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

Okay. Well, welcome once again to the 40th Annual Health Care Conference at Cowen. I'm Yaron Werber, biotech analyst here on the team. And it's a great pleasure to moderate the fireside chat with Mike Morrissey, who is the President and Chief Executive Officer at Exelixis. So Mike, thanks so much for joining.

Great to be here. Thank you.

Maybe the first question to you is, how many years has it been now for you at Exelixis?

Wow, I just celebrated my 20th anniversary on February 3. So it's like 10 biotech half-lives. So happy I made it so far. So a lot more time to go, for sure.

It's pretty impressive. So congratulations.

Thank you. Thank you.

And it's been a quite run.

It has been, yes, quite a run. Understatement of the year.

So let's talk about Exelixis of the future because there's a lot going on in the company and things can -- are definitely on the -- on an upward trajectory, potentially over the next 1 to 3 years. And so I was going to start by talking about ASCO-GU, literally, just ended about 2 weeks ago, roughly. And the data from cohort 6 of COSMIC-021 was fairly intriguing, showing a 32% response rate, about an 8 months of durability so far. Where do you think -- where are you in the durability curve? And maybe give us a sense how many patients are still on therapy of the original 44 to give us a sense kind of what durability could go.

Okay. So before I begin, let me just say, good morning. It's great to be here. I will be making forward-looking statements today, so please see our SEC filings for a description of the risks that we face in our business, to get that out of the way first. Yes. So look, we've gotten off, I think, to a pretty strong start this year. I think we tried to reframe the narrative at the JPMorgan Conference in terms of the bigger Exelixis picture, outside of 9ER and weekly trips into the -- a lot of other options and trials that we've got, talked about prostate. I was pleased to be able to follow it up with data at ASCO-GI and GU. So we've got a lot going on. I think that's the -- that is -- it kind of portends our future relative to all the trials we have ongoing, new INDs, new molecules coming out of discovery. In terms of prostate, the whole goal here is to be able to really help define a new standard of care in that post-NHT setting. You've nailed the top line data really well, 30-plus percent response rate, 80% -- or 8 months duration of response, which, in that population, I think, is pretty notable. In terms of what's going on beyond what we had in the poster, I'm somewhat unable to go beyond that. We still have a lot of patients enrolling. A lot of patients from the first 30 or so patients on trial. So we're pretty excited about where that's going relative to having a combination that really seems to highlight the best of both molecules. Cabo is rapid onset of action, tumor shrinkage, the long durable responses that you can often see with IO molecules combined together in what's been historically a pretty cold tumor type like prostate. So lots going on there. We'll launch the pivotal trial that will be the confirmatory trial soon as well that we're excited about. So it's certainly great to be back in prostate, and we think we've got some pretty exciting offerings on the way.

And as we think about a pivotal trial, what would that look like? It will be a cabo and TECENTRIQ versus what?

Yes. So that's the key issue. Again, we've agreed with Roche to highlight the details of that trial once we actually initiate that. So stay tuned on the actual details around comparators and those kinds of things. What I would say though, is what we've talked about during the ASCO-GU week that we're not looking to compete with chemo, we're looking to provide options that will delay chemo. So I think that goal will be further highlighted in how we actually design and run that pivotal trial.

Does it make sense to go in a patient that has seen an XTANDI or an NHT previously, dual combo versus one of those agents after they fail them? That's what...

Sure, sure, sure. Well -- and in fact, if you look historically at what's been happening either in the recent past or ongoing, there's 13 trials in prostate cancer that we found that are looking in a post-NHT population that are using a second NHT as the comparator. So I think this is a fairly well-established approach in terms of being able to define new active molecules, new active combinations in a way that provides clearly more optionality in that post-NHT setting.

Okay. Any new thinking about what caused the TECENTRIQ cabo synergy? It looks like there was synergy in it.

Yes. That's certainly the feedback that we're getting here. I hesitate to use the "S" word with 44 patients, even though single-agent cabo is about 5% in this population in terms of response rates. And if you look at some of the early data with atezo, it's by RECIST 1.0. So the -- there's something else going on there. Well, I think it's -- in some ways, it's the classic right combination approach, right? We've been talking for the last couple of years now about how cabo really provides an immune permissive environment in terms of going after VEGF, going after MET, going after the TAM kinases. We had literally every cell type in the tumor microenvironment that matters in terms of having this immune permissive environment really be expanded. Rapid onset of action in terms of tumor shrinkage that can provide a very rich opportunity for new antigens to be released and expressed as tumors become necrotic and die and kind of spew out all their cellular materials. So -- and that's really the optimal environment then for a well-established IO agent that can activate CD positive -- CD8-positive T cells, so they get in there and do their job too. So we like that. And think about that, we've seen that or signs of that in renal, we've seen signs of that in liver, we think we've seen signs of that in lungs. So we think this is a generalizable phenomenon that goes with combo and kind of any IO agent in relevant indications outside of those that are highly thought of as being hot tumors, like melanoma, right? So we'll see. But it's certainly something we're following up on. And certainly, it's a big focus for what we're doing with cabo as well as 092 going forward; can we maximize the opportunity here relative to making that synergy as optimal as possible.

Yes. If you had -- it sounds like you've had a lot of interactions with FDA relating the ability to use the study for registration. I'm talking about COSMIC-021. What about EMA?

Yes. Those are different conversations, a, there's really no Subpart H equivalent in the EMA CHMP world. So we haven't focused on that as a regulatory opportunity.

So would that be a full Phase III, I imagine. The confirmatory pivotal would be for European purposes.

Yes. Sure. Yes, yes. For global registration, for sure. Yes.

Okay. And the -- again, maybe just to go back to the FDA registration or the FDA discussions, give us a little bit of a sense maybe, I don't know, if you can, how did the idea of coming through and delaying chemo come about? Was that an idea that Exelixis sort of went to them, it's just something that came out during the interactions with the agency? Because a lot of other companies are doing studies in that area. But they've not been thinking about delaying chemo, at least they haven't discussed that publicly as an accelerated approval pathway. And I'm sure there are going to be lot of companies kind of waking up saying, why didn't we think of this?

Yes. So I won't go into the details of our conversations with the agency. Those are confidential and I never feel good about people talking about subsegments of those conversations and not the whole thing. So I'll leave that kind of on the side. There's certainly a very clear path within the oncology section for Subpart H filings, either single agent or combinations. Since 2017, there've been 39 Subpart H approvals within oncology, with really a focus on response rate with good duration of response for the combinations where you've got the contribution of components well covered. So we feel good about the approach. We've been following, certainly, I would say regulatory precedents in terms of our design of the 021 overall protocol, where we had the ability to rapidly and in succession based upon data, expand cohorts and -- as well as add cohorts, single-agent cohorts or the like that would give us the requisite insight in the totality of data to be able to move forward. The issue around delaying chemo is -- it actually -- I think it's novel, but it's not all that unknown. I think patients and many physicians try that anyway. If you look at the patients we enrolled and talked about at ASCO-GU, we had 44 patients in that poster, half got 1 NHT and then came on the trial. The second half got 2 NHTs and then came on the trial. A small fraction, about 30% or so got chemo, but that was for their castration-sensitive prostate cancer early on in their disease. But if you look at the 1 versus 2 NHTs, half those patients made the choice I'm going to get a second NHT even though I could go on chemo. And then even now when they progressed, they still made a choice of going on this clinical trial versus chemo. So it just underscores the idea that chemo certainly has a measurable level of activity is used, certainly used more in Europe than I would say in the U.S. But it comes with pretty heavy baggage from an adverse event safety point of view that for somebody who's in their mid-70s, early 80s has had radiation and surgery for their prostate cancer, been on ADTs for years, if not a decade, then some of these novel hormonal therapies. I mean they've been through a lot. So if there's other options available that can provide them with clinical benefit that goes above and beyond kind of what they're expecting with -- going down the chemo route, why not pursue that approach, right? So that's been kind of our view. And I think so far we've been, I think, happy with the response we've gotten overall. Yes.

What's the amount of durability that you think you need to show? I imagine you're going to want to follow patients for a year. I mean that's typically, I think -- used to be 6 months, now we're hearing increasingly 9 to 12 months. But what's the right level of durability that you need to have to -- for the FDA to feel good, do you think?

In terms of follow-up?

Well, both follow-up and actual durability?

Well, I think durability of response, I think we've got that. In this case, in a refractory population, I think 8 months is certainly notable, if not exceptional. In terms of follow-up, I would say, 6 months is the standard kind of approach here.

That's enough?

Yes, yes. Sure, sure.

Okay. Okay. I was going to shift over and talk about a little bit 9ER and your latest -- I mean I think we're all expecting data either late this quarter or sometime early next quarter. I think in our calculation, I think what you've said is that the study was fully enrolled around March last year. The study, ultimately, I think the number of events was upsized a little bit as some of the Japanese patients came on. And when we look at some of your -- the comps, they -- those studies were stopping around 9 months to 12 months post randomization or post the last patient in. So we're getting to the same time frame. Maybe help us understand -- there's no question that you'll hit PFS, at least in our mind. The question is, is there just enough events and enough powering and a question of crossover to also hit survival on the first look? So what have you done to maximize those chances?

Yes. So I agree. I think statement of the obvious here is that we need survival as quickly as possible, certainly in the first look, if it's there, to have a competitive offering. That's just a statement of the obvious, right? So we had survival with METEOR. We knew and we saw how important that was in terms of the launch. The main competition here has survival. So I think what we've done is effectively maximize our chance of success there by essentially letting the clock spin a little bit longer to get more events and generate more power behind that survival readout. So it's an experiment. I won't opine upon that further. I have a high degree of confidence in that happening. But let's get the trial done and look at the data and go from there. Again, I would speak to the Phase Ib data for cabo, cabo/nivo, cabo/atezo, looks really good, right, both in terms of response rates, duration of response, PFS, safety. So we feel really good about the horse we have in this race, right? Certainly, cabo has been -- it's an exceptional TKI from the standpoint of it the only one that has survival for RCC as a single agent. It's the only TKI that beat sunitinib head-to-head with a hazard ratio in the 0.5 range for PFS and it's the only molecule that has survival in both RCC and HCC. Only one. So with that kind of profile, I like our chances relative to these combinations, especially when we're seeing the kind of tolerability and safety that we had, say, in the ASCO-GU data for CRPC. Again, different IO agents, different setting, if you will, but still exceptional tolerability in terms of only 1 grade 4 event, small number of grade 3 events, low dropout rates, et cetera.

Okay. What about -- what have you learned from the recent CheckMate 214 data as to how you think about COSMIC-313?

214, in terms of nivo/ipi, certainly, is showing good durability in their responses. The question is how that translates into what's happening in the field, which is always, you see a little bit of erosion, which wouldn't be a big surprise. Look, 313 is simply asking a different question than 9ER is, right? Can you take one standard of care which is IO-IO and make it better across the whole population as opposed to the PD-1 positive population. That allows you to potentially move the needle in a pretty dramatic way across the entire population by adding best-in-class TKI onto what's obviously best-in-class IO-IO combination. So that's enrolling well. Lots of enthusiasm for that trial because I think -- well, 2 reasons. Number one, everybody likes the idea of going against ipi/nivo as the control. And everybody sees the upside if you can apply the learnings from some of the doublets to this triplet in the PD-1 negative population, which didn't really perform as well as the PD-1 positive population in 214. So lots of moving pieces there. Certainly, very excited about that. And great to see it enrolling so fast with the enthusiasm on a real global scale.

Yes. And is the PD-1 status stratified in 313?

No.

Okay. That's across all patients?

Yes. Yes.

Yes. And so in your view, nivo/ipi how is that used now, especially given some of the INLYTA-KEYTRUDA data?

Yes. So what we see in our market research at a relatively high level is ipi/nivo dominated the front line upon launch and before the 214 data led to approval when axi/pem came online. We saw that kind of re-equilibrate to steady state, which we saw, end of Q3, early Q4 last year, where frontline, probably 70%, 75%, 80% depending upon the week of the month was IO combinations, and that was split more or less 50-50 between IO-IO and IO-TKI. And the IO-IOs are used mainly at the academic centers. Those investigators, those prescribers are usually big IO believers and their therapeutic intent, as you talked about, is really on the cure. Whereas in the community, it's much more pragmatic, and they're very comfortable with TKIs and maybe a little bit better safety profile with the IO-TKI combinations. So we see that evolving. And the question then for us is between 9ER and 313, depending upon data and how that looks at the end of the day, how do we slot in, how do we compete relative to the other offerings, either on the IO-IO space or the IO-TKI space as we go forward. So -- but I like having both trials going because under any circumstance that is differentiating with data we could dominate that space going forward.

And so what's the share now of IO-IO in first line?

It's about 30%, 31%. Yes.

And what happens to their share in second line now?

Nivo -- single-agent nivo is decreasing pretty dramatically month-over-month. As the single-agent mono TKI therapy front line kind of goes, those patients progress and new patients come on to combinations, the use of single-agent TKIs is down in the 15%, 20% range. So therefore, the subsequent first line -- subsequent therapy in second line for either IO or IO combinations is shrinking pretty fast.

Yes. What about -- is IO-IO getting more prominence in second line, now that it's being kicked out of first line?

No, not really.

Not really?

No.

Okay. Okay. Going to shift over to -- you're actually running now a study with the HIF alpha -- the HIF-2 alpha inhibitor for Merck, that the Peloton acquisition, 6482, right, with cabo or...

Well, that's the Peloton trial...

Yes. Peloton study.

So the Peloton ran that study well before. They got that going well before the Merck acquisition. So that's really a legacy trial from Peloton.

It's a legacy study in Phase II.

Yes, yes, yes.

What are you expecting to see from that combination?

Yes, good question. There's very little data out there right now relative to what that combination could do. So we'll see if there's any additivity or synergy or pathways overlap in terms of VEGF and HIF, obviously. So we'll see. Don't know. Yes.

And those are in patients that failed the TKI? So it's an add-on?

Yes, yes, yes.

Okay. I'm going to shift over now to COSMIC-312 in liver cancer.

Good.

So give us an update. The study is still enrolling. It's expected to finish enrollment, I believe, this half and hopefully, interim analysis, second half toward the end of the year?

Yes. Correct, correct.

When we looked in our analysis and we looked across the landscape, we thought it was -- there's no question the triple nivo/ipi/cabo looks really good. The question though was dose and the efficacy was really great. There's also some more tox and so as you kind of lower the doses of, let's say, ipi or ipi/nivo/cabo, the efficacy started kind of losing a little bit steam and then when we looked at cabo/nivo, that looked really good. INLYTA, KEYTRUDA looked very good as well. And obviously, you have IMbrave, 150 looks good. So we thought that was sort of the 3-horse race. So as you think about what would differentiate cabo/nivo? And how do you think it's going to fare versus IMbrave?

Yes. Well, first, just to be clear. So 312 is looking at cabo/atezo, right?

I'm sorry. Yes.

We had some legacy data that BMS published at ASCO-GI, looking at cabo/nivo and cabo/nivo/ipi. That was, I think, really encouraging. If you look at the data, at kind of a 30,000-foot level, they looked at 2 different regimens, the double and the triplet in the mix of first and second line patients. So you start slicing and dicing that -- those 71 patients into 4 quadrants. It gets to really small ends. So if you combine appropriately and ask some key questions like we did, I think the take-home messages around cabo is that you had good response rates and these are tough tumors to image. So I don't put a lot of faith in a response rate number in 15, 30 patients. The DCRs were high, which I think is important because your primary refractive disease is down. DCRs were in the 80% range for both the doublet and the triplet. The survival numbers were high in the 20-plus month range and the triplet wasn't even measurable in terms of -- it was beyond, the medium was well beyond where they last looked in terms of that complementary. So I think that data set for all the caveats of it being small and this kind of slicing and dicing of the quadrants across regimens and lines really provides, I think, some pretty compelling insights into what cabo can do when teamed up with an IO. IMbrave certainly is going to, again, set a new standard of care for front line. It will, we believe, effectively grow the market size dramatically because it will pull all the patients who got sorafenib for a very -- or lenvatinib for a very short time frame before either becoming intolerant or progressing or those patients that were going down the taste route first. So you'll see the pie grow pretty dramatically, I think, out of the box. You will, of course, extinguish your IO frontline with that. So that will leave the second line open to things like cabo, which have a survival benefit second line, which we're excited about in terms of kind of early traction. And then we've got 312, which is cabo with survival as a single agent plus atezo versus sorafenib. And if you look at bev/atezo versus cabo/atezo, it's hard to draw comparisons side-by-side, but I certainly like the -- at least my comparison of cabo versus bev in relevant tumor types like renal and like liver, et cetera; cabo performing pretty well. So again, I like the horse we have in that race, and we're going to hit that really hard. It's enrolled extremely well. It was going fast before IMbrave read out. And since then, it's just taken off like the space shuttle. So we're excited about getting that one done. But in terms of actual growth opportunity, this is a huge indication globally in terms of the number of patients involved and it's in terms of the -- its actual market size is relatively small. So in terms of actually growing a market, building a market with better data, with the right level of collaboration, we're really excited about that. Yes.

Yes. And the randomization of 312 is about 6:3:1. Why do a cabo-alone exploratory arm in first line? Is it just to show...

Yes, that's actually changed from 6:3:1 to 2:1:1.

Sort of 1-point -- yes.

And again, it's just to be able to provide the regulatory framework for the contribution of components. That's all. So yes.

I mean theoretically, cabo may show some single-agent activity as well, you would expect?

Sure. Yes. Which is why we're asking the question. Yes, for sure.

Okay. And is the study sufficiently powered to hit survival on the first look? Or is it really more of a PFS look initially?

No, it's -- I think we have enough power and enough -- it's all event-based, obviously, with all the caveats there. But no, we're certainly -- the fact that IMbrave readout with about 500 patients for both PFS and survival. We have more patients, certainly looking at a similar population, same comparator with sorafenib. So again, we'll see the data when it comes out, but expectations, in my mind, are pretty high. Yes.

Okay. I'm going to move to 092, the son of cabo.

Good.

And -- so give us a sense, maybe just what's the next update to expect. And enrollment, I think, is ongoing, it's been -- it's trucking. It's been a little slower. It's obviously competitive to enroll out there in general. So maybe just understanding a little bit the timing.

Yes, I will comment on enrollment. We're making good progress there. 092 is -- you really asked a bunch of simple questions. We understand the preclinical and clinical pharmacology of cabo really, really well. We've spent a decade or more studying that molecule clinically, single agent in combination with a variety of agents. We've importantly learned where the sensitive tumor types are. And that's from a standpoint of how you define your dose to use for chronic therapy. The first thing you've got to find is a sensitive population to be able to ask that question. So we've done all that background work with cabo, and we know a lot more now about both populations of patients and tumor types as well as cabo itself to be able to make a better next-gen molecule. So we've put all that information together between our pharmacologists and our chemists and designed 092 to be essentially a better next-gen cabo, right? And now we're going through and validating that work clinically as both a single agent and soon in terms of combinations with various IOs. So that's all work in progress. I'm excited about that. I think we've made the right molecule and addressed the right, really, pharmacological components of that profile to be able to give us potentially better activity. We'll talk about that later in the year at a major meeting to be able to give you a sense of its profile, both preclinically and probably clinically as well as its PK/PD relationships, which really underscore that. So I don't want to preempt that right now. It's a very competitive space. We're not trying to hype it. But certainly, in terms of the cabo franchise, you think about what Celgene did with thalidomide and Revlimid, et cetera, we're trying to play the same game here. Learned from the first molecule, designed what could be a better molecule and then applied that broadly across different tumor types. So stay tuned.

Is the goal to also sequence to go into a cabo failure and to go into other tumors...

Yes, probably not. I mean again, going into renal, with 092 probably wouldn't make sense as we're so heavily invested there. And we think we've got LOE through 2030. So we're comfortable with the renal kind of cabo owning that space. But new indications, say, like prostate, where you could split between the bone and no bone. For example, with 092, other indications that are coming out of, say, 021, we have plenty of room to maneuver here because cabo has generated so much single-agent activity across different tumor types. We're seeing the special activity in terms of combinations. So we've just got the ability to mix and match and merge and purge here as we go forward. So it's a great opportunity for us, for sure.

Okay. I think the last time we saw you, we were talking a little bit about BD, and Exelixis has a terrific cash position. You're, obviously, cash flow positive as well. But any sense kind of where you are? And what -- how close are you to doing something next to build the pipeline? And what are you looking for?

Yes. Look, we're heavily invested across the organization in finding either assets and/or companies that we have a lot of conviction in, right? And I think conviction is the focus here around making the right investment with the right lens to be able to maximize our chance of commercial success. Because from my point of view, that's where the risk lies today. I think everybody understands clinical risk and kind of managing that and increasing the chance of success there. We all understand that the regulatory risk, if you kind of play by the rules that are well-defined now for the last decade or so the chance of that working is pretty high. What I've seen, and I think what we've seen as an industry is, that's especially in the small mid-cap space, but also within big pharma, there've been a lot of compounds that have been approved based upon good but not differentiating data, which then fall flat when they're commercialized. And the list, certainly, in the small-cap space can go on and on. And I think cabo is the outlier from the standpoint of what we've done with that, where we had differentiating data, we had a great team, we had the right level of energy and focus, and we got that done, and it's $1 billion a year global brand now. So we're looking for assets like -- to a certain degree, like cabo, and those are few and far between. We don't want to overpay whatever value we ascribe, and we see ourselves as value investors. We want to make sure that we're paying the right price, and we're not getting sucked into an option where it just gets beyond our control. So we're careful. We're thoughtful. We worked hard to build this cash position, and we're going to be very pragmatic about how we use it going forward. So...

Yes. Are you looking at solids more than hematology, I imagine?

Yes, absolutely. That's our strength. I mean we could do hem/onc. We certainly have that expertise within development. From a vertical integration point of view, that would be a whole different commercial organization, and it probably doesn't make a lot of sense to do that right now.

Okay. What about ex U.S. rights? How important is that?

I would say, in general, whether it's for internal assets or assets that we're looking to acquire, we want global rights. Partnering cabo ex U.S. made a lot of sense. It's been very financially lucrative for us. We could -- it wasn't -- it didn't make sense to any of us, and certainly not me, to bite off more than we could chew in terms of launching commercially. You see some of the SMIDs that have done that that hasn't been real successful. But -- and it's a huge financial drain. So -- but going forward, to be able to own assets globally, monetize them globally, commercialize them globally, makes a lot of sense, and that's the plan going forward across the board.

Well, terrific. Mike, thanks so much for coming. We appreciate it.

You bet. Thank you. All right.

See you. Thank you.

Elbow bumps.