Good afternoon, everyone. Day three of our Jefferies London Healthcare Conference. I really do appreciate everyone's attendance, and I hope we are all having a good set of meetings. My name is Akash Tewari. I am head of our pharma and biotech research efforts internally in the firm, and I have the pleasure of hosting Andrew Peters, Chief Strategy Officer of Exelixis. I will make note, it is always a tough tape in healthcare, whether it is good or bad. It is very rare to see the amount of outperformance that Exelixis has had recently, and it is worth acknowledging. I will hand it over to you for some intro remarks, and we will get started.
Yeah. Thanks for the invite for the conference. Always happy to be here and always happy to chat. You keep me on my toes, so that is always good. Just as a reminder, going to be making some forward-looking statements today, so please see relevant disclosures in our regulatory filings around risks to our business. As you mentioned, yeah, kind of a really exciting time for Exelixis right now. We are kind of at that cusp of transition between single product, single franchise company to multiple product, multiple franchise company, going from CABO to CABO and now Zanza, and then just really kind of keeping our head down from an execution perspective, where cranking on the commercial side, cranking on the Zanza development side, early-stage pipeline keeps progressing, and then being deliberative and precise with our expense management.
So we can have a high degree of optionality, use our balance sheet if we see something externally that we like as well. Kind of going across all of those cylinders for our business, our theme, kind of the through line of the company is we run Exelixis like a business and not a biotech, and really the heart of that is execution.
Okay. I think that is a very, very provocative statement and worth examining, because we go back to a few years ago, and we always love our legal IP cases, and that is really what I think Exelixis was, right? It was range-bound for several years. There was a question about the patent life on CABO. You and I still remember this, the kind of investor narrative was, "Okay, if they succeed, the stock will go to $25, $30, but then what?" I think it is worth giving your team credit that you took feedback, I think probably from investors, you also took feedback in terms of what your prior development strategy was with CABO and applied that on a go-forward metric. I think that is part of the reason why it was not just $25-$30. You are sitting here in a very different market cap range now.
I want you to reflect on that a bit. When you mean, "We're running it like a business instead of a biotech," what has been, Chris and Mike, what have they implemented from a capital allocation standpoint that has differentiated you from your other commercial stage peers?
Yeah. So there are a couple of things there. I mean, one of the things that we do best is prioritize. We've said from a capital allocation perspective, big picture, we think about the world in three buckets: internal R&D, external R&D, which is BD, M&A, et cetera, and then share buybacks. We don't think any of those three are mutually exclusive. We think the financial tailwinds that CABO has, the upcoming Zanza launch, provide us the optionality to do all three. So they're not really mutually exclusive. On the R&D side, one of the things that we do best is prioritize.
What I mean by that is we've been pretty clear about capping our budget at around $1 billion a year annually. As we grow the top line, we keep the expense line flat, that's when you're really seeing that earnings acceleration that's going to really allow us to invest everywhere else. That prioritization is through two lenses. One, internally fighting for resource. Does product or program X or Y deserve to be funded and why?
Here's the data, here's the biology, here's the vision, here's the commercial opportunity, et cetera. But then the second one is really franchises, and Mike's talked a lot about it, we've talked a lot about it. That's how we see the world. That's how we see the world of value creation. If we have franchises like CABO and Zanza, where they're approved across multiple indications, great. That second degree, though, is if we can have franchises across RCC with multiple products, multiple labels, NET, multiple products, multiple labels, CRC, multiple products, multiple labels. We build this inherent profile of market leaders across multiple different directions. That's really the lens, how we view the world, certainly informed by CABO. Everything we do at Exelixis is through that CABO lens.
The way we grew CABO is adding indications, expanding, being the leaders in RCC, growing into this leadership in NET. That's the dynamic that we want to continue.
Understood. I mean, it's not like there's been lessons, harder lessons learned along the way. I still remember 2020, your team called its shot on CABO being a $4 billion drug. It wasn't guidance, it was aspirational, illustrative, whatever we like to say. But you didn't achieve that, right? I think what was interesting, and I want you to apply that to the Zanza development. When you thought about expanding outside of RCC with CABO and the sets of studies that your team ran there versus what you're doing with Zanza, you have seven different trials that are ongoing. How is the CABO development path from that 2020 timeframe different than the Zanza development path in 2026?
Yeah, I think there's two things there. One, just functionally speaking, we're at a much different place financially, and we're at a much different place from a LOE runway perspective.
The combination of those two things certainly influenced how we thought about CABO development at the time. But there's other parts, like take bladder cancer, something we're really excited with Zanza about. We have a lot of CABO data in bladder, but at the time, PADCEV was emerging as this kind of new, incredibly active modality. As a company, we realized, did it really make sense to go into, say, a CONTACT-04 or whatever we would have named it, if the standard of care would have shifted pretty substantially during that time? That's an opportunity for us, and this is a bigger thing. We have a real advantage of not only using prior Zanza data, but prior CABO data to point us in the direction of where we can go.
Mechanistically, we think the secret sauce or kind of that special blend of all of the different kinases that CABO hits allows a relatively broad activity profile across a lot of tumor types. Then the shorter half-life kind of user-friendliness of Zanza allows it to be a backbone therapy or in a lot of different combinations. So, again, coming back to bladder, that's an example of we can use the data from CABO, the market dynamic change over time to inform a development plan that hopefully we're going to talk about soon with kind of wave two. So, we have this ability to leverage all of this wealth of history and data and experience with CABO, as well as our early experience with Zanza to then form where we're going in the future.
Understood. Now, as we've heard investor debates about differentiation on CABO and Zanza, and I think your point is fair. It's not getting directly tested in any of your trials. So I think from an investor perspective, that's been true. But I also don't think Zanza was designed to simply be a me-too CABO. You looked at feedback you got from physicians about safety issues you see with VEGF TKIs. You're quite aware of down-dosing and really sometimes these AEs start to manifest not initially, but really over time.
So, let's remove the Zanza versus CABO kind of investor debate, but I would actually ask you to just talk about it scientifically. Why is Zanza going to be more user-friendly when you think about it being kind of the VEGF TKI of choice, when you think about in CABO, sorry, CABO indications going forward? Then you have had data so far. What has really stood out to you that kind of gives your team internal confidence that this is a more user-friendly compound?
Yeah, and it really is. I was pointing out to someone earlier. So CABO's Exelixis numbered name was XL184, Zanza's XL092, so half of CABO, kind of making a little internal joke about the half-life difference. But it really is as simple as that. Our hypothesis was we think that how we hit not only VEGF, but MET, AXL, MER, the TAM kinases, really incorporates a lot of that vascular kinase biology, myeloid cell biology, kind of that profile that pairs well in particular with checkpoint inhibitors, for example. CABO's a great drug, full stop. But one of the challenges that ultimately comes up with CABO is it has a relatively long half-life. So from a patient management perspective, virtually every oncology drug protocol is something along the lines of adverse event comes up, dose hold until resolution of symptoms, re-challenge at a lower dose.
Broadly speaking, that is kind of how all oncology drugs work. When the world moves towards multi-drug, multi-modality combinations with potentially overlapping AEs, you are just adding in additional layers of complexity if that dose hold, that washout period becomes 10 days, 12 days, two weeks. It is just inherently challenging because you are not really sure, is it the TKI? Is it the IO? Is it this? Is it this? Some symptoms resolve fairly quickly, hypertension, go on an oral antihypertensives. Some are more challenging clinically. There is a lot of just dynamics and aspects to consider. Our insight was, can we take that core CABO scaffold that phenocopies the kinase inhibition profile, engineer in a metabolic liability that cuts its half-life, that functionally speaking makes it more user-friendly from a dose hold, dose reduction, resolution of symptoms period.
Perhaps luck, perhaps strategically, one of the things we are seeing is differentiated tissue distribution.
Surprise, surprise, if you change the half-life, you are probably going to change a little bit of the tissue distribution. That is one of the things we published on, in particular differences, say, between tumor concentration and things like the skin or the periphery. That manifests, say, in differences in we are seeing in hand, foot, or PPE. That is one, as opposed to hypertension, which is on target with an anti-VEGF, but really easily resolvable with oral antihypertensives. PPE is burdensome from a patient perspective, really hard to manage clinically, and can take a little bit of time to resolve.
If we are seeing improved PPE, for example, with that more user-friendly profile, that allows us to take a pretty broad aperture and see where all the different tumor types that a drug like CABO or Zanza could be active, where could it be combined, where is the biological rationale for the combination, all of that stuff. That is how we thought about wave one, and certainly that is how we are thinking about wave two.
I can't help but hearing Merck talk about what they've learned from LENVIMA and hearing PD-1 VEGF companies talk about, there's areas where there's not been that PFS, that response rate PFS and OS translation. It seems very akin to what you're saying right now with Zanza.
Yeah.
The big debate on the PD-1 VEGF class is, are we going to see a strong correlation between PFS improvement and then kind of overall survival? I think a key point in that is really duration on treatment. So it may not be the absolute number of AEs, it may be effective dose intensity.
Do you feel like there's signs of that with Zanza, with some of your early data? Do you think we could see some early proofs of that type of differentiation with the studies that are reading out over the next couple years?
Yeah. I mean, I think everyone in this room, in this conference, would love to have an easy answer to that. I think the reality is we're only going to know until we run large randomized survival trials.
Because my challenge is we look externally to opportunities to in-license or acquire companies, and your challenge and investors' challenges is to try and how do you extrapolate limited phase II data to the likelihood of a phase III success? Better data, better early data should translate. The honest answer is that's why we run these large studies. We're confident in the biologic rationale, the trial design, all of that, but that's why we're in seven pivotal studies right now, and we want to wait and see how that plays out.
Understood. I'll just push you a little on that. You haven't talked about it openly, but is it fair to say, just because that's not maybe where the investor debate should be directed, is there maybe for your scientific team a view there? What would they say if I asked them that question?
Well, I think they'd say that it's indication-specific as well. I mean, you were referring to the bispecific debate right now in PFS survival. I think there's certainly an aspect of that in lung cancer. How does PFS response rate translate to survival in other indications? It's probably a case-by-case basis, candidly.
Right.
We get asked versions of this all the time around trial design and things like that, and the reality is, we empirically design each Zanza study on its own. We sit and think about dose, combination, patient population, because unfortunately in this business, there are very few instances where there's a one-size-fits-all approach, especially through this franchise lens, and this is how we evolved with CABO, where CABO 60 is a monotherapy in later line patients. We moved to CABO 40 in frontline in combination with nivo, in part because we realize the clinical benefit risk of not only a lot of that efficacy data, but quality of life, tolerability.
Those are the things that actually drive utilization in commercial practice. When we talk about generating a totality of data, that's certainly one of the dynamics that we're always considering, because our experience in RCC in particular, CaboNivo is the number one IO-TKI. When our data and when others' data came out, there was a real debate as to what would be successful, and I think our experience is you really have to think about it holistically.
Right.
Efficacy, safety, secondary endpoints, all of that stuff actually drives adoption in the market.
Understood. Going back to, again, business, not just a biotech, STELLAR-304, right? You have an important study coming up. It's one of the kind of big initial shots on goal with Zanza. I feel like it's crazy. It feels like all of my companies have trials that go on longer than expected. You're in that bucket too. I can't help but think, with your current corporate strategy, something going into phase III, you would have to have a high degree of conviction about where the rationale fits, a good understanding of what the comparator on performance is, and really where this regimen would fit in. We have that framework of your team, and then we also have a study that's going on, I think, longer than maybe your initial expectations and street expectations.
Has there been a shift in your confidence that the trial design was right, the kind of market fit was right with STELLAR-304 in non-clear cell, or do you remain confident irregardless of the delay?
Yeah, I mean, it is really about speculating on blinded data. The reality of non-clear cell is that there is just imperfect data across the board.
There has never been a randomized study in this setting. Your question is really, and I have heard this from KOLs at ad boards, which phase II study do you trust more?
The 24 patients, 210 or whatever it is, from eight years ago, or the single-center CaboNivo experience, or the Len/Pem study that probably enrolled more favorable risk patients that maybe skews medians.
Right.
There is just a lot of dynamics, and it is hard to say, "Okay, what do you use? What do you trust? What is the data directionally show you?" Then factually speaking, the events are coming in slower than we modeled.
That's all we can really speculate on. What we've said and what Dana talked about on the last earnings calls, based on what we're seeing right now, we expect clinical cutoff around year end, lock the database, clean the database, top line it sometime in the first quarter.
Understood. When that top line does come out, I think the two investor questions you're getting, especially because the trial's taken longer, is, A, what will be that initial top line disclosure, whether it's not Because remember, it's ORR, and then that's actually, you can fail on ORR alone, then you have PFS. Then the second part is sort of like, how much can you disclose there? Then I think the second part, which becomes trickier, is really overall survival, right? That, I think we can look at historical data. I don't know if that's necessarily a fair expectation, but I'd love to get your take. A, what's your team thinking in terms of initial disclosure on ORR and PFS?
Then number two, expectation that you could hit on OS at this time point, and then more broadly, long-term expectation of OS given, actually, if you look at other precedents in RCC, they don't always hit on OS, they still get on the market.
Yeah, so a lot there. The easiest answer, I don't know what's going to be in press release. We haven't written it yet. There's always this delicate dance between stakeholder disclosures and, let's say, medical conferences preferences.
Yeah.
Around what's new, what's unreleased, what's released, and there's always a little bit of a back and forth there. Whatever I say, we'll probably change our mind 5x before.
Right.
We ultimately decide on the press release. Stay tuned. I don't know. As it relates to kind of survival, per protocol, when we hit the number of PFS events that triggers the primary analysis, there's going to be an interim done on survival. How mature it'll be, I'm not really sure. I think ultimately it's a statement of the obvious that the more robust the dataset we can generate, the better it'll be commercially.
That's kind of full stop. Our hope is that we can, meaningful response rate, meaningful PFS, meaningful OS. That's kind of our hope. We'll see ultimately what the data show. The one thing I'll say, kind of taking a step back on the Zanza program in general, non-clear cell is the smallest of any of the indications that we have ongoing. The way I think about Zanza is it's kind of this stepwise build of a foundation. The question we always used to get was, "How are you going to replace CABO?
Our answer is, here's seven pivotal studies that collectively not only have the ability to replace CABO, but grow, and then we're layering on this kind of next wave, which substantially increases the TAM of those potential markets. Then you layer in what we can do either with our early-stage program or the optionality that looking externally does, and that's kind of our big picture vision over the next 10 years. How do we grow and evolve Exelixis from a one franchise to a two franchise to a three-plus franchise company? That's kind of the vision.
You've alluded to wave two a couple times, and I think it's notable. Your company has a significant focus on RCC, and I don't think you're backing away from that. But there is, at this point, no CheckMate 9ER replacement that you've necessarily announced yet, right?
Stay tuned.
Right. Stay tuned. Saw some interesting data at ESMO on the T-cell engager side. I do think, like this, in my opinion, the HIF-2α plus PD-1 is not going to be a TKI plus a PD-1. I think we're starting to see that. You already have a partnership with Merck, which has helped you kind of share the development costs with Zanza. You talked about, "Hey, we want to kind of stay at around a billion R&D expense." You've got a tricky challenge, right? Which is, it looks like with the next wave, you're going to go external with maybe not just Merck, but maybe additional partners to think about the right regimen to take forward in a frontline setting. But then you also have to balance that with making sure the deal makes sense from a development cost perspective, right? Two parts there.
A, how do you make sure that whatever trials get announced make sense from a P&L perspective and your own internal discipline? But then B, can you comment maybe more generally in terms of modalities and emerging datasets you're seeing in RCC that get your team excited?
Yeah. I think kind of the I'd add one dynamic that is related. The failure of LITESPARK-012 really opens the door.
Because I think the frontline metastatic investment would have been a lot more complicated had LITESPARK-012 been successful. Certainly would have informed us on a biology perspective around that sort of triplet opportunity. But the failure of that LITESPARK-012 really, and we talked about this in the last quarter, like it has enabled us to think big now.
We've talked about what does that potentially look like. I'd think more orthogonal mechanisms that, say, doubling down on VEGF, where HIF and VEGF TKIs kind of hit that axis really hard. So that orthogonal mechanism dynamic is something to pay attention to. And would certainly agree collaborations is a model we like. It's, at the end of the day, coming back to a word I mentioned before, prioritization. How do we prioritize frontline de novo metastatic RCC from everything else that we're planning? So, that's kind of the fun part of the job is that debate, that discussion, going through all that data internally.
How close are you to wave two?
Stay tuned.
All right. On that note, hey, thank you so much. I really do appreciate it.