Arcus Biosciences, Inc. (RCUS) Earnings Call Transcript & Summary

Good afternoon, everyone. Thanks so much for joining us. I'm Salveen Richter, I cover the biotechnology sector at Goldman Sachs. Really pleased to have with us the Arcus team. So with us, we have Terry Rosen, CEO; and Dimitry Nuyten, CMO.

And in the audience, Jen Jarrett.

So maybe to start here, ASCO, you recently presented updated data at ASCO. Maybe walk through kind of the key takeaways from that data set and how we should put that in context of kind of the overall outlook here...

No, that's an awesome question. So just reminding everybody, specifically, Salveen is talking about our ARC-7 data set which is looking at the first Fc-silent the anti-TIGIT domvanalimab. And we were actually -- I think it was a very awesome data set. It was presented by Melissa Johnson. She did a 6-minute presentation because it was part of the ASCO preliminary format, and then we had a follow-up data set that we put out. And so I'll hit the real highlights. The first key thing is in this interim analysis, the molecule continued to perform as it had in the past, most evidenced by the Kaplan-Meier curve. So if you look, the hazard ratio, when you compare the doublet containing anti-TIGIT versus simply anti-PD-1, there was a 0.67 hazard ratio, so roughly a 33% reduction in events occurring. The other thing that emerged with this data set is that we start to really get a good feel for durability. And I'll point to a couple end points there. So first off, the DOR, very substantial, although immature, but basically a doubling of what we see with the anti-PD-L1 alone with the dom-containing doublet. The other thing that really stood out, if you look at -- now that when you can look at patients that are out in, let's say, the 18- to 24-month period and you look at those patients still on therapy, dramatically more remaining on the doublet than the singlet, which really bodes well as you start to think about what's the approvable endpoint, what's the meaningful thing for patients, which is overall survival. On the other end of sort of the book end of the late versus the early, one of the more profound observation -- this was pointed out, we had an event and Tom Marron, who's one of the KOLs in this field, really emphasized this point. So if you look at anti-PD-1 therapy alone, you see about 25% early progressors. And in the dom-containing arm, we saw that with zim alone, but the dom-containing arm, again, dramatically less early progressors. The other point we would make is that if you looked at spider plots, you would see that a number of the stable disease patients were on the stable disease before they actually progress. And in fact, we have one pretty remarkable patient where it was after 21 months, they've been stable disease, and they had a pretty deep response. And then, a couple of weeks after the data cutoff, that actually confirmed. And so that differential in the Kaplan-Meier curve of about 20% is really meaningful, and from a physician standpoint, almost gets to be like "you're getting that much more of a curative effect." Beyond just the clinical observations, one of the things that we highlighted was we're doing a very extensive biomarker analysis, but one piece that we felt we could talk about early because we felt it was important from being a smoking gun in terms of the mechanism is that it's known -- and this was known even before we did our study, that high CD155, which is the ligand for TIGIT, is a very negative prognostic for this particular setting when you're on anti-PD-1 therapy alone. So let me just -- we will quickly talk about a couple of the key aspects of the biology. So CD155 is the ligand for TIGIT. It's highly expressed on tumor cells. When CD155 engages TIGIT it causes immune cells to essentially become quiescent. So when you disrupt that interaction, that CD155 can now bind to another receptor on immune cells of CD226. CD226 is immune-activating. So essentially, you convert a foot on the brake of the immune system to the foot on the gas pedal of the immune system. It's very well-established biology. It's the reason why people got into anti-TIGIT. It's the reason why it's such a safe mechanism because that CD155 is so highly expressed in the tumor environment so that when you get that switch over, you don't get a whole lot of autoimmunity type of side effects, and that's been borne out across studies. So the reason we pointed out those data is that, in our study, we actually validated what had been reported in that literature before. So if you look at the patients that were on anti-PD-1 alone, those who had high CD155 above the median versus those who were below the median, the hazard ratio was 0.38. So very meaningful difference. You have a very negative prognostic if you have high CD155 and you get anti-PD-1. But what we also showed is that in the dom+zim arm, that was reversed, and you could essentially take those high CD155 patients and turn them back into patients that would look like they had low CD155. So we feel that's a very important observation for the field because it's a smoking gun that links a mechanism in a molecular endpoint to the clinical observations. The final point that I would highlight about our data set is you continue to see really attractive safety profile. So essentially, dom has created something that's pretty unique in our field where you're creating a combination therapy, adding a new agent on top of the standard care, getting improvement in efficacy with no added safety base. That's due to both the mechanism but also the fact that we have this Fc-silent anti-TIGIT. So what's clearly emerging across the field is that those antibodies that are anti-TIGIT that are Fc-enabled cause depletion of T regulatory cells. So this isn't just theoretical. It's been shown multiple times by multiple companies. What you also see associated with that is higher adverse event, particularly immune AEs, things like rash, pruritus. Genentech even reported pancreatitis that all could be Treg-linked. So we feel that the Fc-silent anti-TIGIT, which would -- there's some questions about for some period of time, is clearly emerging to have safety advantages compared to the IgG1 wild-type antibodies. Final thing I'll just highlight is that in addition to our own data set, there were a couple of other important data sets, I think, important for the field. Genentech presented data in HCC, clearly, a meaningful signal, whether or not you want to ascribe quantitative truth to it, definitely from a qualitative sense, it's another very strong signal. BeiGene presented data in the GI setting. Interestingly enough, that's with chemo, and we also see there a high degree of AEs, which again points to that Fc-enabled anti-TIGIT. And finally, a company that's been developing anti-TIGIT a little less under -- a little more under the radar, and they only had an abstract, not a full presentation, but Innovent shared data in the same setting, high PD-L1 non-small cell lung, with a very similar hazard ratio to what we see. So I think the field as a whole, KOLs, investigators, those who have been working in this field, clearly came away from ASCO with -- clearly seeing that the signal in anti-TIGIT is looking very real.

At your last interim analysis, we did see the median PFS pull back a bit by about 3 months. Put that in context for us with regard to your confidence and particularly as you look to OS being kind of the more meaningful measure here?

Sure. So Salveen said, you definitely saw that point estimate for median PFS drop, but what we come back to is really PFS, what you want to think about is the hazard ratio and you want to think about those Kaplan-Meier curves. If you look at the Kaplan-Meier curves from interim analysis 4 and interim analysis 5, they're indistinguishable. The other point that we made is we did roll forward. So we took the full patient population that had been enrolled as of IA4 and then just roll those forward as if that was last patient in, into the same analysis, the data looked very similar to IA4. So between IA4 and the most recent analysis, you have 2 things happen. Additional patients come out of each arm that affected the contemporary PFS. But again, you kept the Kaplan-Meier curves looking similar. And then everything in terms of OS which, by the way, I should remind people, from -- the FDA is black-and-white clear that the only approvable endpoint in frontline non-small cell lung cancer with an I-O therapy is OS. And all of the data, the DOR, those patients still on study in the doublet versus the monotherapy, point to what could be a very strong signal when it comes to OS. And we think the data are pointing towards a very positive outcome on the OS front. And as you know, with I/O therapies, and this is what we're seeing in terms of that tail, that tends to be the strongest part where you get -- once patients have responded, that I/O tends to really prolong their survival. And that's at least as we sit here now, how those -- if you were to look at the spider plots, for example, what you would see.

Great. Clearly, you'll have to see the data, but tell us how you're thinking internally about the impact to your trials on the forward based on Roche's study and the outcome in the second half of the year, if it's positive or if it's negative?

Yes. So we think that like in terms of positive/negative, it's less binary than that. We think that based on everything they've said and what we know and they've talked about numerical differences between the arms that we expect it to be positive. But if it isn't positive, if -- where positive is measured as approvable, we think it more likely than not -- and we'll have to see the data, would have something to do with their design or their statistics plan, et cetera. But we think that one thing about SKY-1, it's been a topic for some time. We think probably overstated because for a while it was sort of the only game in town. We think there's so much data out there that not only is public, but when Merck clearly has a lot of data. BeiGene has data, we have data, Genentech had data. They showed the HCC data which was a pleasant surprise to their community. So we think there's no doubt that anti-TIGIT is going to be important. It's going to be important in this setting. And so we're -- we will look with the same anticipation as everybody else towards SKY-1 but we really don't see that it's likely going to affect anything that we do going forward. Obviously, we'd rather see a positive signal, and it will be good for the field than something that's perceived as negative. But again, we don't think it's going to be negative, positive. We think it's going to be the degree of positive and whether it actually hits statistical significance.

You're going to have a number of readouts coming in the second half of this year and next year for TIGIT, including in other tissue targets, right, or like gastric and so forth. Help us understand what you expect to learn from each of these different studies and the outlook here? And also, just based on what you mentioned earlier, is liver going to be a focus on the forward?

Do you want to comment on the...

Yes, and I can start with the different readouts. So let's start with gastric cancer. So on gastric cancer, we have the Phase III trial going on STAR221. It's a Phase III trial where the standard of care is nivolumab chemotherapy, FOLFOX or CapeOX and we are randomizing patients between the standard of care or dom+zim with FOLFOX or CapeOX, the same human therapy backbone. The trial was started because we thought it was a really great opportunity. There's obviously clinical data to support it, but it was relatively limited at the time. So we felt this is a field where we can move fast and not only bring a meaningful trial to patients but actually be the first ones in this setting. And in order to enable the trial for, let's say, regulators, for patients, for physicians, while the trial startup was ongoing, with a lot of preparations, we started a Phase II trial that is now called EDGE-Gastric. And in that trial, we examined the same regimen. So FOLFOX with dom and zim. That's done in a platform setting. That means there is a standard of care control arm in there, meaning it's zim plus FOLFOX, so we can use that for contribution of components, but it's also a rolling control arm in a platform study. And that particular, or let's say, the first-line cohort, the dom+zim+FOLFOX, we are planning to present data later this year. So that will be the first presentation with the safety and response rate data. It's probably too early at the time for PFS data. And that is, as I said, it's basically a data set that we use for health authorities, for physicians, doing investigator meetings to make sure people see data from that trial, right? Because if you start treating patients in the Phase III, you can't look at the data, right? That's -- it's completely blinded and you can't look at it until the trial reads out. And that same trial, the EDGE-Gastric, also has a second-line cohort in there for patients that either have or have not seen a checkpoint inhibitor. So it gives us the opportunity to explore different regimens. At the same front, where we are looking at the mid-stage development platform studies, Phase II studies, we also have 2 Phase II studies in lung cancer, VELOCITY Lung run by our partner, Gilead, and then EDGE-Lung run by us. And those are examining chemotherapy-based combinations with, let's say, dom and zim, but also with our CD73 programs. And those trials are really set up with the same purpose. They generate early signals of efficacy. They have flexibility to put in new combinations. And while we are thinking about next opportunities, we are generating the safety and early efficacy data in order to move fast if we want to move into a different setting.

In the VELOCITY trial, are you stratifying by PD-L1 status?

So we have different substudies in the trial and different substudies, they look at either all comers or they look at certain parts of, let's say, the PD-L1 spectrum. In VELOCITY right now, it's all comers. In EDGE-Lung, we do have a cohort with PD-L1 high separate where we only explore immunotherapy combinations. And in the other cohorts in both trials, it's the chemotherapy backbone with experimental agents on top.

Outside of TIGIT, you also have a pretty vast portfolio and you're also going to have a lot of reads coming there. Maybe just help us understand what you're most excited about in terms of some of these programs.

Awesome. Love that question. So first off, let me talk about HIF-2. So we just completed -- so HIF-2, to remind people, it's transcription factor. There's an approved drug, belzutifan that Merck has, that they acquired from Peloton. That has well documented, well-described limitation. And that limitation is that it's associated with its PK profile. So it has absorption limited pharmacokinetic basically above the 120-milligram dose that's clinically used for belzutifan. You can't get higher exposure. So it's clear, as you think about clear cell renal cell carcinoma, we think there's a high probability that Merck is going to get approval there, but we also feel like there's a high probability they're leaving activity on the table. We've been in healthy volunteers, and we just completed the dose escalation phase of a study in all-comer patients, and that's with the 100-milligram dose. That's cleared, and we're starting a dose expansion phase of that study. And we're also -- we'll be starting a Phase II study in combination with another agent that we'll talk about later this year that we think actually gives us an opportunity for best-of-class combination. But to describe what we already know and then what we'll be sharing later this year in terms of the data set, so that 100-milligram dose of AB521, as our molecule is called, achieved essentially a three- and probably closer to fourfold the activity that you get with that 120-milligram dose of belzutifan. At the same time, there's a peripheral biomarker for that activity, and it has to do with -- unrelated to the molecular mechanism in the tumor, it's associated with erythropoietin production from the healthy kidney. And what you can see is that we would achieve a roughly at 20- to 30-milligram dose, what Merck would achieve with the 120-milligram dose. The good thing about that marker is that it essentially saturates at that level. So in going higher doses, you can't -- you don't suppress any more of that erythropoietin production. But on the other hand, you're going to be hitting the tumor much harder. So there's a number of reasons based upon the historical data. Most notably, if you were to look at a second-line RCC data set that Merck has in about 30 patients, essentially almost every patient had some tumor reduction, but still you only had a 20-some-odd percent response rate, also pretty substantial time in terms of the kinetics to see that response. We think it's pretty clear that there's an opportunity that activity was left on the table. And so we're very excited about having the opportunity to go with a molecule that's unquestionably better. And so we'll have the Phase II data that when we started the Phase II study, we'll be sharing the dose escalation data prior to the end of this year. So you'll see that PK/PD and safety that you can compare to the Merck molecule. And we're pushing really hard to be in Phase III by the end of next year and looking at both the doublet combination as well as a monotherapy. The other thing that will be important readouts between now and the end of the year is if you think about our molecules in the adenosine ATP pathway, etruma, the A2 receptor blocker, and quemli, the CD73 inhibitor, we'll have -- in addition to the ARC-7 study, we have ARC-6, ARC-9 and ARC-8. As I stated that there would be etruma in both prostate and colorectal and ARC-8 is in frontline pancreatic cancer. All of those are randomized studies. They all have had data where we've seen an interesting signal. The one that we're probably most excited as we sit here today is our frontline pancreatic cancer trial. This was a study -- just I'll remind briefly what we're looking at here, we did a dose escalation, dose expansion where we saw an early signal. We started a randomized component where what we're looking at is gem+abraxane, the standard of care plus quemli, our CD73 inhibitor, then plus or minus zimberelimab, anti-PD-1. So really, what we're sorting out here is whether anti-PD-1 brings anything to the table. There have been a couple of studies with gem+abraxane using nivo or pembro that didn't show any benefit. In our case, we're looking in the context of CD73. So that was an outstanding question. Mid -- so we ran this randomized study. Mid the randomized study, we pointed out that we were not seeing a difference suggesting that the anti-PD-1 was doing anything. And it wasn't clear how exactly things would play out. And we said that the next data cut that we would report on would be when we had OS. So that's very near. I think it's going to be a very interesting data set. That's something we're excited about and something we'll be sharing prior to the end of this year. So I think between all of those studies, that will also define a lot about how our investment is going to look in both of those molecules going forward. So a lot of readouts, and I think a lot of interesting readouts between now and the end of the year.

And I think you mentioned on the last earnings call that with ARC-8, the event rate has been a little bit slower than expected. So -- how should we think about this in terms of hitting the 6-month PFS goal?

So I think -- so here's what I would say is out there, we described that the last patient in was the end of 2021, and we're looking at OS. And so if you think about what you would expect with gem+abraxane, there's been gem+abraxane's sort of been in numerous studies as a control arm. Of course, there was a Phase III approval trial. So in the Phase III trial, the OS was about 8.5 months. There have been other studies. So if you were to ask people what would you expect with contemporary gem+abraxane, they'd probably say 10 to 12 months. And there's been other recent readouts where it showed up 9.5 months OS. And so if you consider that last patient in was the end of 2021 and we've been like waiting to get to a good, meaningful, mature OS, it's a pretty positive, I would say, that you're waiting. So it's a good sign.

Great. You also expanded your partnership with Gilead recently to include immunology targets, I believe, 4 assets here. Could you just talk, even though it's very early, just discuss your work here and what investors should take note of with regard to this partnership?

Yes. So I think it's -- our entire Gilead relationship from day 1 -- and I think you see this playing out, has been -- our purpose has been to build a long-term independent company. And something -- when we first mentioned that, that people have questions, okay, let's just see how that plays out. I think the first part of that is it's enabled us to really build, not only a really strong portfolio molecules, but a portfolio molecules that are competing in major, major markets with major need with major competitors. And look at us, we're running 4 registrational trials that are with huge -- you're competing with Merck, you're competing with Roche, Genentech. We have a mid-stage pipeline with things like our HIF-2 inhibitor, the adenosine modulators. We have early molecules. We have an AXL inhibitor that will be going into the clinic soon, CD39 antibody that started. So we've built a pipeline -- but we have a very much -- something that I think is unusual. We have a discovery group that is capable of creating something from nothing. So it's not like we're just running around looking for leftovers. We're basically looking at innovative targets with potentially best-in-class molecules. At the same time, we've built a very substantial development organization. So if you look at what Arcus does, like we have a very strong small molecule expertise. But also from a biology standpoint, our group is primarily immunology. So despite the fact that we do so much in oncology, a substantial portion of that, because it's immuno-oncology, is studying immunology. So we've always had the intent to get into immunology. The dynamics of the relationship with Genentech, as many of you are aware, it's an extraordinary collaboration. It's not just tossing money over the transom, it's not just tossing molecules over the transom. Gilead has hundreds of people working on our programs. We're well capitalized, largely due to the relationship with them. But it is sort of a soup-to-nuts type of collaboration. And as part of the discovery efforts, very organically, our discovery units, since the arrival of Flavius Martin as the Head of Discovery, there's been very good dynamics both at the oncology level and inflamm level. Also keep in mind that Gilead has built out a very substantial inflammation development group under the leadership of Mark Genovese. So there's a very good symbiosis there where you can imagine they have a development organization. We want to get into oncology -- into immunology in a bigger way. It enables us to do that without having to distract the development organization. And so basically, Gilead enhanced our resourcing, the inflammation end. We have a strong immunology group. We picked the targets together. Jen's pointed out a couple of groups. It enables us to be very aligned on the targets from day 1. So we're working together from the earliest literally basic discovery into the early phases of drug discovery and then ultimately clinically together. So strategically, it really enables us to, in a very continuous way, continue with our oncology efforts at all levels from discovery, Phase III development but to weave in these inflammation and immunology programs in a capital-efficient and operationally efficient way. So we're really excited about that. And we think it's, again, another aspect of building a sustainable drug discovery organization where we can start things and stop things based on their merit, not simply because you have to keep pushing something because it's the only thing you have.

Great. Well, with that, maybe we'll open it up to the audience for any questions. Can we just get a...

So for the PDAC study, I think before you mentioned OR of 40% is what you think is meaningful, PFS of greater than 6 months. It sounds like the OS is going to be probably, I don't know, north of -- I mean, like 15 months would be meaningful in terms of benefit over that 10 to 12. Is that fair to say? And then what about you have both the quad and quemli+chemo. Have you seen anything different between those 2 subgroups?

So what we'll say, and I'm comfortable saying this now, is that the anti-PD-1 is not enhancing the activity. Your points on like what would be a good OS, certainly, that would be fair to say. I think any KOL in the field would say something above that would be appealing, and we obviously don't have the full data set yet, but something along those lines would be exciting. So we shared an update when we looked kind of interim. We haven't done a formal data analysis since that one, but we will be soon.

Just one last one here at the...

Sure. There's a couple.

Yes. We're keeping you on schedule. There's the Phase II EDGE platform study in lung cancer. Just remind us how that kind of fits in your overall strategy for lung?

Yes. So the way we see the platform studies is like lung cancer, obviously, is one of the tumor types we're very focused on. And the platform study fits in, in the way that we have the ability to keep on putting new arms into the study with a running standard of care control. So we're able to say -- to benchmark early signs of efficacy. We're able to generate safety data that typically is needed when you want to start a Phase III trial. So for different combinations that we're interested in, we can basically amend it into the trial, have that flexibility and then wait for the data to mature while we are making decisions about where the combination can go. For example, if we think about, let's say, the CD73 pathway, we've done the experiment in ARC-7, but that's an I/O-only setting and any tumor was added to dom and zim. But I would say, based on the research hypothesis, exploring adenosine-targeting in the context of chemotherapy is a very different and I personally think more attractive setting to explore that. So that's why we have a number of arms in the 2 platform studies that we are combining, let's say, the adenosine modulation on top of chemotherapy with zim and sometimes with dom as well. So then at some point, we have the safety data and efficacy data to make decisions. And especially when we want to move into earlier lines, for example, a Stage 3 setting in lung or, let's say, a periadjuvant setting, meaning patients with curable lung cancer, it is even more critical to have some safety data and early efficacy data in order to start mature trials. So it's really a very efficient way of being able to screen for efficacy. And also if we have new agents that we are -- let's say, that are coming out of the pipeline, they can also go in those trials. So these trials can run for a very long time.

Great. With that, thank you so much. Really appreciate the time, Terry and Dimitry.

Thanks, Salveen. Thanks, everybody.

Thank you.

Appreciate it.