Arcus Biosciences, Inc. (RCUS) September 2026 Conference Transcript

All right. Thanks for joining us, everybody. I'm Terrence Flynn, Morgan Stanley's U.S. Biopharma Analyst. For important disclosures, please see the Morgan Stanley Research Disclosure website at www.morgansdainley.com backslash research disclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. So I'm very pleased to be hosting this afternoon, Arcus. We have Terry Rosen, the company's CEO, and Bob Galtz, the company's CFO. Thank you both so much for joining me. Really appreciate it. And I guess, again, maybe just high level to get started. How should we think about the company, Terry? There have been a lot of changes over the years. Is this really a single product story, or is there a platform strategy here? I know a lot of focus on Cast Data Fan, which we'll dig into, but how should we think more broadly about what the company is doing beyond Cast?

So first of all, thank you, Terrence, for doing this and thank you for all those of you who are joining us and I think very clearly Castatafan obviously is a center of focus for us, for the world we believe very strongly it's going to be a 5 to 10 billion dollar drug so not surprisingly that question will come up but I think Arcus continues to be as we plan since its inception to be a company that's focused with a long-term vision. We have a very sustainable pipeline. We have a program, QEMLI, and frontline pancreatic cancer. It's an interim analysis happening this quarter. And so we feel that CasDaphan is going to lead the way, but that we can build a great long-term and highly valuable company. And if you put it in economics, we can become a $10, $20, $50 billion plus company. And so that continues to be great.

And what do you think is one thing that people are underestimating right now about the investment case?

Yeah, interestingly enough, that question, you know, you get asked that. And I'll share what an investor, what he thought investors were missing, and I think he hit it spot on. So when you think about CasDatafan and that opportunity, if you go back to the earliest data we generated, the PKPD profile, which is dramatically different from that of Belzutafan, and then you think about the data that we generated subsequently, A lot of the questions about the ultimate commercial opportunity with CAS data fan as we think about front line, second line, third line, should really just come from that simple observation because at the outset we just had those data, but now we've done three. We've had an investor event, we've had a nature publication, we've published all the details, And those data drive all this efficacy that reads through to all lines. In the front line, we don't even have a competitor. So I think the thing that's missing is those data tell you that Castatafan is going to win in the front line. No caveats, and I think the data are going to demonstrate that.

Yep. Great. Yeah, I saw the announcement this morning on the timing of the event. So I think, you know, you guys gave a little bit of a preview on the second quarter. you know, recap, but maybe, again, talk to us about, number one, why you're hosting this event, and then just, again, high-level maybe, what is the kind of key things that are going to be in focus, and then I know we'll unpack a lot of those and subsequent questions.

Perfect. So, the reason we're doing this, there's a convergence of data sets that read on all of the lines of therapy front line, second line, and a late line, and we felt like the best venue and best way to communicate all of that was in a single form, and it's over 200 patients worth of data. So since we'll talk about metrics for each of these and what we need to be successful, let me just describe what those various studies will be. So this is all of the data that we'll be presenting come under the umbrella of a platform study called ARC-20. Now, within that study, since we're talking about front line, second line, and late line, you can basically think about those three buckets of being the primary data that we're going to speak to. So on the front line, the standard of care that we're going against is ipinevo. So there's ClearCell, RenoCell, Carson, talking about A-standard care in the front line, the IPI-NEVO, and what we're going to have for that is a data set where we're looking at anti-PD-1 plus CAS data fan without the IPI, and we're going to have about 11-plus months of data to compare it to, and we'll talk about how that fits into the metrics. Now, that will give you a good feel for the efficacy, even without the IPI. The second component there is in that triplet we'll have anti-PD-1, IPI, and CAS, and what we're running right now is what will generate the safety data set that both the FDA and the EMA have told you to study at the end of this year. That cohort is fully enrolled, it's 30 patients, and what they want to see is 12 weeks of safety data. Why 12 weeks? Well, when you think about what is ipinevo, ipinevo is four cycles of ipinevo, which is 12 weeks, and then you get nevo for the rest of the therapy. So they want to see that 25 patients with that sort of follow-up, And by the data update, we'll have roughly 20 patients that have had that 12 weeks, so a very good read on the safety. Now, that brings me to the second line. So, again, if you think about what we're doing, we believe that we're going to in this field, but we're doing it in a way where we're providing a HIF-2 inhibitor on top of standards of care that physicians like. So we've taken a lot of feedback from physicians insofar as what they want to see, and that reflects our strategy. So in the second line, the standard of care, 40% of that second-line therapy is monotherapy CABO. And what we'll have is a data set combining CABO plus CAS data fan with roughly 20 months of follow-up, and that's a 40-plus patient data set. Finally, in the late line, we've used 120 patients' data set to really be between CasDatafen and Velazutafen, and I think that's what really established a major inflection in our program. Now, what we'll have for those 120 patients will now have a median follow-up, And so that's going to lead us to not only show how that data state continues to evolve and how it continues to differentiate itself, but for the first time we'll have an endpoint where we can look at overall survival. And given that Merck with Belzutafan has run three registrational trials, hasn't hit statistical significance on OS yet, we think that could be very, very meaningful.

Great. So let's dig into this. So I think on the frontline study, again, here you have a triplet, you've got a doublet. Talk to us about how you kind of choose between these. Again, I know you've already outlined the design of the phase three, but again, what else informs this doublet versus triplet strategy?

Yeah, so it will be a triplet. So it will be ipinevo cas, and largely that came through again, As I was mentioning, our ad boards, both U.S. and European, felt that no matter how good the CAST data fan anti-PD-1 data looked alone, to enroll the study, they're such strong believers that the IPI contributes to the survival that they felt running a doublet, we thought about a three-armed study would slow the enrollment down. So we see that may be a question after people see our data as to would anti-PD-1 CAS be good enough. It's going to be a two-arm study. The nice thing is they're very – so you can think of bookends on the benchmark front to the standard of care, which is ipinevo. That's really our only – is the existing standard of care. And the data for ipinevo, it has a rate of primary progression just under 20%, about 18%. It has a response rate in the high 30%, and it has a median PFS of 12.4 months. Now, what do physicians want to see? Well, ipinevo is really the preferred frontline regimen. It's most used. Roughly 30% of the frontline patient population gets that. and then the other 60 or so percent is fragmented amongst TKIs. The reason ipinevo gets that usage is because it has the best overall survival. And it would get more use if it either you could improve the regimen where you could bring down the rate of primary progression or improve even just the timing of that progression, which I mentioned was 12 months and change. So what we'll be looking to show is that we can lower that rate of primary progression, you know, down into the 10% range or so, and then also improve with that 12-and-a-half-month median PFS, landmark data would say you've got about 50% of the patients still on at about 12 months. And so with 11-plus months of median follow-up, and I think this is going to be one of the most surprising things, not only the data but the data set itself, it's going to be very clear if CasDatafan plus anti-PD-1, even without the IPI, looks better than ipinevo. Now, another data set that's very important that most people aren't aware of is that Merck has actually run a study in that exact same patient population with just Keytruda. So what's important about that is when you look at our anti-PD-1 cast, that gives you the other end. You've got ipinevo, and, well, what would anti-PD-1 do alone? And it's very different than ipinevo. So for Keytruda in that setting, same patient population, you see a 30% rate of primary progression and a PFS in eight months. So with those guardrails, we think you're going to come out of that October event with a very clear pin if Ipenebo is going to be a success.

And I think on ARC-20, for at least the doublet, you guys already reported that you had, what, 7% primary?

So the doublet showed a 7% rate of primary correction. We reported that at, I'm sorry, at J.P. Morgan this year. We also described that the waterfall plot looked really good. So I think the dots are going to connect very well. The other piece there is that since we will have those 20 or so patients, that we expect a safety profile that should just look simply like ipinevo with perhaps some anemia that comes from the on-target activity of the HIF2 inhibiting component of the therapy.

Okay, great. And then maybe just remind us in terms of that, you know, peak 20 trial, like what's the timeline for that and size of it and anything on, you know, powering, et cetera, is there yet?

Yeah, so we really will share more details on the design, et cetera, as we get closer. But it's on track to start at, you know, the end of this year. And, in fact, the safety data, we've already had the discussions with the FDA. and as I said, that cohort of 30 patients that will provide the 25 with their 12 weeks is already fully enrolled. And the study, the primary endpoint will be PFS. So that tells you not only do we expect a fast enrollment, but the readout should come relatively quickly, knowing what it looks like.

And just for a minute, on LightSpark 012, Two, that was the other Merck study that recently had a setback in that study. What was the driver? What was your kind of conclusion on the reason for that?

Yeah, so I'll answer that in two ways. What do we think happened? And second, I always feel that question behind the question on that is, does it have any read-through to us? And what we feel the primary driver is going back to that what is everybody missing piece, which is that initial PKPD data that we discussed. So what's known is that biomarker data revolves around looking at the ability of your HIF2 inhibitor to diminish the production of erythropoietin. So that's the gold standard for inhibiting HIF2 in a human. And what's known is while Casdatafan robustly inhibits that well out past the year, Belzutafan, those are Merck's own data. So we feel unquestionable lack of that durable effect that really manifests itself, particularly in the front line. So if you look at the various Merck registrational trials, They've gone first in the late line versus something that had a PFS of five months and changed to something that had ten months. So we feel like absolutely the efficacy that shorter time of hitting the target hard is getting diluted out. A lot of people speculate that the triplet might be more toxic than the doublet. But we'll see. But the important thing is we think it has zero read-through for two reasons. First of all, it's a completely different regimen. In that case, you're going on top of TKI anti-PD-1. In our case, we're going on top of ipinevo. And the second thing, most importantly, is we already have a lot of data. You'll see in October we have a lot of data in the front line, Not only that, it shows a very profound effect of castatophant in that patient population.

Okay, great. Maybe we'll pivot to the second line setting now. So, again, just remind us kind of what's the benchmark that you're aiming for here at this update in October?

Sure. So we view second line as a huge opportunity from Belzutophant. And so Belzutafan was successful in the second. We believe that there's two opportunities apart from the Belzutafan limbatinib. Opportunity on the outcome for pain. A little bit about that. We've recently described population populations, the LightSpark 011, and our ARC-20 cohort as well as our registrational trial. The first thing is, in their study, they had TKI in the front line. They excluded patients that had received copozantinib in the front line. And also, those patients that did get TKI, so 25% of the patients in their study, not just 25% of the TKI patients, 25% of the patients got sudintinib or another first-generation TKI. So the first thing that we would just say about the patient demographic, it looks very, very different. In the real world, 60% of the frontline patients are pretty much going to either get Cabo, Lenva, or Axie. And so they have zero Cabo, zero Lenva, and all those patients with the first-generation TKIs. The second thing that we'd emphasize about that is that it's very different than the ARC-20 and PQ1 patient populations. So, and what I would say in the ARC-20 patient population, it looks very much like the distribution from the real world, including a very substantial portion that are LEMBA patients. Now, the important reason that we mention that isn't to set expectations that since we have a more difficult-to-treat patient population that somehow we should expect a lesser performance. But what we wanted to point out is that given those more difficult patients to treat, we feel it's a huge opportunity to essentially show that castatophant and incremental advantages, but completely discontinuous. So one other fact that I can use as an example, and there have been recently two publications, real-world data. If you get limbatinib in the front line and then get treated with CABO, your ORR drops to like 5%, and the PFS drops to four to five months. In fact, your OS is substantially hurt as well. So we feel we have the opportunity to actually do phenomenally well in those patients. So we're thrilled that we have these not only TKI patients, but LENVA patients. Now, part of the reason for our, you know, feeling very strongly that we'll get to the real data, but we've already generated data in the late line that reads on this. So this is part of our Nature publication. So what Merck showed in both LightSpark 05, their approval study in the late line, as well as LightSpark 013, which is essentially the same population, different study, is that they showed an inverse relationship. The more TKIs that patients were treated, so I'm going from three, the more TKIs you had, the worse you did. Actually, what we showed in our 120 patients in the late line is no dependence on that. So whether you had one TKI, two TKIs, three TKIs, and we'll show you the curves, They're right on top of each other. So we think that relates to a couple things. Now, one last point that we showed is that in that same study, we showed that the more HIF2 activity that a patient had when it showed up, the better that patient did. Now, HIF2 is one of the worst prognostics you can have, not only in RCC, but in cancer in general. So the fact that we're doing better in patients with the worst prognostics is generally taken to mean that it's causative. Now, what else do we know about the biology of HIFT-2 and TKIs in general? And what's known is that when you treat a patient with an anti-angiogenic, including a TKI, that causes the patient to have more HIFT-2 activity. The belief that's going on is that those patients that are getting those hard-hitting TKIs in the front line are actually now coming into the second line with one arm tied behind their back, particularly in a world that's HIF2 free. And that's been shown in a number of studies. So if you take one of those patients and you put them on Cabo, they really do poorly. But we feel that with Casdatafan, that's hard-hitting and durable late line to actually do better. in those patients. So I'm not just saying incrementally better. If you look at the hazard ratios in the patients that were treated with TKI coming into that versus IO-IO, the hazard ratio is better in the IO-IO. We actually feel like we have the opportunity in our registrational trial that the CABO patients obviously will perform more poorly, but we feel we could do as well or better. So we actually feel like we have an opportunity to totally shift the paradigm, not just do something that's better, because 60% of your patients are going to get a TKI in the front line. So coming into second line, they should be on a therapy with a F2 inhibitor. The second point that we'll be able to read on is less months of median follow-up. As I mentioned, Merck already we know missed, even though numerically, statistical significance they did not have. And so what we know is that Cabo's OS is roughly 21 months and change. So we will share OS Keppelin-Meyer curves, and so we have an opportunity on two fronts, that ability to better serve patients that have TKI treatment in the front line and the opportunity to win on OS, which we think for all practical purposes, we would own the entire second line opportunity there. So we're very excited about those data. You'll see those in October as well.

And then just any caveats as we do that cross-trial comparison? Like, again, you mentioned some of these baseline characteristics, but anything else as we make that comparison to the 21 months with Cabo?

The one baseline thing that I would mention when we think about what Belzutifan has done is their patient population. But when you talk about Cabo, that's the perfect comparator because that's still going to be. We actually did a survey of physicians. The LightSpark 011 data read out. and we talked to 5-0, 50 physicians, and about half of them said they would go to that LEMVA-BELS regimen. But CABO will still be primarily used. It's well-defined. It's well-understood. Physicians like it. So it's the perfect.

Okay. And then maybe bake that all together. What does that mean for your peak one phase three trial? That's, again, the first Phase III trial that you guys are doing. But talk to us about kind of design of that, timelines. I think you said enrollment completion by year end.

Yeah, so peak one is enrolling rate. We expect it to be fully enrolled by this year. It's a two-to-one study arm versus control, so we expect it to enroll rapidly. and importantly, 20-patient population that will be representative of what you expect to be coming out of the front line. And, of course, we feel that's important because not only do we want to be successful on that ITT population, but those sub-patient analysis, again, are very important insofar as dramatically changing the TKA population. And then finally, you know, we will be able to give you a sense of, since we will be sharing Kaplan-Meier curve and OS, this one, Belzutofan did not hit on this. You don't even have to think about the cross-trial comparison. and it'll just be, you know, how's our hazard ratio going to look versus Cabo, and there's plenty of data out there when you see how our OS Keppelmeyer curve is looking from ARC-20 to give you a quite strong feeling of optimism on that. And we'll talk about, you know, what we're going to learn from the late line that I think we'll read through to that.

Yep. And just the primary of peak one is PFS and then OS. I'm assuming it's powered for OS.

So the way this is set up, it's very nice. The primary endpoint is PFS. If we hit on PFS, all of the alpha.

And just to remind us, so that, again, enrollment completion by year-end, what does that mean for timing of data?

We haven't given guidance, but if you were to just do the math, given that the control arm is about 10 months or so, it probably takes you into early 2020.

And then I guess just anything else on the, you know, I guess market dynamics that we need to consider here as you think about launching into that second-line market, just anything else? Again, you walk through a lot of different variables here in terms of, you know, some of the TKI use and things like that, but just any other considerations that we think about kind of, you know, framing that second-line market potential for us either on the duration side or other dynamics?

Yeah, I think, you know, Terry described the potential for us to differentiate on the majority of patients that are coming into the second line that have previously seen a TKI. We think that would be huge potential. In terms of the actual commercial execution, we think this is an opportunity that Arcus is imminently capable of executing against. You know, it will be a targeted sales force and commercial footprint. We think the path for a HIF-2 inhibitor will be well-worn at that point with belzutafam, but we think we have the potential to generate data that puts us in a strong position to capture the market.

And what is that? Maybe just to elaborate on that a little bit, what does the commercial build look like in terms of rough numbers of people?

Yeah, it's pretty typical for a kind of a large-ish oncology tumor type. But, you know, Salesforce sort of in the 100-ish reps kind of zone, you know, we think that there's been a pretty well-worn path for a smaller biopharmist to execute on that type of indication.

And does that bill, that spend starts next year, most likely?

Yeah, if you look at, you know, a launch that wouldn't be before, let's say, late 2028, you know, basically a year out is when you see the more significant investment for launch preparation.

Okay. And how are you guys? I know you already have a partner in Japan, but how are you thinking about kind of the European side of it?

Yeah, for right now, we're very focused on maintaining strategic optionality, especially with Western European commercial rights. You know, not to say that if we got closer and we launch that we might not leverage somebody in that area.

But for right now, we think it makes a lot of sense for us to retain those rights. and I guess in terms of, I mean, Terry, you already talked about this, but the $5 to $10 billion, maybe just break that down for us by line of therapy, like what you guys think is, you know, second line, third line, first line.

Yeah, so we think second line is, I think, with a very conservative duration of treatment is $1.5 to $2 billion. You know, if you look at just as a marker right now, Belzutafan is doing basically a billion-dollar run rate with the majority of that coming from the late line. There's a little bit of second-line usage monotherapy, but most of it's late line. And obviously their late line, the PFS, was 5.6 months. Our monotherapy has shown PFS that's on the order of double. So we think there's upside from a duration of therapy perspective, and then obviously the market size in the second line is also more significant. So we think $1.5 to $2 billion is very conservative for the second line. And then in the front line, you know, if you look at the Ipenevo regimen, right now it has roughly a 30% share in the front line. We think with compelling data, physicians really want to use that regimen because it gives patients the best chance for long-term survival. If we showed that we can bring near-term disease control, augment long-term survival with a tolerability profile, that's probably more attractive than the PD-1 TKI combinations, that it wouldn't be hard to imagine growing that share from 30% to 50%, which translates into a $4 billion-plus commercial opportunity.

You know, the one thing on the durability that I think really emphasizes the point is in the late line, so we are going to read out on OS on that with the 30 months, But to give you a sense in thinking about this even before we get to October, so approximately 28% of those monotherapy patients were still on the therapy after two years. We've even had a couple patients convert from stable disease response after two years. So when you see the spider plot, it really gets to the point of a very, very therapy. And that's in the late line in a very advanced patient population. I was like, if you look in the late line, we even have compared, when you think about Belzutafan, we have 20-some-odd percent of patients that are fourth line plus Merckhead Well, I think we're up against time, guys, but I'm looking forward to seeing you again in October. Thank you. And we think we're going to have a really exciting data set. We look forward to talking about it that day and thereafter. Thank you, Terry. Thanks, everyone else.