Good afternoon. My name is Peter Lawson. I'm one of the biotech analysts at Barclays. Welcome to the Barclays Global Healthcare Conference in Miami, and really pleased to have up on stage with me the management team from Arcus. I should also mention if you want to ask questions, raise your hand, and also I'll be on Bloomberg and some of my associates that do ping us, if you have anything. With that, I'd love to introduce Arcus. W e've got Terry Rosen, CEO, Jen Jarrett, Chief Operating Officer. I'd love to go through HIF-2α since it's a very prominent part of the story at the moment, and then, of course, gastric and 73 and lung cancer. I guess on HIF-2α, what should we expect to see in the expansion data for RCC in H2 , and what would we see in terms of durability as well?
Great. So let me give a little background into that program. So we've had a HIF-2α inhibitor program ongoing for some time, and I think it's been of pretty broad interest. But at our last earnings call, we talked about our dose escalation data and gave a bit about the expansion study that's ongoing, and that will be, you know, talking about a medical conference second half of this year. T o frame the HIF-2α program, really what we're focused on is a molecule that's an improvement upon Merck's HIF-2α inhibitor, belzutifan. It's an interesting field because f rom our portfolio molecules, it's really given what we've generated to date, we feel very confident that this molecule is a drug.
A t this point, it's how are we going to compete with Merck, how are we going to develop it. W hat are the data that we have, and what are the data that we're planning to share? A t that earnings call, we primarily focused on our dose escalation data. T he key takeaway from that is that there's a pharmacodynamic marker for HIF-2 inhibition. It's production of erythropoietin from the kidney. Peloton's use it, Merck's use it, we use it. It's a very robust marker. It doesn't have to do with the mechanism of action in the tumor.
On the other hand, it's absolutely linked to the AE profile, the molecule. I f you inhibit HIF-2, what you see is an inhibition of HIF-2-mediated erythropoietin production, and that can lead to anemia. W hat's nice is that nature's introduced a break there where that HIF-2 component only represents about 60% to 80% of your total erythropoietin. T he key thing that we established there is that if you look at the Merck clinically used dose of 120mg , that has a PD effect of basically just being on the cusp and actually inhibiting fully that maximal HIF-2-mediated production of erythropoietin. What we showed is that we can achieve that same PD effect with 20 mg of what we call casdatifan. That's our HIF-2 inhibitor. W hat we're dosing going forward is 100 mg.
We also established a very nice, linear, dose proportional pharmacokinetic profile, so we get 5x the exposure. E ssentially, we're at fivefold the PD equivalent of that Merck clinical dose, s o that's the first thing. We talked about effects on there. There were four clear-cell RCC patients. That's the target patient population at this point, for the program. There were four patients in the escalation study. They did see clinical benefit, b ut what we also talked about, and this gets to Peter's question finally, we had initiated an expansion cohort that was just clear-cell RCC patients.
W e had 30 patients that were fully enrolled and, with those 30 patients at the end of November, at the time of our earnings call, what we could already see is that from a response rate standpoint, we were essentially equivalent, if you took all 30 patients in that denominator, to what Merck in their pivotal study, what's called LITESPARK-005, had achieved at the study. T he difference is that if you look at where we stand in the maturity of the data, the Merck median time to response is about 3.8 months. 70% of our patients in this study had only had one to two scans. Y ou're basically talking six to 12 weeks on treatment. W hat we also had, in addition to those patients that had responded, double-digit patients that were already showing tumor reduction, stable disease with the opportunity to convert to response.
The second thing that's important is that in that registrational trial, there was a high level of primary progression. B asically, over 30% of the patients, which was inferior even to the control in the study, didn't reach that initial scan before they had progressed. That's something that won't change, and we've already seen a better number of that. I t's something we've had our first ad board for this program, and investigators tell us that could be an important point of differentiation.
The other key points of differentiating that we're looking at is depth of response, kinetics of response, and finally, PFS. The belzutifan, PFS is on the order of 5.5 months. W hat you should expect to see in that second half will be a very mature, dataset with about seven months of follow-up, potentially, early read on PFS, definitely mature ORR, a sense of the kinetics of response, as, as well as, what those numbers were for that primary progression. I think I probably didn't miss anything.
T here's a 50 mg, 100 mg dose, and 150 mg dose. What were you taking 100mg forwards? Just walk us through.
Yeah. R ight now, what we're focused on is that 100 mg dose. We've actually already fully enrolled the 50 mg dose. I n thinking about Project Optimus, the go-forward dose, and I described how we got to that through the pharmacodynamic analysis, our go-forward dose is 100 mg. W e're also going to have data from this 50 mg dose that'll be far less mature. I don't think that we'll present that at the medical conference, but we probably will speak to it. We'll have enough of a feel. As a matter of fact, we were already at seeing even by the time we shared the 50 mg sort of peak, we know the 50 mg dose was looking pretty good. We will be doing a 150 mg dose, and we'll likely, if that goes fine, we'll probably do a 200 mg dose as well.
I should mention, Peter, one other thing when you think about this program, because we are on a trajectory. None of these are decision-making data from a phase III standpoint. W e are on a trajectory to get a phase III study up and going by early next year, and it really represents another key point of differentiation from Merck. T he studies that we're going to be running are going to be combination studies, and they're going to be initially with a TKI.
Merck's TKI partner is lenvatinib. I think it's well recognized in the field, and certainly with investigators. It's a difficult drug to administer. It's probably one of the least favored TKIs in the field, but it is what Merck has. W e'll be using either cabozantinib, Cabo across our study. T hat's another point of differentiation. We'll be on the combination partner, much less toxic as well.
Got you. W hy the different doses? Why up to 200mg?
In anticipation of discussions with the FDA, they tend to want to see a lower dose, and they tend to want to see a higher dose. I think the Optimus concept was more focused on safety, but when you look at a drug like casdatifan, there it's really in the context of cancer about as safe a mechanism as you can get. As I said, the primary AE is on target, and that's anemia. But we want to be prepared to have discussions where we've gone above the dose that we're going forward, as well as shown something different at a dose that's lower.
Got you. What do you need to see in the dataset, whether it's in H2 to move forward? It sounds like you're already moving forward.
Let's move down up an octave in voice, and maybe Jen can jump in here and talk about what we might want to see.
My answers are a little shorter, so we'll get through it quickly. A s Terry was saying, and I do think we're at a point, just based on everything that we've seen so far, like, we have a lot of confidence in moving forward. And so we're starting to get past the point of looking at data as decision-making. And right now, we're moving very aggressively forward, towards being able to start a phase III in H1 of next year. That said, you know, we know investors are going to be looking for certain things. T he best benchmark study to look at is LITESPARK-005. So that was a phase III study that supported the approval of belzutifan in third-line clear-cell RCC. So similar setting to what we're exploring in ARC- 20.
In that study, they observed about a 21% to 22% ORR. E ven more importantly, they saw a 33% or so primary progressive disease rate. So those are patients that progressed at or before their first scan. And those are like, you know, those are your worst performing patients, and that's exactly what you don't want to see in a cancer study. And so in addition to, you know, potentially improving upon ORR, we also think there's an opportunity to improve upon primary progressive disease and to be able to bring the disease under control more quickly. And actually, when we talk to investigators, they almost point to that as like being the real Achilles' heel of belzutifan and something that can be improved upon because obviously, your goal is to try to bring the tumor growth under control as quickly as possible.
So those are probably the two things that people are going to be most focused on is, you know, does our ORR look better than that 21%, and then does the primary progressive disease rate look lower? And, you know, back to Terry's points on what we said on our earnings call, we've already said that our primary progressive disease rate is lower than what we've seen LITESPARK-005. Given that you know what that answer is going to be at your first scan, and everyone in our study has now had their initial scan, that number should not change. And so, you know, we already know that we should beat the LITESPARK-005 number. And so now, you know, we're obviously just watching the ORR very carefully.
Does that all point to a really good PFS number that could blow the 5.5 out of the water ?
Yeah. So, I mean, you know, if we see, you know, improvement in ORR, and then, you know, an improvement in that primary progressive disease rate, then that should definitely translate into an improvement in PFS. One thing that I'd also say, you know, as a nuance is, even though in ARC- 20, we are looking at that third-line patient population that belzutifan's currently approved in, we will not pursue that from a phase three study perspective. So, you know, Merck has that on their label. It's a relatively smallish market opportunity. So we're really going to focus on trying to get to that second line and potentially first line setting as quickly as possible.
So even though this is a great proof of concept study and will help us be able to tease out the potential areas of differentiation relative to belzutifan, we probably will not take it forward, in that third- line setting. And one other thing I think is important to note, you know, as you compare our dataset to LITESPARK-005 is, you know, while it's a fairly similar setting, our study patient population is a little less healthy. In LITESPARK-005, patients were eligible if they had one, two, three prior lines. So if you had more than three, you were not eligible. In our study, you need to just have more than one. And so, you know, we have patients on the study that have had six prior lines, seven prior lines, etc. About 15 patients, about half the patients have had three prior lines or more.
Perfect. The opt-in from Gilead, what triggers that? Why haven't they opted in? They seem to talk about it in very positive light.
W e've aligned with them on what the package is to inform their opt-in decision, which is a certain amount of data from ARC- 20 and some other stuff that we're doing. So we know exactly what the trigger is going to be. We have not obviously hit that trigger yet. I'd say it's going to be within the next 12 months. You know, we haven't, you know, just having very bad memories of the whole TIGIT opt-in question phase. Yeah, we probably won't provide specific guidance on when it's going to happen, but, you know, it'll be within the next 12 months.
The studies you're doing, does that go above and beyond what is written for Gilead to opt in?
Yes. ARC-20, as we talked about, you know, there's three different expansion cohorts. We're now in the combination study with Exelixis' Zanza, STELLAR-009. So there's a little information from that that they'll want to see. But, you know, we're on a path right now to generate a lot more data than what's required for the opt-in package.
We're full steam ahead. W e're excited. We hope they opt in. We think they're interested, but this is full steam ahead, in any event.
Got you. B ack to TIGIT, gastric, what does success look like for PFS, for gastric?
You want me to?
Say it.
To remind people, the study that we're talking about is a study that we call STAR- 221. It's a registrational study in upper GI cancers. It's actually one that we pre-invested in sort of at risk, and it's playing out quite well. We're probably two to three years ahead of any other competitor. W e're going to have a substantial first-to-market advantage in this. We reported data on a phase II progenitor, same patient population, same combination, same setting, at ASCO Plenary last year. W e're actually at ASCO going to have updated dataset on that. So I'll remind people that what we saw that was extremely encouraging is we're basically looking at two subpopulations or a subpopulation in that study, the high PD-L1 population and the ITT population. And we saw six-month PFS of 93% and 77%, respectively.
Now, recalling that the standard of care there, which is Nivo chemo, is a PFS that is sort of in the order of seven to eight months. Obviously, anything above that, and you can do your own math, what would be substantially better, would be something good to see. We will have that PFS at ASCO. T he study that we're talking about here is called EDGE-Gastric . We'll have that PFS. We'll have mature ORR. And I think that's going to be a very exciting dataset for us. The other thing to juxtapose with that is the registrational study, has enrolled extraordinarily fast, and we actually feel there's a good chance that that study will be fully enrolled by that time period, plus or minus a month or so.
W e could very well be sharing the ARC-7 Gastric data, mature PFS, and at the same time, instead of saying, "Oh, we're very excited about these data. We'll be starting a phase III study X months from now," to actually saying that the registrational trial corresponding to that is fully enrolled. So when you think about that, we're definitely in sort of a pre-registrational, pre-launch, pre-commercialization mode, for this program, given that the standard of care, Nivo chemo, has an OS on the order of 13 months or so. So if we're fully enrolled by, you know, the middle of this year, you can, we'll give guidance later in the year as to when that might read out, but you can do your own math as to when that might occur.
Got you. As we think about that kind of fitting into the treatment paradigm, will you have to do PD-1 status for these patients? How common is that?
How common is?
PD-1 assessment.
Yeah. I t's actually very common today. So, you know, about 80% of oncologists today test for PD-L1, and it's about 60%. It's a little bit lower in the gastroenterologist community. You know, I think it probably depends a little bit also on whether you're an academic center or community. My guess is in the community, they probably test less frequently. They just want to get patients on treatment as fast as possible and not wait for test results. And in that case, you know, they just use PD-1 plus chemo in everybody, especially since Nivo is so safe or PD-1 is so safe in that setting. You know, it makes it a little bit of a no-brainer to just give everybody PD-1 chemo.
Got you. Would that change the way you kind of think about enrolling patients or just the treatment just because you have to do a PD-1 study?
I think it'll be very similar in our situation. F irst of all, you know, just a reminder on our study design for STAR- 221, there's sort of two ways for us to win. So we have dual primary endpoints of PD-L1 high, so CPS greater than 5, and then ITT. So let's say, you know, we just hit in that PD-L1 high patient population, which we don't think is going to be the case, but, you know, let's say that is what happened. Like I said, you know, 80% of people test, so I don't think that becomes a big obstacle. You know, every center has PD-L1 testing there. I also think, you know, as you know, like PD-L1 is sort of a gradient. And a lot of people, you know, especially with gastric, have positive PD-L1.
G astric, unlike lung, is more skewed towards patients, being PD-L1 positive. So about 40% to 50% of patients are PD-L1 high, sort of the equivalent of greater than 50%. Another 40% or so are one to five. And then it's only about 15% to 20% that are less than one. So we think you could still possibly see this pattern where you're, you know, even if we don't have the broad label for ITT, that people still don't want to deal with testing. They see Dom as a really safe agent. You know, chances are their patient has probably, you know, had some PD-L1, and so might as well give them the anti-PD-1 antibody as well.
Perfect. I s there anything coming through, like Claudin- 18, etc., that's kind of changing the landscape in any way?
W e're obviously keeping a close eye on the Claudin- 18.2s, but, you know, Astellas was supposed to get their anti-Claudin- 18.2 approved in January. The approval was delayed, I think, due to CMC issues. But, there's two big obstacles as far as the Claudins. You know, one, there's a lot of toxicity, you know, particularly GI toxicity, nausea, vomiting. So we think that's going to be a big issue for clinicians, patients, especially in the community setting. I mean, everyone we talk to, like an academic center says, there's no way community centers are going to want to use this.
The second thing is it's going to require the Claudin- 18.2 testing, and it's just going to take a while for that to roll out, you know, especially beyond the academic centers. So we think those are two uphill battles for the Claudin- 18.2. So we think there'll be some usage. Like, there's probably a place for Claudin- 18.2, but we don't think of that as something that's going to take a lot of our market opportunity away.
To give you a little sense of things, the enrollment in this, as opposed to normally where, like, you might slip a little, we're probably ended up, we're going to be six plus months ahead of what our initial plan is. So there's been a huge excitement that's only, as we generated that initial, those initial data from ARC-7 gastric, only increased. The second piece that I would say is that when you think about the Fc-silent anti-TIGIT and the fact that when you combine it, especially now, like with chemo on top of anti-PD-1, it's essentially the safety profile so benign that when, if you were to just look at the data, you'd think you were just getting the chemo. So, you know, or chemo plus anti-PD-1, so it brings no additional baggage.
T he idea of being able to come in with something that enhances efficacy in cancer that doesn't bring any additional toxicological baggage is very exciting to feel.
As we kind of think about that phase III, how similar are the phase II patients versus III? Is there any kind of variables we should be considering?
Very, I mean, as similar as you could get. Yeah. So there's not even something I could think about on the top of my head that would make those patient populations different.
Okay. Perfect. Thank you. And then CD73, I guess, what does the registration or trial look like? And we'll start there.
T he registration or trial, we'll be talking to the FDA shortly about that, but our plan is that it's going to be a very simple design. You're going to have Gem/Abraxane, which is a standard of care, plus our CD73 inhibitor. CD73, that enzyme, its inhibition will prevent the formation of adenosine, which is highly immunosuppressive. That study would look at then just the triplet of Gem/Abraxane, quemliclustat versus Gem/Abraxane. I should point out in that context at AACR, so I'll let me just remind people what we did in advance of that trial or what led to that trial was a study of quemliclustat, plus Gem/Abraxane, plus or minus our anti-PD-1 where we were clearly able to show that the anti-PD-1 did not do anything.
We showed a profound effect, whether it was compared to historical Gem/Abraxane or a synthetic control arm, that we generated. So we saw an OS of 15.7 months where, depending on the dataset you were to look at, you would say that Gem/Abraxane is going to be somewhere between nine and 11 months, OS. What's very interesting is this field starts to blow open, and we think that we'll have three datasets this year that read on it. I n addition to ARC- 8, there's a study that Genentech ran as part of their Morpheus platform in pancreatic cancer where instead of Gem/Abraxane, quemliclustat, they looked at Gem/Abraxane etrumadenant, which is our adenosine receptor inhibitor. So basically, you're in one case, you're blocking the actions of adenosine. In one case, you're blocking the formation of adenosine. Those data will be shared at AACR.
They do have a Gem/Abraxane control arm there. W hat you're going to see is a very similar profile with that very prolonged OS while you're not even having much effect on the ORR. So same phenotype of that immunotherapy-induced prolonged OS. The final piece to that, what would be another dataset that we have that we have submitted for abstract, hope to have it out by the middle of the year, is a study where we're looking in third-line colorectal cancer with chemotherapy, Bev, and etrumadenant, again, our A2 receptor antagonist, versus regorafenib, the standard of care there.
A gain, what you're going to see is a confirmatory dataset when you take and ameliorate the effects of adenosine when you're using immunogenic chemotherapy that you can get very profound effects on OS. So you'll have three datasets that are different in very important settings where you have immunogenic chemotherapy that really illustrate the potential for adenosine blockade.
Got you. Fo r 73, what's the real bar? Is it what you've previously seen in the phase I, like the 15-month kind of OS?
Yeah. So it's sort of saying kind of like nine to 11 months-ish, in pink. I mean, obviously, we'll have the control arm in the study. But if you look at, you know, even the more contemporary studies that looked at Gem/Abraxane as the control arm or the experimental arm, you know, it was about nine to 11 months.
Great. Perfect. Thank you. We've hit the end of time. So thank you so much.