Exelixis, Inc. ($EXEL)

All right. Good afternoon, everyone. I'm Stephen Willey, one of the senior biotech analysts here at Stifel. I'm glad to have with us for the next session, Andrew Peters, who's the Senior Vice President of Strategy and IR at Exelixis. Andrew, thank you for the time. It's always appreciated.

Yes, great to be here.

Maybe we can jump right into zanza, which obviously remains the focal point of an expanding development effort across a lot of different tumor types and lines of therapy. So you've got this December PDUFA date that's approaching in third-line colorectal based on the STELLAR-303 data. Just -- how do you think about the value proposition of this regimen in this setting? And do you think that there's a subset of later-line patients for whom this regimen might be best suited?

Yes. Steve, glad to be here. Just as a reminder, I'll be making forward-looking statements today. So everyone online, please see our disclosures around risks to our business and our regulatory [ tiles ]. So it's an interesting place to start, and I'm glad you started kind of the discussion on zanza that way. Highlighting the breadth of what we're doing with zanza is really how we think about the business. Mike talked about it on the last earnings call, and we certainly highlighted at R&D Day, kind of this idea of franchises is really what we're trying to do. The breadth of development with 7 ongoing or planned pivotal studies with zanza. And certainly, one of those core areas is not only the zanza franchise, but the CRC franchise as well. So the foundation of that is the 303 data with zanza, and it's one that we want to continuously build on. I'm sure we'll get to 316 later kind of in that post adjuvant post definitive therapy CRC space. But really, this idea that we want to establish a leadership position in CRC and really built from there. And we think that the 303 data certainly enabled us to do that. So on the heels of kind of the positive data that came out at ESMO last year, showed a survival advantage against a very active standard of care. It's something that as we've done market research and are really starting to ramp up a lot of that launch prep, we're finding that the patient community, the physician community is especially excited about having another option. I certainly wouldn't frame it as we're thinking about it as only targeting a segment of patients. The way we see the landscape is it's actually quite fragmented. I think we've talked about this before, that third-line plus CRC is about $1.5 billion addressable kind of market today, with about 1/3 in that sunlight regimen, 1/3 of TKIs and a 1/3 kind of a mixed match of a bunch of others. And so our job, and I think something that we've been especially good at in the RCC space is just focusing and being competitive and trying to gain as much market share across all of those segments as possible. If you look at our data, kind of the 4 spot shows that there's no real patient population that drives that activity. It's actually pretty robust and consistent benefit across all of those different subtypes. And then layering that on top of zanza/atezo offers a chemo-free option, an IO-containing option. Those are the sorts of messages that we think are starting to resonate in a lot of the market research that we're doing. And we're certainly excited, if approved, to go out and have those conversations with providers about having another option for their patients.

Okay. One of those subgroups is this non-liver met subgroup. And you've got this prespecified OS analysis from which we're expecting data sometime, I think the middle of this year. How, if at all, would you expect this data to impact regulatory discussions or labeling discussions? And again, I know this was prespecified as a dual primary end point, but you did hit [ stat sig ] in the ITT population. So why does this analysis, I guess, even matter to that extent?

Yes. I mean, the simplest answer to the question is the more data, if positive, that we're able to share with providers and patients to kind of complement the story that we've already generated, the better. I mean, if we're able, again, if the data are positive, having a label, say, Kaplan-Meier curve for the ITT, Kaplan-Meier curve for the non-liver met, that just adds to kind of the robustness of that conversation that we can have with providers and patients around this option for patients. Certainly, as it relates to kind of how the NLM data will impact our ongoing discussions with regulators, we certainly see this as kind of par for the course. This -- it's pretty standard in any ongoing review for the agency to ask for updates to data, especially on something like survival. And so we think that this is just something that will kind of fit in with that normal course of review. You did highlight that this is part of that dual primary end point. Factually speaking, part of the reason that data at ESMO, data midyear is really about kind of the time course of disease for these patients. Patients who have, unfortunately, liver metastases just tend to see events much faster. But this is kind of much more of a standard update than anything.

Okay. Maybe just continuing on zanza for a bit here. So we get a lot of questions around safety tolerability at the 100 mg dose. Just given that if you look at median dose intensity from STELLAR 303 and some of the other Phase II data sets that we've seen, it's kind of a fraction of the specified dose. So can you just talk about the dose optimization work that led to the selection of this dose for this initial suite of Phase III studies? And then I know that you've talked about other or additional dose optimization efforts that have been initiated to inform dosing in other combo-based trials and earlier lines of therapy. Is there anything you can say about how those additional efforts have kind of informed zanza dose selection for the next suite of trials that you're looking to enroll?

Yes. So again, I think it'd be helpful to kind of set the framework around how we think about just dose in general dose in 303 and dose for the other studies. So at Exelixis, an incredible amount of work kind of goes into that process, not only around generating a significant amount of early data, but we have an incredible PK team that does exactly this, focuses on a lot of in-depth modeling around exposure response, understanding tumor sensitivity, line of therapy, combination partners. Those are all of things that really play a role in kind of how we determine dose selection, not only for 303, but other studies as well. One of the things that we talk about at Exelixis is kind of this cabo lens helps inform everything that we do at the company. And the learnings and kind of the process of development of cabo certainly followed the same path where we started, say, the monotherapy at a higher dose as we move in later lines as we moved into earlier lines of therapy or different cancer types. And certainly, in combinations, we kind of evolved that to really focus on clinical benefit risk, tumor sensitivity, combination partner, what the biology really tells us and what the modeling really tells us. And so we're kind of following that same playbook, so to speak, with zanza. And so we don't necessarily say, okay, there's 1 dose across all indications. We have a lot of in-depth work that goes into kind of dose selection for each individual study. And we think that, that's kind of the best approach we should take. One of the dynamics that we talk about a lot is -- we want to make sure that our job, ultimately, at the end of the day, to create value for all stakeholders in the company, is to run successful trials, not just run studies. And so dose selection and optimizing dose for each of those kind of components and incorporating a lot of that exposure response modeling and all of the other things that the PK group does, that all kind of fits into how we think about, is this going to be a positive study? Is this a study we want to invest in? And is this study going to have the ability to shift standard of care? Because the only way that we're going to be successful commercially is if we're able to define a new standard of care for patients and help them live longer. And that's certainly kind of what we're aiming to do.

Okay. Kind of another near-term zanza catalyst that we're anticipating is top line data from STELLAR-304, which is in non-clear cell, in combo with nivo. So I know that there hasn't really been any prospectively generated data in this setting from a Phase III perspective, but IO/TKI regimens that are already listed as preferred treatment options and NCCN treatment guidelines. So what kind of commercial opportunity do you see this 304 data unlocking? And how do you think these data will then change the prescribing narrative going forward?

Yes. 304 is an important study for us, and importantly, for patients as well. Candidly, I think as we were starting to strategize around what we wanted to do in RCC, this idea of really no standard of care in non-clear cell honestly surprised me a little bit. The fact that there's never been a large randomized study in this space is somewhat unique in oncology, and it's really a fact of -- all drugs, all combinations that are approved in RCC have a label for both clear cell and non-clear cell. And uses, as you mentioned, is really driven by kind of guidelines. But again, those guidelines are informed mostly by single-arm, unrandomized kind of Phase I/II type studies, which we all can appreciate the challenges of interpreting that data because so much patient selection can oftentimes go into that. And so from the physician and patient perspective, understanding, well, how does this data set apply to the patient that's in front of me can be challenging. And so what we wanted to do with 304 was really plant a flag in the ground and say, okay, with Level 1 evidence against an active standard of care, is zanza/nivo able to show a benefit? And that's what we're hoping to read out, looking forward to reading out later this year. Again, the feedback we've gotten from the RCC community is exactly that, is that they've been waiting for something to help inform a lot of those treatment decisions. Because from their perspective, reading the guidelines, you're basically trying to pick, well, do I trust this small Phase II study or this small Phase II study or this small Phase II study. And all the caveats with cross-trial comparisons, single site versus a few sites, what are the patient characteristics of these, where they preselected. All of those dynamics are certainly important. And so 304 offers a chance for us to really define a standard of care. And again, with Level 1 evidence, if positive, hope to move up kind of and become that clear recommended combination.

Okay. Maybe just sticking to RCC. So we just learned about the second Phase III zanza trial that you are running in collaboration with Merck. I think it's LITESPARK-034. We just saw Phase III data from LITESPARK-011, which was generated in pretty much the same kind of patient population. So how do you think creates a second-line foothold for zanza in the context of this 011 data that we saw I think at ASCO GU earlier this year?

Yes, ASCO GU. Yes. I mean, I think -- one of the dynamics that really -- that informs how we think about zanza development is the idea of cabo's the TKI of the 2020s in RCC. How do we make zanza the TKI of the 2030s in RCC? And certainly, a big part of that, and I'm sure we'll get to it later, is how the way patients are treated can evolve over time. And so a lot of that is understanding, okay, let's not run a study for how patients are managed today. Let's run a study for how we think patients will be managed in the future. So what the LITESPARK-011 data show us is HIF, VEGF TK combinations certainly can play a role in the patient's journey with RCC. What we want to define with both 033 and 034 as well as other studies that we're planning with zanza is how can we improve and iterate on that. The totality of the data that's been generated so far across all of these different regimens, all of these different combinations, whether it's 05, 11, 12, 22, those inform a lot of those decisions. And so we ask questions like every patient who's coming in and trying to understand how they're going to be treated, is it better to combine? Is it better to sequence? This dynamic at ASCO GU that came out after that data were presented was, well, if there's no clear survival advantage based on what we've seen so far, the absolute PFS or, say, a TKI monotherapy followed by what we've seen, say, with LITESPARK-05 with belzutifan, how does that compare with what the combination does? So it's going to be on us in generating data with LITESPARK-034 to more robustly answer that question, and that's really our goal. As I mentioned before, our job is to kind of right shift that standard of care. And so the question is, if LITESPARK-011 shows that this combination can have an effect and positive impact on patients, can an optimized TKI like zanza, best-in-class TKI further improve upon that? And then similarly, kind of with 033, as that treatment dynamic changes and more and more patients get treated, say, with adjuvant pembro in frontline, what does that treatment option say should be for patients? So it's kind of that -- what's the expression in hockey -- skate where the puck is going. That's kind of how we think about development of zanza and RCC.

Okay. So you mentioned 033, which was the first RCC collaboration that you announced with Merck. Just a couple of questions on that, right, because that's looking at belzutifan and zanza as kind of post adjuvant IO frontline therapy. So I guess, number one, we know from LITESPARK-022 that belzutifan added to pembro can improve outcomes, right? So I guess, first question would be, what is your expectation for belzutifan uptake in adjuvant? And then the second question is -- we've known that adjuvant pembro has been around and available and approved, I think, since '21. An IO/TKI is still considered to be standard of care frontline therapy following pembro in the adjuvant setting, right? Like, that's still the tailwind that you're benefiting from [ 9ER ]. So would it be a more, I guess, appropriate comparator arm for this trial? Should it have also included an IO/TKI arm as frontline?

Yes. So there's a couple of different questions there.

It's a very [indiscernible] question.

So one of the...

[indiscernible].

Yes, always. As a former sell-sider, I can certainly appreciate that. One of the dynamics that we've seen and I think the industry has seen is the challenge is kind of frankly, across tumor type of IO retreatment. And so understanding the actual contribution of that IO piece if you've received pembro in the adjuvant space is it's not really seen as kind of being -- contributing a lot. So we certainly feel that 033 is well designed to kind of answer that question of if IO retreatment's not really expected to be successful, what should that standard of care be? This is the first study to help answer that question. So we want, again, want to be involved in defining what that is. Again, this comes back to this dynamic around that TKI of the 2020s versus TKI of the 2030s. We see the kind of world of adjuvant checkpoint use only expanding in the future. One of the dynamics that we've seen is now that pembro does have a clear survival advantage, it's driving -- increasing and accelerated adoption. But beyond that, as the number of kind of de novo versus prior diagnosis kind of change as well, we expect that number to increase over time. And so it's this matter of if our view is that the number of patients who are treated with pembro in the adjuvant is expected to grow over time, we want to make sure that we're helping to define the answer on kind of what's that standard of care. As it relates to kind of the adjuvant [ belz/pem ] use, our observation has been that kind of that hockey stick acceleration of adoption for pembro in the adjuvant space was really when they saw a survival advantage. Because, again, these early-stage patients, as I mentioned before, it's always going to be about clinical benefit risk. And so when you're adding on an additional mechanism, you generally see additional tox in understanding, well, if I'm adding on additional tox, what is the actual benefit that I'm getting, especially with the 5-year survival rates of RCC continuing to improve, especially relative to, say, when cabo was first introduced, understanding the combination versus sequence dynamic is always going to be important. And so that's something to consider. Again, as we think about not how are patients treated today, but how will they be treated in 2030 and beyond.

Okay. STELLAR-311. So this trial is evaluating single-agent zanza versus ever everolimus as first TKI therapy and neuroendocrine tumors. I think unlike the other STELLAR trials, you're characterizing this as a Phase II/III. So is there a formal Phase II portion of this trial? And if so, what is it actually assessing?

So we haven't broken out kind of the specifics around what that Phase II/III gate is. But in general, like all Phase II/III studies, it's a way to be reasonably efficient from a timing perspective to understand, is the study likely to be successful? What are some of the metrics that we can take kind of an early look at and then continue to enroll? It's just a mechanism that we can use to really be efficient, both from a risk perspective, but importantly, from a time perspective, how to kind of get that completed as quickly as possible.

Okay. Maybe the last -- or maybe second to last zanza question. So STELLAR-316, right, this is the trial you referenced earlier, where you're looking at zanza and zanza plus PD-1 in early-stage colorectal. It's -- conceptually, I think it's a very interesting study. There's certainly a lot of observational data that supports -- supporting worse outcomes in these MRD-positive patients. But I guess when we've talked to KOLs about therapeutic intervention in these MRD-positive CRC patients who don't have evidence of clinical progression, the feedback there has been pretty mixed, and maybe that's just a byproduct of some of the historical data that's been generated in that setting to date. But how do you just think about kind of the mixed opinions that seem to be out there around slotting something like zanza into treatment here? And whether or not that could have an impact on the pace of enrollment and just how you think about the market opportunity?

Yes. I mean this is one that I'm particularly excited about. We're particularly excited about it. Again, this is a study where Exelixis is really kind of defining a new standard of care. When we've done kind of our pretty robust KOL work and really understanding that opportunity, I would frame it a little bit differently. If you look at the data that Natera has generated and published, those Kaplan-Meier curves for kind of that MRD-negative versus MRD-positive are striking. And kind of the current standard from the patient's perspective is watch and wait. These patients who are MRD-positive, ctDNA-positive kind of know that they're an exceptionally high risk of their disease recurring and can't really do anything about it. What we're offering is in potential intervention, where we know that zanza is active in a much later line population showing a survival advantage. So it's a sensitive tumor type, and then we're offering them either zanza monotherapy, or as we've announced, zanza with pembro. And so offering kind of that opportunity to really right shift again that DFS interval, which is unfortunately very short for these patients. That has a potential to be quite meaningful. I think you've heard Dana on the last couple of earnings calls talk about our view on potential enrollment being particularly brisk. That's, one, informed by this high unmet need from a patient's perspective around what do I do if I am kind of ctDNA-positive. But two, operationally, the benefits of working with Natera and working with Merck to kind of help identify those patients and really accelerate operationally what that could do. And so I'd actually frame it the other way. This is one where we're really excited to enroll. And as we reflect on kind of the 303 data, what we've observed is KOLs were really excited about 303 start learning about 316 and get even more excited about zanza as a franchise in CRC. And similarly, as those docs who are interested in kind of that 316 post definitive therapy CRC space, they start to learn about zanza in the later lines. And it's kind of that echo chamber that we're hearing a lot about that really kind of, I think, is getting us excited about what we can do as kind of having a real franchise in colorectal cancer.

Okay. Maybe last question here. So the pipeline behind zanza, you've got a few different modalities. There seems to be a bit of a build in concentration on biotherapeutics, right? I think you've got ADCs that now represent the majority of these earlier stage pipeline efforts. Should we expect to learn anything about kind of the 4 assets that are currently in Phase I development? Will we learn anything about those things before the end of this year or maybe first half of next?

Yes. I mean, I think the way that we view it, and again, as you can appreciate, you've known us for a long time, we're a little bit atypical for a biotech. One, we're profitable. Our focus, candidly, is on execution. And so we're not going to dribble 10 patients' data here and there to try and raise money or use the stock or anything like that. What we want to do is once we have mature stable data that reflects this is what the profile is versus what the profile could be, that's when we'll present it. I think what we've shown ourselves to be really able to do is to prioritize. And so if any of those early-stage programs aren't going to potentially define a new standard of care, we'll move on to the next thing. We're going to kill things quickly is kind of one of the mantras that we have here. And that's how we can be efficient is just this really stringent prioritization kind of across the board. And so whether that's moving to the next -- seeing us start a pivotal with one of those programs or seeing us cut it, that's the sort of thing that you'll see. And then we'll present when that data really is mature, and you guys can get an understanding of -- it's not just a couple of patients here or there. It's really a robust data set to help inform why we're making these investment decisions.

All's right. That's all we have for time. Andrew, I always appreciate it. Good to see you.

Good to see you.

And thanks for listening, everyone.