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

Welcome to day 2 of our Annual JMP Oncology -- Hematology and Oncology Conference. Thanks for joining us. My name is Silvan Tuerkcan. I'm a senior analyst here at JMP Citizens covering precision medicines. If you would like to ask a question anytime during today's discussion, please e-mail me or submit them via the chat. Today, right now, it's my pleasure to host CSO, Peter Lamb and Andrew Peters, Senior Vice President of Strategy of Exelixis. Thank you so much for joining us today.

Yes. Thanks for having us.

So obviously, regarding your cash flow generating commercial capital franchise, you're on track to hit the topline of your guidance. And you've -- I'm sure we've been talking a lot about the continuous growth of the commercial franchise. So today, I would like to focus very much on the pipeline beyond cabo. And starting here, maybe, could you please talk about your follow-on TKI-XL092? Is the goal here to eventually replace cabo or focus on opportunities where cabo is not yet used?

Yes. So thanks for the invitation, and thanks for having us. Maybe I'll kick this one off, and Peter can add as necessary. And before we continue, I just want to remind everyone on the line as we may be making forward-looking statements today, so please see our relevant disclosures in our SEC filings for risks associated with our business. So XL092, that's a program we're quite excited about. As you mentioned, it's sort of a follow-on opportunity from cabo, but really not designed or not really set out to replace cabo but more to kind of expand the opportunity set on where cabo could go. If you think about kind of what we started out at the onset to do with 092 was really improve upon kind of one of the kind of core challenges with cabo and that it has a long half life, about 4 days, and so 092 was designed to create a cabo-like molecule in terms of phenocopying the kinase inhibition profile of cabo with a much shorter half life. And so the data that we presented earlier this year, certainly, we think achieve that goal. The side-by-side kind of the early cabo experience versus the data that we showed at ESMO certainly would suggest that this is a cabo-like compound with a much shorter half life and then still early days, but also encouraged by the tolerability that we're seeing. So when we think about owning 2 in kind of development going forward, it's going to be about where are the opportunities, where are the indications, where a cabo-like compound could have activity. But importantly, how could a kind of more user-friendly, shorter half-life compound that's active in novel doublets, novel triplets be used. So that's kind of the broad strategy that we have going forward. Peter, would you add anything to that?

No, I'll just say, obviously, over the years, we have a broad set of knowledge and experience with cabozantinib, clinically. And it's a broadly active molecule versus a single agent, and it's turned out to be a pretty effective combination partner for IO-type approaches. Obviously, we've not explored every single one of those indications, combinations in registrational trials with cabozantinib. So the intent with XL092 is really to do that. And I think that's one of the things that with the STELLAR-303 trial that we started earlier in the year as a good example of that. As Andrew mentioned as well, it's a very vehicle for exploring novel additional combinations that we never explored with cabo.

Great. And I have a couple of questions about the ESMO STELLAR-001 data that you were referring to this year. First of all, how does the tolerability compared to cabo in the data that we've seen? And what's the importance of potentially lowering the diarrhea and PPE incidents? And my second question is on the efficacy side that we've seen. What about the fact that ORR was higher with the monotherapy versus the atezo combination. Is that just due to the a small number of patients with small follow-up? Or what can we tell about the efficacy side of things here?

Yes. So maybe I'll take those, and Peter can add on as well. I think just on the efficacy side, it's challenging, I'd say, to kind of compare across early days, early data, small end population, especially the monotherapy versus the combination. I think it's just still too early to draw any sort of meaningful combinations. Like I said before, kind of the takeaway that we had from the data, in particular, was this is a very cabo-like compound from an efficacy perspective. And then obviously, again, still early, but certainly encouraged by the kind of tolerability and early differences that we're seeing so far. I think that we're going to get a lot more information from the ongoing STELLAR-001 or 002 combinations kind of going forward to help further frame out the clinical profile, both from an efficacy and a safety perspective. But as it relates to the data at ESMO, certainly, we're encouraged that, again, it's a cabo-like efficacy compound with some encouraging but still very early potential differences on the safety side.

Great. And obviously, as you mentioned, you've initiated Phase III STELLAR-303 as you promised by year-end. Maybe moving on to your strategy after the pan TKI, so ex cabo and XL092. You've 2 -- around more than $ 2 billion in cash, and you've already a quite large pipeline for a $5.5 billion market cap company. Could you please walk us through your portfolio just at a high level and that spans anything from small molecules to ADCs and antibodies as well? And yes.

Yes. No, I'd be happy to. I mean if we go back a few years to when we were in a position of being able to build the pipeline beyond cabozantinib. We felt strongly that we will need to have both small molecules and biologics in the pipeline, just gives us access to a wider variety of target space, of therapeutic modalities. And obviously, if you look at the way kind of oncology development and obviously commercialization has gone, biologics probably represent somewhere over 50% at this point of the successful -- really successful drugs. So we definitely want it to be there. So then forward -- I mean you can expect to see us advancing both small molecules and biologics or doing partnerships around both those kind of modalities. I would say second, we're in a happy position, I think, from a science point of view that we are able to look fairly broadly at different mechanisms of action rather than having to constrain ourselves for example, I don't know to be a tumor metabolism company or [ exel ]. There won't necessarily be single themes from a mechanism point of view. We can go where we feel that the science is the strongest and clearly feel there's a reasonable development and commercialization path. And I think that really kind of also just helps amortize the risk of different mechanisms in different tumor types. So I think that's obviously sort of an exciting aspect to it, and you can sort of see it emerging. I mean I would actually argue that whilst I think we made a great start on the early pipeline. Our intent is to broaden it further, so I don't think we're done by any means. So you can expect to see us advancing additional molecules into the clinic and then hopefully potentially partnering with folks or licensing things to add to the early clinical pipeline. As it stands right now beyond XL092 that we just spoke of, we do have our first biologic XB002 or ADC, which I think we'll talk about in a moment, which is kind of maturing some says [ wrong ], it's maturing nicely. We will now have a couple of small molecules, XL102, our CDK7 inhibitor and newest into the clinic is XL114, which has a novel mechanism of actually kind of focused in the lymphoma space right now.

Great. Yes. Maybe it's a good segue, just jump right into the ADC side of your pipeline. So XB002, you got from iconic, use Zymeworks, ZymeLink, right, which has been somewhat derisked in several pipeline assets and other compounds. Could you please summarize the key takeaways from the first in human data that we saw recently?

Yes, absolutely. So we initially saw this opportunity, kind of triggered for us a few things. One was a pretty broad analysis of the whole ADC space and how comfortable do we feel getting into it. And it came out in the end for us as being certainly a big area of focus for us on the biologics side, so it's an area we're investing in fairly heavily and are going to be expanding. For XB002 itself, again, some of that is pretty compelling from a science point of view. We see it sort of intended as the next-gen version essentially of tisotumab vedotin actually which has approval already in cervical cancer. So from a target point of view, with the target being tissue factor here, there is already clinical proof of concept, right? It's a target that's known to be overexpressed certainly in cervical cancer, but actually a broad range of solid tumors beyond that as well. And [ 002 ] has several differentiating factors for tisotumab. Primarily, it's a different antibody, of course, and it binds to a different epitope on tissue factor. So tisotumab binds in a way that blocks the interaction of factor VII and thereby interferes with the coagulation cascade, and you can see that in the AE profile for that drug where there's reasonable rate of bleeding and epistaxis. [ 002 ] , the antibody doesn't but by nor it does not interfere with Factor VII binding. So we thought we'd have a differentiation based on not interfering with coagulation. Beyond that, you mentioned the ZymeLink technology from Zymeworks. This is a modified or a statin. So it's not the standard MMAE that has been extensively used clinically. It's certainly a modified version of that and a modified linker. And they had pretty compelling data showing that, that linker payload when conveniently attached to antibodies was highly stable. So that's a very desirable property in ADCs. You don't want the payload dropping off. You want to minimize the amount of free payload that you have circulating because that's certainly something that contributes to side effects. So that was a preclinical data. And I think what was just released a couple of months ago now, the Phase I data from the JEWEL trial, I think really starts to demonstrate that, that is playing out clinically. I would say the key takeaways are: one, we haven't seen bleeding so -- as expected. So that part of the differentiation so far looks very, very solid. I would tell you the second key takeaway really came from looking at the pharmacokinetics of 002. When we look at the intact ADC, the free payload. And I would say, I'm going to take one thing away from this, if you look at the 2 mg per kg dose, which is the dose of tisotumab that's used, at that same dose, we have twice the exposure of the intact ADC and 1/10th the exposure of free payload. So there's very low levels of free circulating payload. And although it's still relatively early, dose escalation probably starting to get up there. If you look at the emerging side effect profile, you can see at very low rates. In fact, we haven't seen any alopecia or diarrhea in the dataset that was presented, very little peripheral neuropathy. These are all polysome side effects for patients, obviously. So the emerging from the early [indiscernible] safety profile, I would say, looks very encouraging, especially given the amount of the exposure that we have to the intact ADC. And I would say, again, that's one thing that from an efficacy point of view makes us excited about what we're going to see going forward. We really think by being able to go forward with a higher exposure, we should see improved efficacy and not just in cervical cancer, but there are many opportunities in other solid tumors as well.

Great. And maybe talk about the other ADCs that you have -- that we've not heard very much about. You have 3 new ADC programs from Catalent's SmartTech platform. How do they fit in? And how did this -- I'm just going to lump that in with ADCs here, the recently licensed CBX-12 peptide-drug conjugate. How does that fit into -- what else you want to do?

Yes, in the last few years, we've been building a network of collaborations that we really intended to be sort of an ADC engine for us as well as building some internal expertise and capability, and Catalent was certainly part of that. We wanted access to a number of technologies and platforms in order to be able to advance kind of contemporary cutting-edge ADCs. So obviously, we have sources also good antibodies, which mainly comes from our collaboration within [ Benma ]. We've also in-licensed individual antibodies. We need to have access to a range of different payloads with different mechanisms of action come to us variously from collaborations with Catalent with NBE BI. And most recently, you may have seen we've done it, we did a deal with [ Briview ] to access their STING agonist payload as well so we can move into kind of immunostimulant break type of payloads, which I think is a very exciting area going forward. And then thirdly, getting access to the ability, 2 sites specifically conjugate payloads to antibodies, I think is important. One of the historic issues with ADCs has been they've been heterogeneous that's created some issues both on the manufacturing and can produce issues on the [ TORK ] side as well. Catalent. But Catalent have pretty much cutting edge, site-specific conjugation platforms enable you to precisely control the drug antibody ratio. So yes, we now have some experience with the Catalent platform. We advanced late last year, a new ADC into preclinical development. It's a known target, 5T4 and there have been attempts to develop ADCs against this particular antigen before. Again, it so widely expressed on multiple solid tumors. ADCs, I would call first gen ADCs and a number of them were drawn in the clinic for inadequate TI. But our belief is that by making or by constructing an ADC using contemporary technology, we'll be able to unlock the potential around that particular target. So that's XB010, which is kind of moving through our preclinical pipeline right now. We have several things kind of coming on behind that and not quite at the preclinical stage yet but will be in the not-too-distant future. As you commented, we recently kind of expanded our relationship with Catalent to add 3 additional ADCs against 3 targets. We have been disclosed the targets yet. But in each case, it relies on that same technology that I was referring to earlier. So this is ADCs, absolutely going to be a point of focus for us going forward. And our intent is to continue to broaden access to payloads, antibodies, technologies as necessary and have a real suite of ADCs moving forward in the clinic. Just to finish up on the Cybrexa deal. We have been taking the approach of looking at a lot of late preclinical, early clinical assets with the aim of doing deals where we could provide some funding to complete the sets of clinical data and then essentially license assets. At that point, if we like the data, Cybrexa was one of those deals. We also did a deal with Sairopa at the same time that fits into that category. Yes, this does broadly fit into sort of the ADC except it's a peptide-drug conjugate, not an antibody for a conjugate. And conceptually, what was compelling was with this particular approach in technology, you don't need an antigen overexpressed on tumor cells. So CBX doesn't bind to a particular protein on tumor cells. Instead, it detects the lower PH that's present in the tumor microenvironment. So tumor throw out a lot of acid, lactic acid. Peptide kind of detects that and forms a structure that enables it to translocate through the tumor cell membrane, dragging the payload, which in this case is exatecan with it. So again, the potential here is really broad. Most tumors have an acidic microenvironment. So Cybrexa recently put out some early Phase I clinical data exploring a number of different dose schedules and doses. Early signs of activity, certainly in CR, 1 PR, 1 near PR. I think most encouragingly, the PR was actually at 20 mg per kilogram once weekly dose, which I think will be a good schedule to show it. So that's -- they're continuing to expand that particular dosing. So early days for that, but I think encouraging today.

And maybe in the last minute that we have here together, could you please talk about the switching small molecules because of some internal breast cancer conference? Could you talk about the CK7 program with XL102? And just tell us about, at least from the data in the abstract, what we can -- if you could just briefly summarize the data that we'll see and why.

Yes. So it's the CK7 obviously, we think is an interesting target. It sits upstream with a number of other CDKs, so with activates CDK [indiscernible] and also active CDK4/6, which we are probably the most familiar ones from a drive point of view. It also regulates the transcription of kind of a number of nuclear hormone receptors, including estrogen receptor and androgen receptor. So again, the potential for CDK7 inhibitor is broad, both with a single agent, but also in combination with things like abiraterone, fulvestrant, [indiscernible] for example. In terms of the data, I mean this is an ongoing trial. So it's the first look at the ongoing Phase I, so it is relatively preliminary. We're presenting data on. I think there's 4 cohorts of patients dosed on a daily regime. We've got a cohort in there that's being dosed twice daily. I will say that this compound is one orally bioavailable. It's highly selective. It's potent. It's also covalent, so when it binds to CDK7. It does so covenant in terms of stick on the protein for as long as the protein's around. So our plan was to have a compound that would clear out of the circulation fairly rapidly but would have a long duration of pharmacodynamic activity in the CDK7 for a long while as one strategy one can use to try and optimize the therapeutic index by just not having a lot of circulating drug that could then have off-target activities. So in the data, you're going to see some PK data, which I think the paradigm in the halfway between 5 and 8 hours for that compound. We're also going to show some initial pharmacodynamic data that do show sustained inhibition of CDK7. So the innovation of the binding to CDK7 lasts longer than the compound is in circulation just as we had hoped. And I don't even though we're still in the dose escalation phase somewhat. You will see that we have pretty high levels of CDK7 target occupancy at the doses we're showing. So that's got a brief break of the abstract at least.

Great. Thanks, and we'll learn more tomorrow. With that, I'd like to wrap up. Thank you so much, Andrew and Peter. Thanks so much for joining me.

Thanks so much.

Thank you.