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

[Technical Difficulty] Health Care Conference, and it's the first one at TD Cowen. I'm Yaron Werber from the Biotech team, and it's a great pleasure to moderate the next fireside chat with Mike Morrissey, our President and CEO of Exelixis. We have Chris and Susan here as well. Mike, it was good to see you.

Yes, you too.

Thanks for joining. I appreciate it.

Yes. Great to be here live again.

So lots to talk about. Cabo continues to grow annually and the pipeline is coming now into late-stage development, early-stage development, multiple different new platforms coming through, but let's maybe talk first on cabo because it's still very much the growth driver. You gave guidance of growth about 13% to 20% year-over-year. What's driving that? Is that still share gains? Is it durability? I don't know how much you can comment between renal and liver as well?

Yes, for sure. Well, again, great to be here. Congrats on the meeting having it live again. It's great to see the hallways buzzing again, so congrats there.

Yes. [indiscernible].

I'll be making -- yes, for sure. I'll be making forward-looking statements today, so please see our SEC filings for a description of the risks that we face in our business. Yes, cabo growth has been strong the last couple of years. Overall, we've seen a double in net product revenue since the end of -- since 2020. So that's been nice. Obviously, the 9ER data is -- was strong, continues to be strong. We had the 44-month median follow-up data at ASCO GU. About a month ago, that was the hit of the meeting. The maturing survival data there looks impressive, median of about 50 months now. So really challenging a lot of the dogma around expectations for survival with an IO/TKI best-in-class molecule in terms of RCC, top TKI for RCC in terms of a combination with an IO and then top in the first-line setting and then top second-line drug as well. So we're pleased with all that. In terms of growth, again, we've seen strong growth since the 9ER approval, right? So if you look at that over the last 8 quarters, first 2 quarters, we grew about 5 share points, and then we've been growing a single share point since then over the last 6 quarters. So that's -- again, a single share point might not seem like a lot, but when you look at what that's worth to us in terms of top line revenue, that's a big number and to see that consistent growing. And that's due to both very solid NPS growth. So we're starting a lot of new patients in the frontline setting and then the fact that we have long duration of therapy with the PFS in the 16-month, 17-month range from the 9ER study. You've seen this compound stacking of patients year-over-year helps too. So in terms of '23, the growth is a mix of both certainly more NPS, more new patients start as well as that continued compounding now 2 years in, have a lot of patients who are on study and are on drug and doing well in their frontline setting. So we're excited about that, not certainly going to rest there in terms of being content with that data, but we're very aggressively marketing the drug per label and hoping to see that continue to grow in the years ahead.

Is most of the growth from first-line renal from new adds or is it from durability?

Yes. It's a combination of additional, yes.

Yes, for sure. And what's going on with share in second-line? I assume you're gaining in first and second is sort of stabilizing or maybe...

Yes. If you look at the BrandImpact data, which is probably the most concurrent publicly available data where they look at share point and you can argue about the methodology, but it's a good, I would say, a benchmark. Frontline, we're in the 20% to 25% range consistently. Second-line, we're in the 40% range overall. So that's a very strong place to be in terms of having a majority of impact on patients in those first 2 lines of therapy for RCC.

And is -- and what's -- and how are the trends? 40% sounds like it's been fairly stable. So is the growth really just capturing share in first-line or you're seeing a little bit of a shift?

Yes. No, it's really growth capturing the first-line setting and keeping the second-line constant. Yes.

Okay. So let's talk about maybe COSMIC-313, you [ upbeat ] on PFS survival, obviously, it's been a little too early. We're about 20 months so far in terms of follow-up. The activity was overall, but there was obviously a better activity driven by intermediate risk than poor risk, and you're still waiting for an update on survival, I believe, last year.

Right.

What are you expecting to see? I don't know if you can share with us? And what do you think you need to see to get a label?

Yes, it's a good question. Obviously, we've had a lot of discussions internally and with the FDA on that topic. So I don't want to provide too much oversight on that right now. I think what's safe to say is that the mature data is the most important thing as we accrue more events. And we have the second interim coming up in the second half of this year. So we'll see how that data looks. If it's appropriate, we'll talk to the agency about that. If not, we'll go to the final analysis and see how that goes. So obviously, bar is pretty high here. PFS has historically been enough for approval. I think the winds are shifting a little bit in terms of expectations and the overall risk benefit here. So survival is an important part of that equation, and we'll get more mature data and see how that looks.

Right. I mean the hazard ratio was already 0.7 on PFS, right? So that's unlikely to change at this point, right?

Yes. Exactly.

So it's a question of -- I know you haven't commented a lot. The second analysis usually sort of 50% of the events. Is that even in the [ black ballpark ]?

I would say a bit more.

A bit more.

Yes. Yes.

Okay. And the final sort of in the 70s would be or...

Well, final would be 100% of...

And see it all the way, yes, okay.

If we have to go that far. So again, I don't want to project too far. Let's get the next interim done and see if that...

Yes, next one should definitely maybe more than 50%, probably even -- okay. Got it. I mean survival, if I remember correctly in this setting is -- well, the data has been sort of maturing, right, but it was almost 40 months in first-line when you look at some of the competitors. Am I off?

It really depends. And the question is, say, with ipi/nivo, you're looking at intermediate poor with the TKI/IO combo as you're looking at all 3 with favorables, obviously, will right shift the [indiscernible] pretty dramatically. So it's all trial dependent and situationally dependent in terms of what you're looking at and the ratios and those kinds of things between the 2 -- [ 3 ] different results.

Yes. And is there a chance -- and I know I'm jumping ahead here, but is there a chance to get a broad label or you might still be driven by the intermediate risk cohort?

I wouldn't want to speculate on that right now. That's a discussion for us and the agency at the appropriate time. Look, we're looking to enhance the opportunity for patients to access cabo as a single-agent, as a doublet and potentially a triplet. So I think we'll -- we'll get the data that we need. We'll work with the regulatory bodies to figure out what makes the most sense there and then move forward appropriately.

Okay. Right. But in principle, this was a study in essentially both cohorts, right? So it wasn't really cutting into a sub-segment is not probably where you want to go?

Ideally, the bigger population is better, but that's going to come down to the data and how the other various stakeholders view that data relative to standard of care and risk benefit.

Okay. If there's any questions in the audience, just yes, by all means, let me know. Let's talk about CONTACT-02 in metastatic castrate-resistant prostate cancer. I think we're expecting an update later on this year. Just given the historical data, right, about 18% response rate, what -- from the prior cohort I'm talking about COSMIC-021, what are you expecting from this? And how much of an important indication is it for cabo at this point?

Well, it's certainly a large population, even considering the subset of patients that we're looking at with measurable disease by RECIST 1.1, either visceral mets or extra-pelvic lymph node mets. So it's an important opportunity here. Obviously, we have a lot of history with prostate cancer with the cabo molecule going back to the early days in 2011 with the bone scans and the common trial. So yes, it's been a long story. Again, the primary endpoint for CONTACT-02 is progression-free survival. We're going against the second NHT, which is, I think, is a relatively low bar. So we'll see how it looks, right, from the standpoint of how PFS for the cabo/atezo doublet compares to a second NHT in this relatively defined population of patients with measurable disease. So it's the cleanest experiment we could run to be able to really answer the question, the impact of that doublet on disease progression in that population. If it works, we have lots of opportunity with XL092 relative to how we might play that zanza, how we might play that going forward. And we'll talk more about that later, I'm sure. But prostate cancer is an important opportunity. You've seen some -- we've all seen some field trials there recently looking at IOs by themselves or in combination with various chemotherapies. So it's an area that really needs new therapies, new approaches, and we think this is an important one to test, and I'm looking forward to having those results later in the year.

And what's considered clinically meaningful in this? Is it sort of 6 months is the bar or you want to see kind of a 3 months difference from...

Yes. Take the p-value framing the right way, and we'll navigate the rest. Yes.

Okay. Great. So let's talk about zanza. I mean, you've naturally alluded to that already.

Or we go on to, let's talk about CONTACT-03 real fast.

Okay.

Second-line renal trial. I want to just -- it was a trial that read out last week. It's interesting dynamic and certainly opportunity for us. We ran that trial for a couple of reasons to first ask that it was a very clean way to ask the question. The sequencing 1 checkpoint inhibitor after a patient has failed a first or second checkpoint inhibitor, does that actually add value in terms of risk benefit for the patient? And secondly, it was really asking the question for single-agent cabo, can we actually benchmark what that would do in the context of a patient who has progressed on a checkpoint inhibitor. And the important biology here is that it's CD8-positive T cells live a long time, even after you stop dosing with a checkpoint inhibitor. So it really asked the question in a very broad sense. Can you magnify the activity of that relative to adding on a compound like cabo. So it was a very interesting experiment. Trial didn't work in terms of showing a differentiation between the doublet cabo/atezo versus cabo. There's been a lot of chatter around, well, you guys used the wrong checkpoint inhibitor, atezo wasn't that good, blah, blah, blah. I think the reality is in our hands in the couple of cohorts from COSMIC-021, we were able to actually show that cabo plus atezo either at the 40-milligram dose or the 60-milligram dose of cabo actually had very similar activity in terms of response rate in PFS to what we saw with cabo and nivo in that combination, response rates in the 50% to 60% range, PFS in the 15-month to 20-month range. So we feel like it was the right checkpoint to use. And the question is, what does the biology tell us about what's happening there? And we'll talk about the actual data when it comes out this summer. But truly, I have to note that cabo by itself performed really, really well. And in fact, single-agent cabo in the control arm actually met all of our hopes and aspirations for what the doublet would do from a duration of therapy point of view. Okay. So it wasn't so much that cabo/atezo didn't work, but that cabo by itself worked really well, and that's probably because all these patients have activated CD8-positive T cells already in place. They don't need to be activated further by a checkpoint inhibitor. It's already happening, right? So data coming out shortly, but I'm really proud that we and the investigators did the right study, asked the right question, it's going to be really important for, I think the community because a lot of patients are sequentially treated with one checkpoint after another. In this case, I think it really was at least for renal that, that's probably not the best way to go.

Okay. So there's still sort of a win here for cabo alone.

Well, from a duration -- yes. From a duration point of view, if we're seeing in this population, the duration with cabo that we were hoping for with the doublet, then yes, cabo does well there too.

Okay. So let's now move to non-clear cell RCC. So moving on to zanza, obviously, the STELLAR-304 study zanza plus TIVO head to head against Sutent. And maybe put that also in context of the prior COSMIC-021 data with Cabo TECENTRIQ, and sort of -- I'm trying to -- cabo TECENTRIQ is obviously not approved for NCC moving on with the next-generation. So maybe just kind of set for us what do you want to see from a bar? And then how much cabo is actually used right now in NCC?

Yes. So the latter question is, first, it's hard to actually break that out. We don't have real good analytics on the breakdown from a commercial point of view between clear cell and non-clear cell of different types -- all different types of non-clear cell. Historically, it's a 15% to 20% mix of the overall population. So I would assume that kind of tracks the same. The label for cabo is broad enough to cover all forms of RCC. So that's not really even off-label per se. It's just part of the overall experience. But there's very little randomized data in that setting. And again, from the standpoint of pushing the wave of zanza pivotal trials forward, we thought that made a lot of sense as the lowest hanging fruit in RCC. We have a lot of interest in looking at novel triplets in RCC with both zanza plus a doublet of IOs versus zanza plus an IO plus other modalities. So more on that as those trials evolve and get started. But again, I think this is a great place to start and to kind of reinforce our interest in the franchise of kidney cancer and then to build out the zanza pipeline within RCC as well as other tumor types in GU and GI and thoracic and beyond.

Yes. So with zanza, again, you're sort of going back old school and doing this study with OPDIVO and now with TECENTRIQ.

Right.

Why is -- so why is that? Are you just trying to build on what the already approved...

Yes, I think the overall approach is to be somewhat agnostic in terms of the checkpoint that we use. So we've got nivo with STELLAR-304 in non-clear cell RCC. We're using atezo in STELLAR-303 and the third-line CRC, and we have plans to use other checkpoints in future pivotal trials that we'll announce this year. So it's really a catch-all from the standpoint that we view those as all more or less the same, more or less interchangeable, and it's just kind of mixing and matching based upon the different situations and opportunities.

And STELLAR-304 is Bristol providing OPDIVO. Okay. And I assume Roche is providing TECENTRIQ for STELLAR-303.

Yes.

But the commercial rights are just yours?

They're all ours. Yes, exactly. Yes. Yes.

This could be label-enabling for the checkpoint as well, that study or...

Sure. Sure. It's the combination that is involved, right? So...

Got it. Okay. And so the -- going back to STELLAR-303, right? So now you're talking about zanza and TECENTRIQ and third-line CRC. This -- I assume you're building on TECENTRIQ from the prior study, and that's why you're...

Well, really, it's more the cabo data that we had at ASCO GI in '22. We had 2 different studies published where we showed a response rate in the 25% to 50% range with 2 different checkpoints, strong PFS data, all the caveats of it being non-randomized, but strong PFS, strong survival. So the unmet medical need there is very, very high. So we feel like that's, again, relatively low-hanging fruit that if we can win, it's a big move for patients and certainly a great start once we file and would get that approval on the market.

Okay. So far, what you -- it sounds like you're going to colon, you're going to NCC RCC, what's next for zanza?

Yes. So that is to be defined this year. So we've got at least a couple of more trials I'll start this year and then certainly more after that. The overall framework here is to look at both existing cabo indications where we think we can design better trials, better combinations, better doublets and triplets that would allow us to raise the bar in terms of standard of care for patients as well as new indications, where we have some activity, some historical activity with cabo, having pursued full development, so we can then engage there with zanza plus any combination partner, which makes sense. So -- but again, I'll just reinforce, we're not looking to just be as good as cabo, right? We're looking to design better trials, better combinations. We have to improve standard of care, right? We have to be able to move the needle for patients because that's the only way you're really going to be able to make a difference from a pure marketing and overall commercial point of view. The days of having need to drug [ STELLAR ] just over, right? So it's all about bringing more value to patients, right? So...

So data from STELLAR-002, that's the Phase Ib testing combination with zanza with OPDIVO or a combination of OPDIVO and YERVOY or with Opdualag. Are we -- this -- and you're doing obviously dose escalation with all the cohorts. Any data released this year or sort of what's the next update?

Yes, we'll see. It's all about execution this year. So if we have the opportunity with mature data, then we'll certainly present that. It's really important for us to launch pivotal trials, move compounds forward, that's the priority, right? So -- but again, you mentioned this on the call and certainly, again, today, the combination of zanza plus rela plus OPDIVO is a very attractive approach based upon some of the early melanoma data, have a great relationship and history with BMS. So we're excited about that. And if we can move that into multiple pivotal trials in the short term to be defined across tumor types, that would be a great way to go, right? If you look at the some of the COSMIC-313 data, ask some of the questions about where you can improve upon that just in general, well, maybe swapping out ipi for rela might be a better way to go having zanza shorter half-life, maybe better tolerability for cabo. So again, we're looking for winning combinations that we can then go against standard of care and improve outcomes for patients.

Yes. Okay. And so with STELLAR-002, it sounds like you're still in the dose finding and a top point, you'll do a whole bunch of escalations, COSMIC-021 or...

Well, the question is once we have a dose, what do we do, right? And we're not -- see, we're not shy about moving forward aggressively into pivotal trials if we have a dose, and we did that with cabo/nivo, cabo/nivo/ipi, right? So I would expect we will act in a similar fashion here. Yes.

Okay. Got it. Any questions from the audience? Then let's talk about -- the one element I did not talk about is just litigation. The next trial has now gotten pushed from the spring to the fall, right? And...

In October, right.

It's in October. So I'm thinking the decision is probably more sort of next year, the way -- well, the last judge took 6 months, 7 months, 8 months to roll. I know you can't say a lot about that. The litigation with Teva essentially stayed pending the second case with MSN, right?

Yes, that's right. Yes.

So that's probably not going to -- and I would imagine even pending a resolution with MSN or a verdict that's still going to be whether that litigation starts right away, whether it waits for an appeal?

Yes. Again, I wouldn't want to speculate on what timing wise or what could happen? Yes. Yes. Sorry.

Okay. I got you.

But you're right. So the first case was a split decision. Second case is pending a lot of confidence in the data and the depth of the IP. And certainly, the team is both internally and externally is first class. So we're confident in what we've got and understand the value proposition. So off we go.

Okay.

Okay.

Before we dive into the early pipeline, you're sitting on a lot of cash. I get questions all the time, does it make sense for you to do something bigger? And it's something we've talked about for...

Yes. Sure. Sure.

Valuations have come down obviously. Valuations are fluctuating up and down all the time.

Yes. Yes.

Why not go out and buy something in Phase II?

Yes. Well, the issue isn't -- for us, it's not valuation. It's conviction in the asset, right? And to be frank, there are a -- there aren't really a lot of high conviction assets out there right now, right? So -- and I think that's not just our view, but that's the view of the general biopharma world because the number of big deals has been relatively small in oncology for the last few years, right? So we're looking closely. We have a pretty short target list of, I would say, high-value assets that we are pursuing. Those might work. They might not work. We'll see. The 2 option deals we did at the end of last year with Cybrexa and Sairopa, I think are a good example of how we can put the right amount of cash to work to be able to have access to clinical stage assets and then essentially pay for success if that comes to fruition later. We're looking at later-stage assets, those that are in late Phase Ib, Phase II, have reasonable good data, and we'll see if we transact. Those would come at a bigger price tag, but we think that's warranted based on our lens of what success could look like there. But it's all about -- again, we're really informed by the cabo experience. And again, cabo is a drug that was returned to us twice, failed pivotal trials, but we had a lens to the activity of that molecule that allowed us with the data we had and the belief that we had in the overall approach to be successful commercially for a lot of patients and generate a lot of revenue. So I think that same lens is being applied to here. We're very disciplined. And if we do transact on a bigger scale, then it will be because we have the conviction that it makes sense from a pure clinical commercial standpoint.

Yes. So let me actually -- one thing I'll take a step back for a second and back here. One of the things that we don't have visibility on and obviously, of your planning internally that we don't have, we obviously don't have a view at this point of what the IP is going to be, right? Is it [ '26? ] Is it [ '28? ] Is it [ '30, '31? ]. And it's obviously 4 years that make a big difference, right, in terms of not just where the TAM and area under the curve and talk about your ability to then replenish, right? You're in Phase III now in colon, third-line is interesting. That's not going to be a cabo replacement, right? What I'm going with this is that you -- over the next 12 months, you need to start moving in the Phase III. They can have data on about 2 years launch, right, and be ready for the '26, '28 time frame, the worst case, '28, '30 in the good case, right? So how are you thinking about all that? The early pipeline is interesting, but it's just to risk into...

Yes. So the clock number stops, and we talk about that a lot back in Alameda. The temporal nature of what we're doing now internally, the value of speed and moving zanza forward, XB002 forward, the early-stage pipeline that is the -- in the IND-enabling phase right now. The beautiful thing about oncology is that if you pick the right targets, you make the right molecules and you do Phase I the right way, you can get hints of activity early on and move quickly into Phase III. And that was the cabo story to a large degree, right?

Yes.

So we feel like -- first of all, we're very confident in the 2030 window in terms of LOE for cabo. So we've got room to maneuver. And we have a very deep pipeline of assets that we think can really generate the right level of data that we can then monetize appropriately if we move them into pivotal trials quickly. And that's been our MOA for years, right? There's a lot of companies and no judgment here, but there's a lot of companies that almost avoid making that transition because it gets expensive fast. You've got to put all your chips in the table and say, I'm going to bet the farm on this approach. Well, we've done that. We have the confidence in that and the conviction that we know what we're doing there relative to -- every trial is not going to work, but we know how to play that game pretty well. So all that being said, I think acting -- we don't feel any level of desperation to put money at work, take big bets. We have a sense of urgency and a sense of focus and a sense of energy in doing the right science, the right clinical science, asking the right questions. But we've got -- we have time and the luxury of having something like cabo that generates free cash every quarter. So we can take these shots on goal in the appropriate manner, and we hedge our bets well with the internal pipeline, things that we can source externally and have a broad portfolio. It's all about the pipeline and playing that pipeline game successfully.

Yes. So let's talk about -- the early targets are super interesting. The bigger question is just how fast to your point you can move them forward. With XB002, you only differentiated, you can see an early look versus TIVDAK. The question TIVDAK in cervical is not going to be big and also therapeutic window and cervical is just not a big indication. What's next? And I know you've got to figure out the dose still between 2 mgs per kg and 3 mgs per kg, but you already got a differentiated profile early on.

Yes. Yes.

Over and above cervical, what's the next tumor type that makes sense?

Well, we have -- so the dose expansion cohorts include 10 different histologies. So we need to lock down the recommended Phase II dose and all the Project Optimus machinations just complicate that to a certain degree, and we'll do that as part of dose range finding and then certainly in the dose expansions. But to have 10 different tumor types that really allows us to ask the question, including cervical. So we benchmark there. But are there tumor types because XB002 has a differentiated safety profile, has arguably better exposure and a more potent toxin, right? Do we see tumor types that are insensitive or just modestly sensitive to [ TV ] actually have over sensitivity to XB002. That's the home run, where we can go from cervical plus indication X, Y, Z versus just cervical by itself, right? And that's either as a single-agent or a combination.

So Project Optimus is -- look, it's spot on, but it complicates life when you're looking at 10 different tumors, right? And so is it that you need to find -- do you have to find the right dose in each tumor? Is it that you find...

Well, I think you would use each or a subset of those tumor types to triangulate to the right dose for all of them. I think that's the issue. Correct? And again, the question is when you do that and how you do that. And that's -- it's really hard to do that unless you have a sensitive tumor type because how else you judge activity. You can judge safety and tox and tolerability, but if you can't judge activity, then it's really hard to do that, right?

I guess cervical is a good starting point, but it is an approved drug?

Maybe, yes, maybe. Yes. I'm not convinced to that.

You're not convinced to that. Okay. Is 2 mgs per kg sort of the lower end or it's unclear. Could you potentially be even...

I think it's the lower end right now, but we'll see. That was the data from last year, right? So -- and 2 mgs per kg at the exposure that we're getting is actually it packs a punch, right, relative to what TV has, right?

So -- and I think a lot of companies, and we're asking everybody the same question, it's -- some companies have to contend with that when they're doing fairly rare tumor-agnostic approaches, it's not that easy to roll.

Yes.

But how do you determine the lower -- lowest effective dose in the Phase I based on durability to...

I believe we can't do it in Phase I. You probably have hints of that, but you really refine that to run, when you have more patients, right?

In Phase II.

Yes.

So it's more than a single dose, you got to do a bigger Phase II.

And the question is how and what's the most efficient way to do that, right? So...

Yes.

Yes. And that Phase II is really Phase Ib to a large degree because if you want to define that window, that therapeutic window across cohorts, then you can generate a lot of data across cohorts, you pick the winners and you go to Phase III, right? I think that's the optimal, and we did that with MTC, with RCC, we actually ran very few Phase II trials with cabo. And we were on -- kind of at the forefront there, that was Gisela's [indiscernible]. That was her kind of insight into how to play that game and she was spot on.

So in the novel world, this is -- we're talking about 20, 30 patients, multiple dose Phase II.

Yes. Yes.

And it doesn't have to be statistically significant versus placebo at that level, it's got to be telling?

Yes. Yes. You'll see activity when you've seen this at the right dose. Yes.

But I'd say we've about a minute left. Let's talk about CDK7. Where -- the big question is ultimately which indication. Obviously, those seem there's some competition here as well, but more -- what do you think the biology is most...

Well, I think that's the issue. I think the big issue is, is it a target of interest, right, because there's no clinical POC, no clinical proof-of-concept here at all. So we have the right probe, it's potent, it's selective, it's irreversible to ask the question, does this molecule, does this approach of basically irreversibly inhibiting CDK7 do anything? I think that's the first big question. So -- but I like the scientific part of that. We're asking a very fundamental -- fundamentally important question around the biology of cell cycle, and we're going to get that, I think with this probe. It may be good news, go into indication X, Y, Z, it might be [ bottom ], and we'll find out, right?

And how -- do you tell that from Phase I? I mean, the magic here is you want to...

Well, I think we'll find it within Phase Ib, yes, sure, because we have the right expansion cohorts to be able to feel good about that. Yes. Yes.

Yes. Okay. Great, Mike. Good to see you.

It's always nice to see you. Thank you.

Excellent.