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

Well, good afternoon, everybody, and thank you once again for joining us for the 42nd Annual Cowen Healthcare Conference. I'm Yaron Werber, Biotech analyst and it's a great pleasure to moderate the next fireside chat with Mike Morrissey, President and CEO of Exelixis. Mike, as always, great to see you. Thanks for joining.

Yes, Yaron, great to be here. Thanks for the invite again, for sure.

Last time, virtually, not going to wave.

Yes, last time, I agree, I agree.

No excuses.

Before we start, I'll just remind everybody that 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.

And if anybody has questions, feel free to e-mail them directly to me or use the Wall Street Webcasting portal and I can actually see the questions. So Mike, lots to talk about. This is going to be a big year for Exelixis, execution-wise. Obviously, cabo patent wise, and also data-wise. So let's maybe just start commercially with cabo. We've seen cabo continue to gain share. KEYTRUDA's approval from the CLEAR study hasn't made a huge dent. You've continued to grow share and ultimately, NRx continued to grow. The market has been a little lumpy during COVID, expanding. It's been a little -- the IQVIA data has not been very predictive. Why do you think that is? And what do you continue to see commercially?

Yes. So I think that's always been the case, though. I think the -- that snapshot that IMS or IQVIA generates is based on some data and a lot of modeling. They look predominantly at the specialty pharmacy network, which is a fraction of what kind of our book of business. So it's just an estimate. And sometimes just directional and then just not. So it's hard to say. Look, we've been pretty pleased with the overall performance. NRx wise, we grew 40% year-over-year in '21. TRx was a 50% growth. Obviously, the impact on revenues were hard to miss, but it's great to have the data come together so well and to be able to educate physicians and health care providers to be able to see the benefits of the combo cabo/nivo regimen in frontline -- in the frontline setting for RCC. So we're pleased with that. We're in the business of really focusing on only one thing, and that is to improve standard of care for patients. And as you would imagine, it's incredibly rewarding to see the impact of the 9ER study on the frontline setting. But it's just one study, and we have a lot more to do and a lot more going on this year. So we're happy, but not satisfied, just keep cranking.

Okay. Let's move to 313, that's the cabo, Opdivo, Yervoy triple and first-line study has been fully enrolled. Data is expected in first half, I'm assuming it's Q2. You enrolled 840 patients When we looked at Checkpoint or CheckMate 214, there was about 1,200 patients in the head-to-head ipi/nivo study. Why is 840 the right number?

I guess 840 is the right number. Well, I think that's the powering assumptions that go into the sizing, obviously, right? So I think if you look at the 214 study, there were a lot of uncertainties there relative to when that study was started versus what they found at the end. And as you'll recall, they actually won on survival, not on PFS. So that whole framework was really groundbreaking research at the time and certainly changed standard of care, again, for front-line RCC patients in a pretty dramatic fashion, which has been then further complemented by the TKI/IO combinations like cabo/nivo. So again, the sizing is all about the powering assumptions for PFS as well as OS. And again, it's been fully enrolled for a while. So the [ chasm ] truth will be known soon enough. Again, we're guiding for the first half of '22. And we'll get the data, and I mean that's what we do. We're in the p-value business, and we're going to have one of those coming up soon for 313.

Yes. And can you remind us, at least what you've said publicly, I mean, the primary, as you said, is PFS by IRC. What -- how did you power the study? And what is the comp for PFS? And maybe to remind just the audience on 214, the doublet arm against sunitinib did 11.6 months on PFS.

So we have not talked about the assumptions and the stats around the trial. We never do that. We talked about the sizing and the inclusion, exclusion criteria, those kinds of things. But we don't go into detail about that until after the fact. So stay tuned there. I mean you quoted all the right numbers for 214. It's a reasonable place to look in terms of the control arm. The question is how much that standard of care evolved since then, and that's why you do a blinded randomized global trial is to get the most contemporaneous data for the external arm -- [ external metal arm ] versus the control arm. So we're doing that now, and we'll know soon enough. So stay tuned.

Do you have enough powering for the secondary endpoint of survival too? And I assume we're not going to get mature survival data right now.

Well, as you recall, we added a couple of hundred patients to the trial kind of during -- while we're in progress largely because BMS finally hit the medians on the 214 survival data. So again, in terms of the estimates that we had based upon their immature survival data became much firmer when the medians were known and based on the 4- or 5-year follow-up. So yes, do we have enough powering? I think so, but that's just -- it's just part of the overall hypothesis and we've used the most contemporaneous data from 214 to be able to resize as appropriate. So certainly, survival is a big part of the process here. Certainly, that was a big win for us with METEOR as well as in 9ER. So we'd like to see that here, again, how that separates from the ipi/nivo curve. If it does, when it does? All the details will be defined, but we'll look at the first interim when we look at PFS and go from there.

Did the Apollo study, that was at ASCO GU last year. That was the triple Phase I. That showed a 57% response rate for the triplet in about 9 patients or so. I think there was about 13 patients who had different permutation of the doublets or triplets. Did they ever release the PFS data from that study? Again, the response rate was 57% and to compare it for the audience CheckMate 214 had a 42% response rates for ipi/nivo. Ipo...

Yes. So all that data has been presented and published. I don't have it on my fingertips. I would caution everybody to make any kind of crosstalk comparisons between that initial Phase Ib data in 214 and whatever else. I mean you're looking at trials done at different times by different investigators, be aware. Those are always tough ones. And people always do it, I do it, you do it, but I have to do it with your eyes wide open. So...

And the tolerability again in that Apollo, I mean that's the only one we've actually seen a triple for. The tolerability looked like it wasn't adding a lot of talks. I mean, again, you're only getting 4 cycles of ipi. Any thoughts there?

Nothing that I would want to really opine upon now. I think what's always the case is you have to look at the totality of data at the end of the day, right? So it's the efficacy, safety, tolerability, all those things combined are important to look at in the context of any pivotal trial that you run on a global level. So again, I would just highlight for everybody to look at all the data when it's available and in its totality and people can draw their own conclusions based upon that trial.

And in your view, CheckMate 214, that was intermediate and poor risk as well. That's a good comp, right, for the study in terms of once the data is out.

Well, again, 214 was in all risk types, right? As you'll recall, the favorable risk population actually did worse on survival, at least at the primary endpoint at the primary look. So the label is in poor and intermediate risk. But the trial, the 214 trial with that all risk types. Just to be clear about that.

Yes. Okay. Got it. So we're recording a 42% response rate that was across the entire study.

I don't know what you're quoting if that's the overall ITT population or the poor intermediate risk, again, I don't have those numbers in front of me.

I think that's actually the intermediate and poor risk population. Yes, I think we split it out. Okay. Okay. Let's move on and -- now move on to CONTACT-01 and CONTACT-03. For the audience, as you know, cabo and TECENTRIQ are in 2, which actually 3 different randomized studies. CONTACT-01 is in second-line lung cancer. CONTACT-03 is in second-line RCC. CONTACT-01 is TECENTRIQ Cabo head-to-head against docetaxel. And in second-line RCC TECENTRIQ Cabo versus Cabo. It's essentially both studies require failing first-line therapy, which presumably is a checkpoint inhibitor included.

Exactly.

Just remind us, for CONTACT-03, do you need to be naive to cabo? Or do you need to be naive to a TKI or no?

You need to be cabo naive to get into the trial, obviously, yes. But other TKIs in a frontline setting are permissible, yes.

Okay. And in CONTACT-01, it sounds like you're planning on having interim OS whereas in CONTACT-03, it will be mostly PFS data. Is that correct?

Yes.

Okay. What are you -- as you're thinking about -- and maybe let's double-click first on lung cancer. Cabo showed a 14% response rate, mono. And Cabo TECENTRIQ had a 27% response rate. PFS was 4 months mono and 4.2 months in the combo in COSMIC-021. The mono was in the E1512 study. As you think about sort of powering for CONTACT-03, is it powered for survival? Is it powered for PFS? And kind of, again, how did you think about what to expect?

Yes. So yes, again, for CONTACT-01, overall survival is the primary endpoint. Everything else is certainly nice to have, want to have that to collect all that data, but it's -- all the powering is done, and all the sizing was done for OS, absolutely, yes.

And can you give us a sense what is clinically meaningful? Or what are you looking for?

Yes. We haven't -- again, we haven't shared the details of the [ SAP ] broadly publicly at all. So I don't want to do that now. Again, we'll highlight all that when the top line data is available.

And maybe the same question for CONTACT-03, the primary presumably is PFS?

It is PFS. Yes. I guess that one I believe we have primaries for both PFS and OS. But again, either one of those can lead to a successful study as well as, certainly, again, if we saw a PFS with no decrement in survival, that would be certainly perceived to be beneficial for patients based upon past precedents for RCC.

So the primary there is sort of like DTC, it's going to be PFS and/or survival, but they're not co-primaries. You can hit one or the other. You don't have both?

Exactly. Yes, yes, exactly.

And that's prospectively defined with FDA?

Yes.

Okay. And second there is, I assume, more response rates.

Yes, you collect all the data, all the PFS data is by blinded independent review, et cetera. So yes.

And where would you think -- I mean, in RCC, it's pretty clear, patients get cabo in first-line or in second-line. Just remind us what percentage of uses is first-line right now versus second-line with cabo?

Yes. It's approximately in the mid-20s for first-line in terms of market share. And second-line, it's -- gosh, it's the market leader, 40% plus or minus overall may be higher. So again, we like to be able to frame that opportunity as we -- based upon the totality of data from METEOR, CABOSUN, 9ER. Patients see cabo at some point in time in their journey between first-line and second one. It's a mainstay. It's very gratifying for us to be able to see really no erosion in the second-line share as we grew the first-line share over 2021. So -- and the question is what do we need to do to continue to be competitive in the space? And again, raise the bar for standard of care for patients with RCC, whether it'd be 313, whether it'd be CONTACT-03, we are committed with a long-term commitment to patients with kidney cancer. And I think just showing the cadence of trials here with cabo and potentially 092 going forward really reinforces that. We're very, very serious about helping patients with kidney cancer. It's a mainstay of our business, for sure.

And for CONTACT-01, if that regimen is positive, again, head-to-head against versus docetaxel, where do you think that might find prominence in...

Well, that's -- I mean that's a big opportunity, right? If you consider that most patients get IO upfront with lung cancer, either IO, either ICI by itself or some combination with either a second ICI as in the case with ipi/nivo or with, say, chemo, whether it's chemo combination or chemo sequentially, it's a population that is very well served by IO therapies, right? The question is what are those patients do when they progress, and the vast majority will progress. There's not a lot there for them. So there's certainly a very large unmet medical need for patients who are progressing either resistant to or refractory to frontline checkpoint inhibitors. And we think this has got based upon the 021 data that's been published and more that we'll talk about later in the year. We think we've got a really good opportunity here. But we got to run the pivotal trial and get the data and with the p-value, right? So that's in process. And again, we hope to have top line data in the second half of '22 here.

Right. Let's move to 092 and STELLAR-303. And so that development now in third-line plus, you're targeting RAS wild-type head-to-head against regorafenib, regorafenib [indiscernible] at 2% response rate, it's about -- just about 2 months PFS in the wild-type overall, just in the all-comer CRC population. What percentage of patients are RAS wild-type in CRC?

Yes. By our math and by our market research is between 50% and 60% of patients are RAS wild-type in the CRC population. So it's a very sizable population. It's one that we think we can based upon the data we had at ASCO-GI earlier in the year potentially impact, and we're super excited to have 303 be the first pivotal trial that's earmarked for 092. So one on the way and many more to come for sure.

And so can you talk a little bit about STELLAR-303? I mean that's 092 with the centric at the head-to-head against regorafenib 600 patients. The primary is survival, and I guess, RAS wild type. What led you to be -- to that scheme? And maybe you can talk a little bit about what you saw in COSMIC-21 and in the CAMILLA study?

Yes. Well, it's -- certainly, it's -- 303 is informed by the data we had at ASCO-GI, both from the 021 cohort as well as the CAMILLA study, where we're seeing pretty significant response rates, 25% up to 50% with all the caveats of small patient numbers and everything else, but certainly compelling relative to what you would see with third-line standard of care. We looked at PFS with OS, again, even certainly be more careful with those data from single-agent non-randomized studies. But everything was moving in the right -- kind of pointing in the right direction. So the unmet medical need is large. The presumed activity, certainly with cabo, cabo ICIs are significant here. So it made a lot of sense to move forward there with 092 as the first one.

And so both of those cohorts, COSMIC-21 with TECENTRIQ and CAMILLA with Durva, both PD-L1s, each cohort was 12 patients. Was that -- it sounds like that was compelling enough to then move into a 600-patient study. What led you to be comfortable? Is it just -- what's the point in doing another Phase II that's going to take 1.5 years? You might just do the big Phase III.

Well, it's always -- we've been pretty aggressive over the years. We haven't done a lot of Phase IIs with cabo in general. Think about METEOR, think about CELESTIAL, even CABOSUN and 9ER. Small Phase IIs, which can be 100, 200-plus patients, you don't learn that much more from -- by adding an extra cohort of patients from doing the pivotal. So time is of the essence. These patients, especially with large unmet medical needs really require sub level of risk taking on our part in terms of the financial risk of going from Phase I to Phase III. But if you think about what we've done with MTC, with RCC, with HCC, DTC, that's, I think, worked out very well. So we've been driving that process forward. Obviously, we're very careful on the safety side to make sure that we're not putting patients at any undue risk from a safety point of view. And we've got, I think, really good characterization with both cabo and now 092 relative to the earlier works. So it's the best way to help patients and build the business. And I'll take a step back and just -- I think that's something that we've been very facile at doing, right, is going into pivotal trials. We're in the p-value game. All value for patients, all value for shareholders is really derived from that number, right? And I think a lot of companies are hesitant to do that to a certain degree, and we're kind of seeing the outcome of that relative to, I think, the way the market is trading right now, but we're not afraid of that. We revel in the opportunity to run the right experiment, put the money at risk and get the data we need to help patients. And that's the only way it's going to happen is by really these big trials. So I'm going off my [ ranch ] here but it's something that we just do, that's our business is to generate p-values, right?

And this year, you're planning on starting other Phase IIIs across other tumor indications. Maybe just remind us from STELLAR-001 and 002, you're going to have some updates, obviously, this year in a Phase Ib. Without trying to get ahead of disclosing what you're going to do in Phase III, but what comes to mind with 092? In the past, you've talked a lot of broad host of opportunities.

Yes. And that's still the case. And I don't want to front run. But we're thinking we have the second pivotal trial details more or less locked in right now, and we're working on the finalizing details for the third and fourth as well, doing those internal debates and studies between development and commercial. So we make the right selections. But again, we've done, I think, a really good job of learning how to use cabo and navigating cabo in some of the early tumor types that I mentioned this, right, thyroid cancer, renal cancer, liver cancer, lung cancer, prostate cancer. There's another whole host of opportunities we could do in terms of other novel doublets and triplets with Checkpoint inhibitors and other kinds of modalities that we think can drive value. So we're just really at the tip of the iceberg here. I don't want to get ahead of that. But I think the broad activity of cabo, we have [ bona fide ] RECIST 1.1 responses in 20-plus different tumor types. Broad activity that we've talked about previously. So the question is really now can we take all that learning, all that knowledge and all the momentum and then use it with what we think is a better cabo-like molecule 092 and be able to build the franchise even further to help patients and drive value for shareholders. That's the whole goal.

Right. Let's now shift over to 102. Mike, the CDK7, we're going to have initial Phase I data second half of this year at a medical conference. Can you talk about maybe #1, the therapeutic window with CDK7? And how the covalent versus noncovalent debate relative to your competitors, where you shake out on that side?

Yes. So again, this is -- I think this is a very unique target. It's certainly implicated in a variety of different tumor types, pathways, obviously, very well known and is one that has been, I think, studied more upstream than downstream in terms of the CDK4/6s. But it's similar from a basic biology point of view as any pathway you want to look at. It sounds like a related. So potential tops profiling might be different than what you've seen previously with others. But again, there's certainly precedents for that. I think the big challenge here is to put the old data in the context of some of the newer molecules. And I view XL102 is really an optimized tool compound and optimized probe to really ask some of the key questions around the biology here, right? So we have a very potent, selective, irreversible inhibitor that was designed to have a relatively short half-life. So basically, the molecule gets in, covalently binds, links bonds to CDK7 and then the unbond -- unreacted material then is basically washed away with the short half-life. So you -- with the high selectivity and the short half-life and the covalent linking, you have arguably an ideal probe to start asking some of these questions around an orally bioavailable molecule that you can dose [ chronically ]. So I don't want to get ahead of the -- and speculate on what the tops will look like, all that kind of stuff. But I think from our point of view, it's just -- it's the right molecule to start asking the right questions in a tumor type -- a range of tumor types, either as a single agent or combination that have pretty significant unmet medical needs. So we've got to work through all that stuff. I think the molecule is coming before us. Had certain liabilities in terms of their potency, their selectivity, they're [indiscernible] half-life and/or route of administration. So we think 1 or 2 is the right tool for the times, and we're going to profile it. And if it looks good, we'll push it forward. If not, then we'll take it out back and bury it and focus on something else. So it's just all based on the data.

And based on the preclinical data, do you expect single agent activity? Or is the combination with fulvestrant, let's say, in breast cancer or is [indiscernible] prostate going to be where it's really going to be a combination approach?

Yes. I would say, in general, I think the combination approach makes the most sense, right? So in terms of large tumor types, that could impact the largest number of patients, the broadest number of patients going forward.

Okay. Let's move to XB002, your tissue factor ADC using the Zymeworks linker. And TIVDAK has now been approved through Seagen and Genmab. Give us a little bit of a sense relative to TIVDAK where are you hoping to show differentiation? And it looks like there's going to be data second half?

Yes. So very pleased with the early data we've seen so far, molecule entered the clinic and the around middle of last year and had a very productive second half in regard to the dose escalation with 002. Obviously, what we saw helped us understand the potential value for us as well as ways we could differentiate from the first-gen molecule. That prompted us to work with Iconic Therapeutics where we licensed the compound from to tie up all the loose ends in terms of having broader rights with the molecule, the ability to make other ADCs from their binder, again, which is -- again, is binds in a noncompetitive manner with Factor VII for tissue factor. So clear enhancement in the mode of binding, and that could lead to better -- potentially better safety. So we're really doubling down here based upon the data we saw in the back half of 2021. Differentiation could take place in a number of different ways. We think we're seeing that already from an exposure, safety, overall totality of data perspective. It's still early, right, obviously. But it's one where we have a very broad early clinical program in terms of expansion cohorts, and that's all been informed by our knowledge of the biology, but we've seen so far clinically and the overall profile to date. So we're excited about that. And obviously, as sooner we can move this into full development, the better. And it's a very, very big priority getting a lot of airtime within the company so we can make this go as fast as possible.

And which tumors do you think come to mind for tissue factor inhibition?

Well, based on our analysis and if you look at some of the slides we presented both at JPMorgan earlier in the year and then on our Q4 call, thinking about lung cancer, ovarian cancer, certainly cervical cancer based upon some of the early TV data, bladder, head and neck, pancreatic, either gastric and/or esophageal prostate cancer looks really interesting. Triple-negative breast cancer as well as hormone-positive breast cancer. So our expansion cohorts are broad here based upon, again, our knowledge of the biology and what we think is the opportunity with very significant populations where the unmet medical need is strong. So again, look at COSMIC-021 with cabo as the model for how we go forward, right? We get to a dose that we like, good safety profile, good exposure, and then we will have a broad signal searching approach that gives us the flexibility where we can -- if we see some good activity early on, single-agent or combination, we can boom, expand and potentially have a accelerated filing opportunity if it's there, but then certainly, then use that to then drive full development in pivotal trials. So the 021 kind of framework is how we're going to proceed going forward, right? Very aggressive, put some money at risk. We've got a lot now. So it's all about time and speed and getting the right data as quickly as possible to really prioritize our efforts going forward.

And Mike, we can't do a session without bringing the court case, which is expected in May. Can you give us a sense how long do you think a case like that normally will take? And maybe share whatever you can about your positioning, your level of preparation. And what are you seeing that you can share with us?

Yes. So I can't say a lot. Obviously, don't want to litigate this publicly. And what we said previously is that we are very, very aggressive at our resource allocation here to make sure that we've got the best team, the best experts, the broadest possible approach here, that complements the strength of the IP that we've got and the internal team that we've got. So look, it's a no-brainer that this is important to us and that we're putting a lot of muscle behind this. And I won't dive into all the details, the sell sides writing on this. The docket is publicly available, people can see that. We'll keep our head down and just keep working hard here. And again, the trial is in the middle of May. How long it goes for that's not to me. It's really up to the judge and that overall process, but we're going to continue to be very, very focused here and make sure that we do everything we can to win.

Okay.

Okay.

Mike, thanks for joining us. Really appreciate it. Those have been...

Okay. See you next year.

See you next year in person.

All right. Thank you. Bye.