

And then it was felt that combining with an mTOR inhibitor in second line, third line, very few investigators and treating physicians were using everolimus anymore at that time. So, it wasn’t really appealing to have everolimus plus placebo, and especially in US, as you know Brian.

Nizar Tannir:
So, we decided okay, there is strong pre-clinical data, as I said, that supports the use of combining Cabo plus Telaglenastat. And of course, Cabo is an established agent.

Brian:
Sure.

Nizar Tannir:
So, it made sense to really… And, very importantly, again, to go back to that phase one combination trial. In a small number of patients, Telaglenastat plus cabozantinib produced 50, 5-0% objective response rate, and 100% disease control rate. Obviously …

Tom:
Nizar, with hindsight was the randomized phase two with everolimus, because you did a classic randomized phase two, randomized phase three approach, but you switched the drugs between the randomized phase two and the randomized phase three.

The randomized phase two had some positive signals, but by switching the control arm to cabozantinib in the combination, it really meant you were somewhat in the dark. Is that a fair comment?

Nizar Tannir:
Yes. Obviously, hindsight, as you said Tom, again maybe we could have done that. But I think already the train, as they say, had left the station.

Tom:
I agree. I agree. I think the trial design was right, actually. I think it was right. If you said to me, “Was the randomized phase two with everolimus encouraging, and was it the right combination,” I think the answer to both of those questions, I don’t think it was encouraging, I don’t think it was the right combination.

But it’s easy with hindsight to do that. And I totally agree, I think the drug company, the sponsor and the investigators who led the trial, you did the right thing. It was the right study design. It didn’t turn out the way you wanted it to, but I don’t think there’s any problem with it.

The question is, I don’t think the randomized phase two with everolimus helped very much. I wonder, had a more pragmatic or a more ruthless group of people got together, whether they just said, “We should just drop it here and not do the randomized phase three.”

Nizar Tannir:
Well Tom, you did not participate in the Imprata trial with everolimus, but you did participate in the Cantata trial with us.

Tom:
Yeah, so Nizar don’t take my contribution to the trial as an indication of an endorsement. It doesn’t necessarily mean there’s going to be success, because I have an unfortunate record …

Brian:
As I described it, let’s move on to the phase three data you presented. Obviously, we talked about the lead up. Give us basic trial design and then just high-level results.

Nizar Tannir:
Sure, sure. Obviously, we wanted to make sure this would be a registration trial if positive. So, we powered it to have 416 patients, and that sample size was based on a power of 85% and setting the hazard ratio of 0.69 and one-to-one randomization between the experimental arm of Telaglenastat plus Cabo versus placebo plus Cabo.

This was straightforward a trial, with placebo controlled, double blind, as I mentioned. And it was global, so we had US sites, we had European sites, Australia. And Tom, you joined us as well. And so …

Tom:
Thank you for asking me.

Nizar Tannir:
Sure. Yes, I told them, make sure Tom is onboard. And so we, as most of these trials are in RCC, the primary endpoint was PFS adjudicated to by independent radiology review. And then secondary end points were at the usual, PFS by investigators, assessment and OS objective response rate, quality of life, et cetera.

The trial actually occurred very briskly. Again, it was embraced by the investigators in Europe and in the US, because obviously the same thing was Cabo was basically embraced as a modified second line, third line, if you will.

So, it was easy. And this is a new drug, a new class of agents, because obviously we wanted, we all knew that the, I-O-T-K-I, the I-O-I-O combinations definitely moved the field forward, helped a lot of patients. But still, there were many patients who had relapsed, many didn’t respond …

Brian:
Definite excitement about the drug and the pathway.

Nizar Tannir:
[crosstalk 00:13:44] … A new target, a new agent …

Brian:
Right.

Nizar Tannir:
… and a new class of agents. And that’s where I think the community of RCC embraced tumor metabolism and this agent.

Tom:
Nizar, let’s jump onto the results of the study. Were the patient populations balanced? Because it’s not a big, randomized phase three. And so, were there any imbalances in the patient population?

Nizar Tannir:
No, no. Actually, the two arms were very well balanced for the co-variables, the usual co-variables. And again, we stratified, which was important, by IMDC, obviously an established stratification factor. We had to do that.

And also stratified by prior immune checkpoint inhibitor exposures. So, I think 62% of the patients on the trial had prior immune checkpoint inhibitors, and 21% of those actually had nivo IPI.

Tom:
And Nizar, what’s with the primary end point, the PFS and LS results?

Nizar Tannir:
The primary endpoint was PFS, and there was no difference between the two arms. As I showed yesterday at the oral session of ASCO, the median PFS in the experimental arm of Telaglenastat plus Cabo was 9.2 months and was 9.3 months in the control arm with placebo plus cabozantinib. Objective response rate was 31 versus 28% respectively.

OS, as you well know Tom, from the data you were privy to before the presentation yesterday, the data on OS still lacks long-term follow up so it’s not mature at this point. But there were very few deaths, 34% that’s on the experimental arm, 30% on the control arm.

And safety-wise, there was really no signal of anything unusual that we didn’t see within broad phase one trial. And with what we know about monotherapy with Cabo and monotherapy with Telaglenastat. So …

Tom:
Nizar, can I ask a question?

Nizar Tannir:
… overall, a disappointing result, but nonetheless we learned. Because this was a large phase two or a small phase three, any way you look at it, but 444 patients ultimately were enrolled on the trial. So, we learned, and you can say that benchmarks were established for Cabo post IO, because again, about two thirds of the patients had prior IO before Cabo.

And again, I would not close the chapter on this. I wouldn’t say tumor metabolism is dead in RCC. I think in hindsight, maybe the Cabo partner was not the best partner. Maybe Cabo is more of, as we know Cabo, the OA of Cabo is VHFR inhibition and mat and axle inhibition, and maybe not as much glucose blockade as you would find with mTOR inhibitors, for example. So …

Tom:
Nizar, does that mean you would go back and do another trial with the drug, with an mTOR inhibitor, for example?

Nizar Tannir:
I don’t think the company, with its stretched resources, is going to do that. But I’ve been talking with them about using a different partner. And I mentioned that yesterday in the discussion, in the session, as you’ll recall. We’ve been negotiating to do single arm studies in tumors, which I believe, among all the RCC sub-types, are more metabolically active than clear cell RCC.

Clear cell RCC is an active tumor, but I do believe, for example, HLRCC, hereditary myoleiomyoma RCC, is the example, par excellence I would say, of a highly metabolic tumor. Renal medullar carcinoma is also another one.

So, we’re thinking of partnering Telaglenastat with bevacizumab plus [elotanib 00:17:55] in HLRCC, and in RMC, and in [inaudible 00:18:00] tumors. To so see if really, we can validate that concept of dual blockade of glucose and glutamine with a better partner in a more metabolically active tumor.

Brian:
Nizar, I have a question about this in the clear cell data. Do you think the trial was negative because the pathway isn’t as relevant as you thought, because the drug doesn’t inhibit the pathway to a complete degree for whatever reason, or there are compensatory mechanisms, or there’s something about Cabo as a partner that somehow negated its anti-tumor effect? Which one of those are more, I guess, do you think, is at play here?

Nizar Tannir:
Yeah. Excellent, excellent question, Brian.

Brian:
See that, Tom?

Nizar Tannir:
And this is what we’ve been grappling with here with Califera leadership to see what went wrong, or why was the trial negative? I believe the dose and schedule of Telaglenastat in combination with Cabo was not a problem. So, I don’t think it’s a schedule or a dose problem.

Brian:
Yeah.

Nizar Tannir:
I think Cabo probably was not the best partner, because Cabo is not probably, and maybe this is the reason why the trial was negative, did not block glucose utilization as other agents would have done. So, that’s one. Two, I believe maybe clear cell RCC, in an unselected way, is so heterogeneous that you really, this is not the best tumor type to really validate this concept.

And this is why I think maybe going after biomarker selected, like for example they’re doing now, we’ll wait and see the results of the lung study, where they have decided to go after Keep one and RF2 mutated non-small cell lung cancer, [inaudible 00:19:58] in combination with Pembro and chemotherapy.

Brian:
[crosstalk 00:20:04] renal patients? Is there a biomarker selection strategy in renal patients?

Nizar Tannir:
No, we do not have, at this moment, a biomarker selection for renal cell, clear cell, or others. But I really, as I said, I do believe that a HLRCC, renal medullar carcinoma, SDH, I think these are probably more… highly symbolic tumors, more than in unselected. We know the heterogeneity of clear cell RCC, we almost have it clinically.

Tom:
Nizar, there were some talk in the meeting about the immune pre-treated cohort. I have a question for you in a second, Brian. But the immune treated cohort who had previously had immune therapy, there was some suggestion that those curves were going apart and that’s clearly exploratory analysis. With hindsight, do you look back on that and just say, “That’s just [inaudible 00:21:05],” or is that a population you’re genuinely interested in?

Nizar Tannir:
I really, think this is an important observation that is probably going to inform us moving forward if the company will have resources to explore this. I don’t think it’s noise, Tom. I think in the phase one trial, again, going back to the phase one, there were anecdotes, you would call these anecdotes, but there were three patients with melanoma who had progressed on PD-L1 antibodies. And when they were treated with Telaglenastat plus nivolumab, one had CR, two had PR. So, I think …

Tom:
And that would be single arm data, so you’d go back, and you’d do single arm trials. You wouldn’t do a randomized [crosstalk 00:21:53] …

Nizar Tannir:
Yes, I think we should do a single arm and see what happens.

Tom:
Yeah. So back to …

Nizar Tannir:
But there is pre-clinical data to support combining a PD1 antibody and Telaglenastat. And the data is, again, from a colorectal cancer model, where when you block glutaminase in tumor cells, there’s a shift of glutamine utilization from tumor cells that are blocked now, from using glutamine, to the immune cells, to the lymphocytes.

So, I think there is science behind it, and I think Brian, Jeff Rathmell and Kim Rathmell, at [inaudible 00:22:39], have done a lot of work with Telaglenastat, glutamine and the immune microenvironment, where they show that there is something like this happening where glutamine is preferentially taken up by lymphocytes. And then you can think of invigorating, enhancing the immune system when you block glutaminase in tumor cells, and then you enhance the immune system.

Tom:
And I …

Nizar Tannir:
So, I think there is a scientific rationale to combine an immune checkpoint inhibitor plus Telaglenastat. And why, obviously the proof is in the pudding. We just have to see, if when we do the single arm study with, say, nivolumab or pembrolizumab plus Telaglenastat, we’ll have to see if we see enhanced activity.

Tom:
My take on this issue is it’s probably a bit like Johnny Depp turning up to your birthday party. In that it sounds like a really good thing, but actually when you look at the detail, there are probably better guests.

And what I mean by that, and what I think the provocative question is, this is a randomized phase three, the phase two didn’t work, the phase three didn’t work, and is this a target which we need to be investing in the future? Or are there better targets for us to be going onto? Are there better guests, is Johnny Depp the right guest?

Brian:
Tom, you’re going to get us into trouble from the lawsuit from Johnny Depp’s agent on this?

Nizar Tannir:
I think …

Tom:
Oh no, he’s welcome to our party.

Brian:
I don’t know if you’re asking me this question, but I think tumor metabolism …

Tom:
I am asking you the question, Brian.

Brian:
… tumor metabolism is, I think, the next big area of therapeutics in kidney cancer. Clearly this drug, in this setting, with this partner wasn’t right. But I think there’s a whole lot more to explore. I think this is the first foray into this area, and there’s going to be success here. It just wasn’t at this first attempt.

Tom:
Okay. So, Brian, question direct for you. You’ve got 100% of your academic resource and drug development. In the view of what we’ve seen, what percent of that do you think, of our academic resource, you think we should be putting into metabolism as a target in clear cell renal cancer?

Brian:
50%.

Tom:
5-0?

Brian:
Yeah.

Tom:
Wow. Of all of these things in front of us, you think it’s got 50% of the pie chart?

Brian:
Yeah. Again, I think it’s the next big area. I think it’s totally unexplored. There’s the Rathmell data that Nizar talked about, and interactions with the immune system. So, I think it’s just, it’s untapped. There’s only so much further we can go with our doublet therapy, et cetera. Obviously, we’re exploring triplets, that’s probably a big bulk of the other 50 …

Tom:
And so, despite what you’ve seen, you still think it’s the best untapped resource?

Brian:
I do.

Tom:
Okay. And Nizar, do you agree with that?

Brian:
But remember, it’s a big category. It’s not a specific drug, right?

Tom:
And Nizar, do you agree with that?

Nizar Tannir:
Yes, I do agree. I don’t know about the percentage, half my resources going to one class of agents. But again, you look at the landscape, Tom.

Tom:
Yeah.

Nizar Tannir:
We have gone as far as we can go with a doublet, even with a triplet of I-O-I-O-N-T-K-I. And of course, the HIV2 alpha inhibitors are there, but again, what we get is some partial responses and stability of disease.

So, I think tumor metabolism is not just going to be relevant in RCC, but in many tumor types, as I said. And I think it could find a way in triple negative breast cancer, where there is an unmet need, and lung cancer and other tumor types.

And I think for RCC, before I invest a lot of resources in clear cell, I would go, as I said, with tumors that are less heterogeneous than clear cell. Because we know the heterogeneity of clear cell.

Tom:
Brian …

Nizar Tannir:
[crosstalk 00:26:22] RCC, the renal medullar carcinoma, that will be our proof of principle. If we combine Telaglenastat with say, [inaudible 00:26:31], clearly a regimen that’s established as tumor metabolism for HLRCC.

Tom:
Last question …

Nizar Tannir:
If then we also confirm HLRCC in a recent publication, if we then see that there is a synergy between Telaglenastat and [inaudible 00:26:49] in these two tumor types, then I think we can then expand and try it in clear cell with a different partner than cabozantinib.

Tom:
Last [inaudible 00:27:03] Brian, to start with you. What did we learn from this trial in terms of cabozantinib, in terms of treatment refractory disease? Did it take us forward at all in terms of the way we treat patients? Because it is the most recent randomized phase three in treatment refractory disease.

Brian:
Yeah. I don’t know. I think, as Nizar said I think, in his presentation, it establishes a benchmark. I think IO refractory is very different than VEG-F refractory. I think we’re seeing that. We’re going to see longer PFS, and higher response rates, especially with the lingering IO effect in some patients.

So, I think it is good to have new benchmarks. I think that’s probably the biggest contribution. In a positive sense, obviously again, it wasn’t the right partner, the right setting, but I think there’s a lot more work to be done.

Tom:
Nizar, this has been fantastic.

Nizar Tannir:
Thank you, Tom. Thank you, Brian, for the opportunity.

Tom:
We’ve been lucky to have you.

Nizar Tannir:
I really like this, and I like your British humor, Tom. More of this on the next podcast.

Tom:
I wasn’t trying to be funny, that’s the point. Nizar, it’s great to chat, as always. I’m looking forward to seeing you soon.

Nizar Tannir:
Yes, likewise.

Brian:
Thanks Nizar.

Tom:
Congratulations on the trial. Listen, these big negative trials, they do help us. I actually think it is a big step, because it has put, for me, the brakes on some of this metabolism as a, and it would be less than 50% of my resource.

And prior to this, it probably would have been 50%. So, I think it is contributing. I think negative trials have a really important role to play. And so, congratulations for your hard work.

Nizar Tannir:
Thank you, Tom. Thank you, Brian for the opportunity. You guys stay safe.