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

Good afternoon. Welcome to Barclays Global Healthcare Conference in Miami. Just e-mail me if you have any questions that need addressed. My name is Peter Lawson. I'm one of the SMID-cap biotech analyst at Barclays. I'm really delighted to have on stage with us the management team from Exelixis, Peter Lamb, Chief Scientific Officer; and Andrew Peters, Chief of Strategy. So thank you so much today for your time.

And I guess maybe the first question just around the cabozantinib franchise. It's been a fantastic growth engine cash generator for the company, just like -- kind of the puts and pulls -- pushes and pulls we should think about for '23 and the priorities around cabozantinib and then we kind of -- so I'll ask other questions around the deeper in the franchise.

Yes, happy to. And again, thank you for having us. And I think there's most number of Peters on a stage we all have.

Never have too many.

Yes. Before I begin, just a reminder to everyone, we may be making some forward-looking statements today. So please see our relevant disclosure in our SEC filings. So for 2023, as we talked about on our fourth quarter call when we gave guidance, we do expect to see continued cabozantinib growth, primarily driven kind of by 2 functions and a continuation of both from the success that we've seen in the past. First is really around continued market share gains and the strength of the 9ER data, as well as kind of that longer duration, that stacking of patients as patients are living longer, staying on drug longer, you're seeing kind of that continued build and growth over time. I'd point to the recent data, the 44-month update that we presented at the ASCO GU Conference, really a well-received update and is a chance for our sales force, our commercial organization to continue to kind of reach out and touch base with physicians, patients alike, and really kind of emphasize that balance of data that we really think has driven a lot of the adoption so far. That's a combination of the efficacy, the tolerability and importantly, key things like secondary influence the quality of life that really have resonated with patients and physicians alike. So it's a continuation of that, so that market share gain, longer duration, continued traction in the community.

What's the trigger as we kind of think out beyond '23 to continue to drive that growth around cabozantinib, puts and takes around...

Yes. So without kind of getting into kind of long-term guidance certainly, we'd expect a continuation of the exact dynamics that I talked about. And then beyond that, we'll see how the cabo franchise continues to develop from a clinical trial perspective. We've talked about our expectation for more mature survival data from the 313 study that's the cabo/ipi/nivo study expected later this year. As a reminder, we did see positive top line data last year that was presented at a plenary presentation at the ESMO Conference, which showed positive data on our primary endpoint of PFS after discussing with regulators, we wanted to see a more mature cut of that data. And so that's expected later this year. So that's certainly a dynamic that we could look forward to. Similarly, CONTACT-02, that's the cabo plus atezo versus a second NHT and NHT refractory prostate cancer. So that's another one that we're expecting data later this year as well. So that can certainly influence the opportunity. And then we're planning to present data from the CONTACT-03, which recently read out. But without kind of getting ahead of that data into the specifics, certainly, we want to look towards how cabo performed, in particular cabo monotherapy to see if that has the potential to influence things like duration and market share in that second-line setting going forward. So kind of a combination of all of those.

And just remind us where we stand for like IP and kind of how that then marries into how you think about investing in cabo versus the pipeline, and particularly...

Yes. So kind of to take that second part. So the CONTACT studies, those are kind of the final cabo studies, so to speak that we're going to be running with kind of a focus on the rest of our pipeline and portfolio going forward. From an intellectual property perspective, we really see kind of 2-key buckets or 2-key updates with respect to kind of the ANDA challenge or MSN. So MSN-1 was a trial last May that the decision from the judge came out earlier this year and upheld kind of the core composition of matter for cabo to 2026. And then issue in MSN-2 is a broader set of polymorph claims as well as some high-purity cabo claims around 2030 or so. So as we think about cabo and kind of the Exelixis story going forward, we model 2030 for all of our modeling, and that just really reflects kind of our continued belief in the strength of the cabo franchise. But taking a step back, I think you may have framed it well and that we view cabo as kind of the gas of the innovation -- Exelixis innovation engine going forward. And so that's kind of how we see the success of cabo. The continued financial success of cabo really is a way to invest in the future and really build towards being a multiproduct oncology company. I'm sure we'll get to it later as it relates to kind of a pipeline. But we're really excited about what we think are differentiated and relatively lower risk value propositions, whether at zanza building upon all of the foundation from cabo or XB002 building on Seagen's TV. So that's -- it's using that success of cabo to kind of invest in the future.

Got it. Thank you. And then just, I guess, on zanza kind of what -- now we should really be thinking about that, filling in all the gaps that cabozantinib kind of left and newer gaps that may be emerging greenfield opportunities?

Yes, maybe I'll take that one. So yes, when we started the program that led to cabozantinib, obviously, we were sitting on a very significant amount of cabozantinib clinical data. We knew at that time, cabozantinib have been explored fairly broadly as a single agent. We've seen evidence of activity, and I think it was 20 different tumor types. We also knew it was a very effective partner for immune checkpoint inhibitors because of the spectrum of targets that inhibits. Cabozantinib on its own has some pretty interesting activity on both adaptive and innate immune cells in the tumor microenvironment. And obviously, that then played out in the cabo/nivo 9ER study very, very nicely. So yes, we really felt we had a kind of target profile in cabozantinib that was special. And there's a core set of targets that we believe are very much involved in that activity. And in developing benzolindinib, we were looking to very much keep those targets in there in about the same ratios as they are with cabozantinib. So it really enabled us to exploit the clinical knowledge that we had from the years of developing cabozantinib. So that's what zanzolintunib is, has the same target profile as cabo. What we did do was to fine-tune the pharmacokinetics. So cabozantinib has about 100-hour half-life in people. And that has a couple of consequences. When patients initially come on, we certainly spend some time educating physicians to stay on top of any emerging AEs because during the first 3 weeks of dosing, cabozantinib accumulates. So if you start having an AE during that period, you really want to dose reduce or whole dosing pretty rapidly. Secondly, of course, if you have an AE during the course of treatment and you need to dose reduce or hold, there's a washout period of a week or so for cabozantinib. And then that's some inertia there essentially in terms of getting the patient back on to cabozantinib, does it give them an opportunity to switch to a different TKI. So we felt having a version which had a shorter half-life would really enable physicians to enable adverse events much more crisply. So that's what zanza is, again about a 24-hour half-life. And then in terms of the development path of zanzalintinib, it's very much along the lines of what you're outlining. I mean, we're really looking to fill in the white space where we have not run registrational trials for cabozantinib. So it's not a straight drop in replacement, it's really looking at areas where we saw some evidence of activity with cabozantinib, but never put it into a registrational trial. So if you look at the first 2 Phase IIIs that we initiated last year, with zanza with STELLAR-303 in colorectal carcinoma combination with atezo and the comparison there is regorafenib, that was really based on a CRC cohort from COSMIC-021 with cabo/atezo, for example and also some data from cabo and Durva from the CAMILLA trial in MSS stable CRC that looked quite promising. And then likewise STELLAR-304 which got kicked off very end of last year, that's in non-clear renal cell carcinoma. And actually, there is no drug that's registered. There's no registered therapy for it. I mean, people certainly use cabozantinib and they use other TKIs as well in that setting. So we know cabozantinib is active there, obviously, strongly we believe that zanza is going to be active as well. So we're looking -- doing a zanza/nivo versus sunitnib type of comparison, a sort of low-hanging fruit way of starting to filling those gaps. And then I would say, finally, the other big opportunity for zanza is to really go into novel or contemporary combinations that we haven't done historically with cabo. So starting to have one, yes, we've got cabo/nivo and we obviously got a little bit of experience with zanza and nivo and that's actually ongoing. But then what novel MOA do you add on to that in RCC, for example? So we certainly also have ongoing in collaboration with BMS combinations with LX, we have people on relatlimab just as an example. So looking to get into novel combinations that will really move kind of the needle clinically when combined with zanza and that would be novel rather than just trying to replicate the exact things that were doing with cabo.

Got you. And then what would be the first pivotal data set we should expect?

Well, it will -- for zanza, that will be the STELLAR-303 data set. So it will be the CRC data set. The primary endpoint there is overall survival.

And the timing on that?

TBD, we haven't kind of guided on the exact timing, yes.

Okay. Where are you from, like enrollment and...

Well, enrollment is ongoing. I mean, we don't -- I can't really say much more than that. We generally don't give precise updates on where we're standing with respect to enrollment.

But I guess, once it's complete, that's when we get an idea that -- you mentioned it's complete and then from there, we can kind of work out when we could potentially see OS data.

Yes, that's generally what we do. Yes.

And then the contemporary combination. So I mean, I kind of read into that, we're going to see a lot more of these combination strategies develop pretty quickly for zanza?

Yes. So I mean, if you look at some of the tumor types here and there to our -- RCC is obviously a great example of that. I mean, we're now in the era of looking at triple therapies, right? And again -- so doing the cabo/nivo/ipi kind of combination is where it's at. Registered staff is already doubled. So looking ahead, we're going to be looking at more -- probably more triplets in RCC and then even adding on from them. So it's a matter of what are the new or emerging mechanisms of action in RCC or any tumor type where we have some belief that mechanistically, it would be a good combination with zanza, and then we'll go and get some initial clinical experience and hopefully advance it from there. But yes, I think that's going to be the future in these kinds of tumor types for sure.

Should we think of zanza just replacing cabo eventually in RCC and the various combinations that cabo has in RCC?

Well, I mean, ultimately, what you're looking to do again is to provide a clinical advantage in whatever setting you're in, right? So provide additional clinical benefit over what's already there. So in the cabo sense or the cabo/nivo sense, with the zanza plus, whatever the combination ends up being, the test is -- does it move the needle clinically ultimately? If it does, then yes, it will take that. If it doesn't, the cabo will stay where it is. So that's usually -- we're not going to do. For example, we're going to do zanza/nivo versus cabo/nivo, that doesn't make any sense because we're not really expecting that to provide a substantial benefit in terms of PFS or OS. So I don't think that would be good for patients. But zanza/nivo plus something else would be a very kind of rational way to approach it.

Got you. Okay. So it's not IP replacement for where cabo is now post 2030...

Clinically meaningful replacements...

Yes, clinically meaningful replacements. So you could potentially stack another drug in there and then that could potentially replace that drug...

Yes, yes, exactly. And then if it's clinically meaningful, of course, along with that potentially comes longer duration of therapy and the...

Got you. Have you already kind of talked through this year for additional registration or status for zanza?

We -- I think what we said publicly is it's our intention to start additional studies and hopefully multiple additional registrational studies for zanza. We haven't specifically laid out in what tumor types or patient populations those are going to be in, but they will fall under the general parameters that I just discussed in terms of how we're looking at zanza development.

Got you. Okay. I'd love to move on to your tissue factor ADC. I guess, particularly in light of the Seattle acquisition, I guess, where you could differentiate versus Seattle's molecule?

Yes. So very excited about XB002, it's the first biologic in our pipeline and obviously, the first ADC. But I took out of the acquisition and said ADC is a good space to be in, but that's something that -- it's an area we've been enthusiastic for a few years now and obviously done a number of deals in to give ourselves access to a variety of antibodies, payloads and conjugation technologies. XB002 was a tissue factor targeting ADC that we in-licensed from a company called iconic. And we thought the rationale and the story there was really compelling. As we commented, obviously, TIVDAK is out there, it's already been approved in cervical carcinoma. So there's a sense in which the target has been somewhat derisked, right, targeting tissue factor for an ADC has been effective. I also knew a lot about the clinical profile of TIVDAK and the spectrum of adverse events that it has. So as we were looking at the 002 opportunity, we could see a number of areas of pretty crisp differentiation. The first really relates to the antibody. So just to go back to TIVDAK , that antibody binds the tissue factor and it does so in a way that blocks binding Factor VII. And it's that binding that initiates the coagulation cascade. So TIVDAK interferes with coagulation. And you can see in that data and in that label, fairly high rates of bleeding in patients because of that. 002 antibody binds to a different site on tissue factor. It does not interfere with Factor VII binding. And -- so preclinically, you see no interference with coagulation whatsoever. And then gratifyingly in the initial clinical data that we put out late last year in that kind of Phase I patients, we've not seen any bleeding at all to date. So that seems to be transferring pretty well to the patient population. I think the second main point of differentiation is really around the link of payload. Again, in TIVDAK, it's the classic [indiscernible] auristatin MMAE type payload, the microtubule disruptor. What we've used in 002 has been technology that came from Zymeworks called their ZymeLink linker-payload. It's still a microchip disruptor and it's still in the auristatin class, but it's fairly heavily modified. And the net result of that, those modifications, I think has been an antibody with different properties. And in particular, it's a very stable antibody in terms of the payload staying on with the antibody and not falling off much in circulation. I think what got us really excited from the data set that came out last year with some of the PK data, particularly, we got patients at the 2 mg/kg dose cohort. 2 mg/kg is the dose that is approved for TIVDAK. When you look at the pharmacokinetics 002 at that dose, what you see is the exposure to the intact ADC is twice that of TIVDAK at the same dose. While at the same time, if you look at free circulating payload, the free circling payload is 1/10 that you see with TIVDAK. So we think we've got a much more stable ADC. We think it's going to be important for 2 reasons. First, just from our side effect point of view, our view is that some of the AEs and side effects do come from free circulating payload. And some of them are like classic TECAN side effects, like peripheral neuropathy, alopecia, for example which are certainly present with TIVDAK. We've seen no or very limited cases of those so far. So that could be one potential point of differentiation on the AE side as well. I think if you go to the efficacy side of the equation, the PK is giving us some confidence that we're going to be able to see efficacy outside of cervical carcinoma. The original TIVDAK Phase I, they did look at cohorts of non-small cell lung cancer patients, urothelial carcinoma patients. And we did see some activity, but it wasn't really compelling. And so it really developed as a single agent in those. But we think with the added exposure that we're getting at a tolerable dose gives us the opportunity to build on that initial efficacy signal and then -- potentially then to see efficacy in a wide variety of solid tumor types. So the plan this year with 002 is very much to push it forward. We've got to get to a recommended Phase II dose. We're then looking at opening up to 10 different expansion cohorts in a variety of solid tumors. And then we have started combination studies, both with a checkpoint inhibitor, which is a very interesting area for combining with ADCs generally. And we've also started a combination with bevacizumab. And actually, that was really enabled by the fact that we don't have bleeding as a side effect. It would have been tricky to do if we had significant bleeding. So we're excited about that. So stay tuned.

How should we think about the, I guess the initial pivotal studies or where could it be approved first? Is it kind of -- think about it as TIVDAK plus kind of thing, so it's cervical cancer? Or is it elsewhere that we should be thinking? So these are the 10 different tumor types you were kind of talking through...

Yes. I think it will be data driven. So we're not going to stage how we enroll in the various different expansion cohort tumor types. So we'll certainly enroll. One of them certainly will be cervical carcinoma, so we can sort of benchmark if you like. But beyond that, it will be on the basis of -- where do we see good efficacy and tolerability, what are the competitive dynamics in that particular patient population. So that's a decision we'll make at the time and will be informed by the data that we have. We haven't pre-decided yet.

And what would cervical cancer trial look like? Would that be kind of head-to-head eventually?

It's -- again, it's -- and I mean, to be clear, again, we're not really seeing this as just a cervical cancer drug. That would be a relatively modest kind of opportunity. So I think we're more excited about the potential beyond cervical carcinoma. But again, exactly how the trial will be designed really depends upon the data that we have at the time.

Got you.

I think, Peter, kind of the key point again to what I was referencing earlier is we're building upon a foundation of kind of this validated target with this differentiated technology, differentiated antibody different to the link of warhead. So we're kind of building upon and expanding the opportunity set as opposed to just trying to iterate on just the cervical opportunity as an example. So that's why we're so excited and enthusiastic about the program. As Peter mentioned, pharmacology 101 twice the exposure of the 1/10 the amount of free drug is a really good place to start. And that's why we're excited to move into expansion cohorts and kind of build from there.

Got you. And so it's that early data PK/PD that kind of gives you that confidence that this is a potentially broader...

Yes, absolutely. Absolutely. And like I said, there were [indiscernible] activity with TIVDAK recently and we think -- well, if they could have dosed higher, the activity should get better. And that's essentially what we're going to be able to do. We're going to give patients more exposure to drug at a level that's still tolerated, and then there's more opportunity for clinical benefit. And yes, I think if you think broadly and you think about what we're doing in ADCs broadly, we think there's a fair amount of opportunity there for coming in with next-generation approaches to known ADC targets. The ones which historically, people have made ADCs have bounced them into the clinic, but they never got out of Phase I because the therapeutic index just wasn't there. I mean, there may have been activity, but there was just too much toxicity. So another program we have in preclinical development right now is a next gen 5T4 ADC, for example, spills upon exactly that same premise. We're now making -- using modern site-specific conjugation technology using the kind of latest knowledge with respect to warheads and payloads and the like. So it's one example. I've no doubt we'll be doing more in the future as well.

Got you. And so the problem with that therapeutic window for many of these ADCs is just the toxin just ends up diffusing off?

Yes, it's been twofold and this is something that's sort of bedeviled to feel for a long time. I mean, the idea of ADCs goes back, what, to 25, 30 years. And there are a couple of initial ADCs that got registered like Mylotarg, Kadcyla. But after that, there was just a valley of death basically where nothing got registered. And it was mainly the therapeutic index issue. It wasn't necessarily lack of efficacy. It was just too toxic. And there are a number of things that play into ADC toxicity, but one is how much circulating free payload do you have. Others can be nonspecific uptake of ADCs, which is much better understood now. And then the third is, of course, if you have residual target expression on normal tissues that can play in as well. But a lot has been learned in the last 20 years about what drives therapeutic index and what drives the kind of toxicities of ADCs. And I think that's -- you can see that in the renaissance in the field, right? We've got some dozen or so ADCs have been approved in the last 3 years. And truly built on that and improve manufacturing processes.

Got you. Is the industry getting better at kind of threading that needle between finding the right target that's not widely expressed and having a tighter bind of the toxin, et cetera?

Yes, I would say, absolutely. I think it's evident what I just said in this kind of burst of approval. So this thing has a great time to get into a space, right? Sometimes it is first, you're faced with having to slog through a lot of the problems and kind of issues with a technology platform or a therapeutic modality. But yes, we're building now on 20 years' worth of experience in ADCs and a lot of modern technology. So again, if you look at the deals we've done in the space, with Catalent with NBE, with Redwood, with Invenra. I mean, our aim really is to assemble panels of well-characterized antibodies, have access to kind of contemporary site-specific conjugation technology so that we can control the final product to have access to a suite of different payloads and linkers that we can really assemble the right ADC for the right tumor type and for the right patient population. And we think there's a lot of opportunity there.

And you don't think it's one particular part of that equation that drives it, whether it's the site-specific conjugation, is that...

No, I think the integration of everything. It's -- that's why ADC historically has been hard. You have to get everything right to get -- to be effective. The upside is when an ADC is effective, you know pretty early on. I mean, you can see in Phase I if you're getting activity, you've got a reasonable handle on what your talks is. So it's not normally a guessing game with ADCs, which is kind of attractive from a portfolio management point of view. You kind of get early data and then make a go or no-go fairly crisply.

Perfect. Thanks for allowing me getting some extra time.

Thank you.

Thank you.

Thank you.