Exelixis, Inc. (EXEL) Earnings Call Transcript & Summary

Good morning, everyone. Bright and early in Times Square. The Elmos have not come out to Times Square yet. They will be there in a couple of hours, do not engage with them. That's my only advice. My name is Akash Tewari. I am a pharma and biotech analyst here at Jefferies. And it's my pleasure to host Exelixis. And Andrew, thanks so much for joining us. We are both fresh off our trip from Chicago at ASCO. It's been a really great -- probably 1 of the best stretches for Exelixis as a stock. Andrew, why don't I hand it over for you for some brief opening remarks, and then we'll get started with more Q&A.

Yes. Thanks. And I'd echo the same, kind of bright and early this morning. But candidly, just excited to be here, excited for a great day of meetings and for the conversation. As a reminder, we'll be making some forward-looking statements today. So please see relevant risks and disclosures in our regulatory filings. So yes, as you mentioned, coming off the heels of ASCO, which is always an exciting time for an oncology company, certainly exciting time for Exelixis, presented what we think some pretty interesting data for our lead pipeline programs, zanzalintinib. But I think taking a step back, Exelixis has kind of been on this cusp of a breakout for quite some time. As you can appreciate, prior to October 2024. We were a little bit of a holding pattern externally as a company as we kind of waited for some of the ongoing patent issues to resolve. But really kind of we came out of that October time frame with a favorable outcome in that. We signed a big clinical collaboration with our partners and, frankly, biggest competitors, Merck, and really just found ourselves at the point of about to break out. The underlying business of cabo is certainly doing well, kind of humming along, and we're executing on all cylinders there. We have 6 pivotal studies up and running with -- are up and running or soon to be running with zanza, and 2 of them actually expected to read out later this year and then kind of the early pipeline progressing and just kind of everything from a capital allocation, from an investment, in Exelixis execution perspective, really a lot of momentum behind it. So it's an exciting time to be at Exelixis, it's an exciting time to be in my seat where I have an opportunity to both look internally on kind of how we're focused whether it's cabo, zanza or the pipeline, but importantly, externally as well. So I get to go to a conference like ASCO and just understand where is the future of oncology going? How do we play there? What are the sorts of things that we can do as a company to be opportunistic with our balance sheet, to be opportunistic as a big, small company preferred partner for a lot of smaller biotechs and really just understand how do we grow Exelixis. One of the things we talk about is drug -- our business at the end of the day is simple. The more patients you treat, the more patients you help, the more value you can create. So we think about it and drug 1, cabo is doing really well. Drug 2, zanza is really poised for the key readout. Drug 3, 4, 5, whether it's in our internal pipeline or external, that's what we also need to focus kind of going forward. So looking forward to the discussion today.

Awesome. As am I. Why don't we start off with -- and I think this is probably where I get the most questions and what we're working on. So CRC, obviously, you're going as regorafenib, not particularly the best -- it's a good comparator to go against. It's a very tough disease. Obviously, on the Q1 call, the dual endpoint changes, I think a question people have. And the way I kind of see it is a company generally does not split their alpha if they're worried about the powering for their trial. But then you also have kind of 2 distinct populations, liver mets and non-liver mets that might have 2 different event rates. So it's a bit complicated, I think, for investors to understand. I'd hand it off to you, Andrew. How should we think about that dual endpoint change? Is that a sign of confidence? Is that a sign of speed? Is that a sign of -- well, we want to put this drug in the best position to be static and we might be concerned about powering in certain populations?

Yes. So Amy talked a bit about this on the earnings call. I mean, taking a step back, 1 of the things we always talk about is, ultimately, we and everyone in our industry is in the business of P values. We want to run successful studies. We don't want to run studies. And there's an important distinction there that I think oftentimes, a lot of companies, a lot of investors may miss. And so we certainly have that kind of dynamic of we're in the business of P values as a through line every study we design, everything we do. And so when we were looking at the 303 study, we've actually amended that trial several times now. If you recall, kind of that first amendment was when we got a pretty significant kind of update for the industry about how IO-TKI behaves in that later-line colorectal population. And so we made certain changes to that trial to essentially, in our mind, maximize the probability of success, understand what are the really key drivers there? Is it RAS status, as had been kind of our hypothesis before. Or was it more related to this emerging kind of view in the colorectal clinical community that patients who have liver metastasis in patients who don't are functionally 2 different patient groups. And it actually makes a lot of sense. If you think about patients who unfortunately have tumors that have gone into their liver, you would expect compromised liver function and given the role physiologically of the liver, they tend to die faster, expire faster. And so you end up seeing kind of this difference in accrual of event rates. And so as we were seeing these events coming in on a blinded basis, 1 of the things that we realized is given the size of the study, given the number of patients that we had enrolled we had the ability to. On a temporal basis take an earlier look at a much bigger population. And so any time that we're able to do something which we think, one, increases the probability of success for the overall study and be able to look at a much larger population, we think that's a good thing, right? And so to me, the best way to think about it is more of a temporal shift than anything that allows us to also look at that broader ITT group as opposed to kind of the -- just the non-liver mets, for example. So 1 of the things we have said is from a market opportunity perspective, CRC is probably about $1 billion, but it's actually roughly split 50-50 between the NLM and the patients with liver metastases. That's in part because the NLM patients tended a little bit longer. They tend to be on drug longer. So kind of the market opportunity is about the same. But from our end, if we're able to kind of left shift, get to market earlier in a bigger population, that's a dynamic that we tend to pay for.

So bottom line, sign of confidence and speed rather than necessarily worry about how events are accruing. Would you agree with that or dispute that?

Yes. I mean, ultimately, the cosmic truth always comes out when that envelope is read, so to speak. But I think it's a change that we're excited about and, again, kind of tend to see it as more -- how do we do what's best for patients.

Got it. I like the cosmic fun, given [ answers exactly. ]. Okay. Got it. Now maybe just stepping back, and this is a conversation I have with investors, I'm like, I feel relatively confident on CRC, given -- you probably know what the event rate on regorafenib is -- so as events come in, you can kind of figure out what the delta is. Not a clear sell, again, I feel pretty confident there. To me, where our team is doing the most work is really head and neck and then RCC. And it's interesting. Everyone is like, well, is cabo differentiated versus zanza, and I look at your clinical development plan. where are you directly testing that is I don't see those studies. What I do see is what you kind of mentioned earlier, which is zanza with novel targets and synergy where we maybe have not explored with VEGF TKIs, and maybe potentially a better therapeutic window. That's the case study in RCC. So why did you choose Merck as a partner, your biggest competitor? What did you see with WELIREG, which intrigues you in terms of its potential in RCC? And can you give us at least some outline of what that study design looks like, given that could make zanza not just a plus 1 in indications where cabo didn't in work in, but really a full switch?

Yes. I think the best way to think about it is to maybe take a little bit of a step back because that cabo/zanza dynamic is probably 1 of the most frequent questions we get as a company as well. And I think you framed it appropriately is that we've been very thoughtful. We've been very strategic and specific about how we've thought about zanza development. So given the original hypothesis around zanza or XL092, is that the main liability, so to speak, of cabo is it's relatively long half-life. And so what we sought to do with zanza is really to phenocopy the kinase inhibition, kind of the efficacy profile of cabo with a more combinable potentially safer version that we could say, okay, here's all of the 10-plus years of data that we've accumulated with cabo across these various indications, look across that landscape and say, okay, where are the areas that we can be pointed in the direction of where a drug like this would be active, but would be more favorable with a kind of more user-friendly, more combinable drug. And so you mentioned the 6 studies that we have up and running, 303, 304, 305, frankly, all of them are certainly informed by our cabo experience. So we actually see that as kind of a big point of value creation in that we've essentially enabled us to accelerate development because we can learn a lot from cabo to point us in the direction of zanza. So it specifically relates to kind of the HIF-2 alpha in the RCC space, we've agreed with our partners, Merck. Obviously, it's a competitive area not to get into the specifics of the design of the study until those are up and running later this year. But what I can say is, there's been a lot of intriguing early data of cabo and velzutafan that we can use as kind of a starting point to understand how those sorts of drugs behave in patients. And importantly, the signal is clear, the combination is very active and the combination is tolerable. And so if we think about zanza as a potentially best-in-class TKI, and we see the success that WELIREG has had as a monotherapy and in some of those early combinations, we and our partners have been intrigued about the potential for the combination of those 2 mechanisms being able to benefit patients. And so again, the way we think about it, cabo will never have a label in combination with a HIF-2 -- with velzutafan. And so we want to develop zanza in a way that maximizes its opportunity, but also has the potential to candidly establish a new standard of care. Because at the end of the day, drugs are successful because of that. Cabo is successful not because it's approved, it's successful because we were able to generate a data set with 9ER that establishes the standard of care in frontline RCC.

Yes. What's most interesting there because I think when those studies got announced, my head was like, okay, you had those ipi/nivo cabo studies. Nice response rate, durability not there because the side effects weren't working. And I would thought, okay, you can plug in HIF-2 alpha, but it doesn't necessarily seem like you guys think about this as a triplet, you think about this almost more like a doublet. Is that fair to say?

Yes. Again, without kind of getting into the specifics of the trial design ahead of kind of the launch of those studies, I think the more important dynamic is to think about zanza in combination with the HIF-2 alpha and kind of stay tuned on all of the specific details there.

Cool. Looking forward to it. Now maybe just on zanza and differentiation. The way we kind of think about it is, it's not like, okay, if you have a table and at any point, did a patient have Grade 3 proteinuria or do they have hypertension, it's -- zanza is a shorter half-life drug, right? So I don't think you're going to have a patient not have a Grade 3 adverse event. The point is once that Grade 3 adverse event occurs, how quickly can you down dose a patient and keep them on a very high dose of VEGF inhibition while still being adherent on the medication? So to me, this might be not a safety -- in terms of just absolute safety differentiation, it's more about discontinuations and then a better translation from PFS to OS. Is that the right way to think about this? Like how -- if zanza will differentiate in some of these indications, is that being all you're going after?

Yes. I mean I think 1 of the dynamics there you're kind of spot on. So imagine a patient who's on cabo/nivo for frontline RCC and they develop an adverse event and kind of the treatment algorithm, the AE management algorithm basically suggests dose hold to resolution of symptoms down dose to a lower dose. But given cabo's half-life that I mentioned before is roughly 100 hours. And given its accumulation in plasma over time at steady state, that washout period towards resolution of symptoms can often be 10 days, 2 weeks, even longer. So imagine you have a patient who has an adverse event in all patients with TKIs get them, that has potential overlap with a checkpoint as well. And so as a physician and as a patient, you have to try and understand was this cabo related? Is this nivo related? Do I hold one? Do I hold both? And if the patient has metastatic cancer, it's a challenging just dynamic to kind of work through. Obviously, cabo is successful and it's something that physicians have learned to kind of do. But our perspective was, if you can impact that half-life that period of time where you're kind of in that dose reduction, dose management, AE management phase, that's the more important thing because, simplistically, patients generally don't benefit from drugs that they're not on. And so when we thought about why we wanted to develop zanza, it was really about that combinability, that dynamic around having a more user-friendly and more patient and physician-friendly TKI that acts as a backbone. And so that was really the dynamic that kind of shaped a lot of it. And so again, we'll see in the cosmic truth on how this reads out in Phase III studies. But that ability to have a more kind of user-friendly experience because of the shorter half-life is something that we always sought out as kind of the core driver around development. All of the other stuff on the margin of maybe we're seeing some differentiation in certain areas and safety, certainly potentially intriguing signals, but one of the common themes I have as I look externally at a lot of these kind of smaller company assets, the caveats and challenges of interpreting small and unrandomized clinical data is significant. And so we got a lot of questions coming into this weekend around what's going to be at ASCO. And we're excited to share the data, really think that it shows that zanza is an active molecule. But at the end of the day, the much more derisking events for us as we think about zanza longer term are the Phase III studies.

So a couple of -- 1 that sounded like head and neck, but number two, I just want to put a finer point on this. It seems like the safety differentiation should manifest in terms of efficacy because patients are going to be on a higher dose for a longer period of time. Like if again, if you're an asset investor like because it's not like you don't say you're not -- when I talk with you, you don't say zanza is not differentiated. You're just saying it's too early, but is that fair to say it's going to be PFS and OS and discontinuations that we should be paying attention to? Or is there something else?

I certainly think there's a point of that dynamic. But again, taking a step back to where we started, We, again, don't really think about it as zanza versus cabo and the differentiation because it's really about how is zanza different and/or better than rego. How is zanza different or better than SUTENT. How is it better than pembro monotherapy in 305? How is it better than ever alignments in NET for 311? And so to me, that's actually the much more relevant dynamic. Certainly, there's going to be a lot of comparisons across the -- across our comparisons all this stuff. But when we, on the third quarter call, talked about zanza having a $5 billion market opportunity across all of the 6 studies that we're running, that's actually in reference to all of these standards of care that we're comparing it against and not this.

That makes sense, yes. And by the way, for that $5 billion, what are the -- is it fair to say the Merck studies are the 2 biggest contributors to that number? Or are we maybe underestimating zanza in nets? How do you break down that number?

Yes. So we think about it roughly is about 45% GI, 45% GU and 10% head and neck. So the GI piece, if you think about our business right now, it's probably 90-10 GU/GI. Our aspirational goal over time if we think about zanza and cabo growth in NET is a much more balanced dynamic between kind of those 2 categories. So the GU piece is obviously the 3 RCC studies. It's the to Merck studies in the non-clear cell. Nonclear cell is about 20% of all of RCC, utilization right now is a little bit of a hodgepodge across all approved therapies in kidney cancer, just because of the dynamic around labels are inclusive of that and then use is driven by kind of guidelines around on unrandomized single arm data, things like that. But the GI piece is then kind of the 311 study in 303. And so as I mentioned before, kind of a $1 billion opportunity across CRC. And we framed at least kind of that later line current NET opportunity at around $1 billion. And so as we think about moving earlier, tend to have more patients, tend to be on drug longer, that sort of dynamics. So it's kind of that 50-50 balance or plus minus.

I think the most interesting thing about the breakdown you gave there is you would think head and neck would be bigger, right. I mean if you think about what are the darlings in oncology and there are not that many. There's a few of these bispecifics. And you guys have a frontline study you're running, right? And I've talked to Mike about this before, and he's like, look Akash like you talk about single-arm uncontrolled data, you'll often have sites enroll healthier patients. I mean it is not just a comment on this line. And specifically, you see the oncology across the board. You have a 53% response rate with cabo and nivo, right? You have a lot of end in that indication. So when you think about the Merck partnership, it sounds like zanza plus pembro, so you kind of have the same 2 components, maybe version 1.5 version of some of these drugs. But when you're seeing all these other data sets and they're showing higher response rates, how, as an investor, can we feel like Exelixis has a plan to win in that head and neck arena. And it almost implies you're going for self populations in head and neck, maybe not the whole pie. Help me understand that.

Yes. I actually think it's not really about kind of that latter comment. If anything, when we look at that $5 billion market opportunity, it's a good example of candidly us being public-facing and passing the red face test is kind of what -- when we talk internally about market opportunities or projections, what's realistic? Does the model need to have 65%, 70%, 80% market share to kind of make the math work? And so I think it's a relatively believable, reasonable, conservative assumption is to say the reality is you can draw some comparisons between, say, RCC 10 years ago in head and neck. The RCC pie has grown over time because there have been drugs and therapies to really help patients live longer. And so hopefully, we're successful in kind of growing that pie overall. But I'd much rather be in the position of, say, we think it's a $500 million opportunity. But if we're wrong on the upside, that's great. I'd rather not set an expectation that assumes very challenging kind of share or duration or pricing assumptions. I'd rather say, okay, let's start from a place that $500 million is a reasonable large market and hopefully will generate a data set to improve upon that. But it's just...

But you're talking about like 10%, 15% penetration.

It's something that we want to make sure that we're realistic in setting expectations. But kind of getting back to your point on the pembro and zanza piece. Again, we looked at Merck's prior data set from LEAP-010 as a signal that clearly an IO TKI is active in this space. They saw a robust benefit on response rates. They saw robust benefit on PFS. And unfortunately, that didn't translate to overall survival. So the question that we asked is basically using the lens of the prior cabo/pembro data as a starting point. And I certainly echo Mike's comment that there's inherent challenges in interpreting Phase I data from here and here and here. To us, they're probably more similar than different. But we can look at that data as a starting point layer in the learnings from LEAP-010 and ask the question, can it potentially more user-friendly TKI plus pembro replicate a lot of that benefit on responses in PFS but also translate to survival. That's really the question of 305. Again, we'll find out -- the cosmic truth is going to come out when that card flips, but it's really using all of that information we have at our fingertips to drive investment in a market we think can grow.

A couple of things just to wrap on that. Number one, there are 2 cuts in terms of head and neck where I feel like none of the companies are really thinking about, a, you've just seen pembro get approved in first line. So a lot of first line is going to change. It's going to be basically post PD-1. And it's interesting, your data sets post PD-1 might be much more comparable to some of these like early bispecific data sets. So is there any appetite to run studies post pembro given that the adjuvant approval is going to really change what first line looks like? And then number two, we saw this with the Pfizer PD-L1 ADC that had data at ASCO. The HPV positive and HPV negative, there's different profiles that some of these treatments show in those populations. Is that -- so post pembro and then HPV positive and negative, are those areas where you feel like a VEGF TKI combo could actually differentiate here?

Yes. I mean the -- a couple of dynamics within that. I think our experience with at least taking the learnings from LEAP-010 is probably HPV status is less relevant than, say, maybe the EGFR bispecifics are trying to kind of frame that dynamic testing. Maybe it is, maybe it isn't. But we haven't seen as much of a kind of clear difference at least looking at the mechanisms that we're evaluating. As it relates to kind of other studies, other opportunities in head and neck, I think the way that we think about zanza is wave 1 is kind of the 6 studies that we've outlined right now, expect a wave 2, expect a wave 3. We think that zanza has the potential to be a broadly active backbone therapy, so to speak, across a wide range of tumor types, head and neck certainly is 1 of them, especially if we're able to kind of read out positive data in 305. The other dynamic that I'd kind of layer in there is that take the Merck collaboration as a model or even going back to 9ER in Bristol. One of the things that we've talked about is there's this through line or kind of tongue-in-cheek comment that we make sometimes is that we run Exelixis like a business and not a biotech. And so we're particularly thoughtful about expense management, how we think about new studies, how we think about just running the company. And 1 of those dynamics is if you look at a lot of these studies that we're running, they're actually in combination with partners, where there's risk sharing where there's cost sharing, right? So as an example of the 9ER study we ran that with our partners, Bristol or Bristol ran it, we paid for half. They paid for half. But actually, Ipsen and Takeda, who have ex U.S. and Japanese rights paid for half of our half. And so that was a really capital-efficient way to develop a large -- develop and run a large pivotal study. Similarly, with Merck, functionally, we're running 3 studies with them. We're each paying for roughly 1.5 and sharing free drug. And so I would say an important dynamic to look for going forward is we like that model and we want to replicate it. And so it allows us to have a broader ambition for zanza for that next wave and that next wave and that next wave, while remaining capital efficient and candidly just thinking about spend in a probably more disciplined way than many of our peers.

So is it fair to say in terms of potential partnerships going forward, I mean, you've done mostly with large companies when we think about cabo development, the Merck deal does not exclude you partnering with other large cap players in the space when you're exploring that kind of VEGF signal across the board, because there are certainly other indications.

Yes.

Let's just hit on commercial performance. I thought what was most surprising last year, your -- the management team was pretty clear. First-line RCC, we're seeing uptake kind of start to plateau. We shouldn't expect that to grow much more come to 1 year later, you're actively taking share in first line. And so I want to understand, is that patients waterfalling? Or is that actually new patient starts are now starting to trend the other way. How does that occur so late into a launch? I think investors don't understand.

Yes. I mean, again, a little bit tongue-in-cheek, drugs are successful or drugs perform based on the data and based on the team. And the data, we obviously had a 5-year update to 9ER at ASCO GU earlier this year. And I think it's a data set that, frankly, continues to resonate with that physician prescriber base. And then on the team side, I think we certainly believe that we have a best-in-class commercial organization, not only in biotech, but across pharma as well in that kind of GU landscape. And we have the ability to do things from an analytics perspective that allow us to continue to kind of be very targeted and take share. And certainly, we see that stacking of patients of long duration over time. But our team continues to have their foot on the gas continues to focus. And I think 1 of the challenges that our bigger competitors, bigger peers have, is that we have the ability to just focus on cabo, right? They don't. And so having a team that is completely dedicated and focused to maximizing the value of cabo, that's an advantage.

All right. I'm going to stick in 1 more question. I know we're out of time. NET's launch. I think the big question is there is obviously use of cabo in that population already, but there's something about getting it reimbursed, getting it on the label. And you guys were pretty clear the increase in guidance was not necessarily NETs driven. Is there going to be a bullish on the NET's population once that approval gets online?

Yes. So P.J. talked a little bit about this on the 3Q call. So given the later line in more advanced stage of disease that these patients, generally, we don't expect a bolus with kind of these lip-line patients. But what I can say is that the dynamic that we have talked about around cabo and NETs is that there's this inherent familiarity with cabo that physicians have. And 1 of the things that we've seen early on that we've talked about is we're seeing scripts for the 40-milligram dose. So approvals at 60, but 40 milligrams is indicative that these are physicians that use cabo for, say, 9ER are familiar that, that 40-milligram dose is still very active but more kind of more tolerable, gentler. And so it's kind of an early sign that I think is really positive that those kind of NET physicians have this familiarity. And so we've said that we'll update the street when we have a better sense of that launch trajectory. But some of the early signs that we're seeing on KPIs are really encouraging.

Got it. Thank you so much. I really appreciate everyone for joining us early in the morning. And Andrew, thanks.