All right, let's kick off our next session with me, Terry.

Welcome. Thank you. Very exciting times this year. And before we go through the question, which I have a lot, I'm going to turn it to you guys for opening remarks.

So let me just introduce Arcus to those of you out here who might not be as familiar with us. I'll be brief. The company was founded about 10 years ago. We built ourselves around a very strong R&D capability, particularly discovery. That's reflected in the type of programs we choose. We focus on programs, particularly with small molecules, where there's been great validation, but perhaps not a great molecule yet. And I think that leads me to CasDatafan, which is clearly our most valuable program right now. And that kind of fits that paradigm where we think we've demonstrated unequivocally that we have a better molecule than the one other molecule that's out there. That's Belzutafan. It's Merck molecule. It's validated the program clinically. It's validated commercially, and we think we've shown a lot of data, and we have a lot more coming later this year that show not only do we have a better molecule, but it enables a better development strategy. So as we sit here now, we look at CasDatafan as a $5 to $10 billion opportunity commercially. Finally, while we talk less about it, and, you know, next year, for example, 80% of our resources will be invested in casdatafin development, we've somewhat quietly built a very strong immunology and inflammation portfolio of molecules. And they all fit that same paradigm where there's strong validation. In this case, a lot of it coming from biologicals where you have double-digit billion-dollar drugs even, but where there's something that could be addressed that might be a disadvantage of those programs with a small molecule. And so, for example, we have an X2 antagonist that's heading into humans imminently. We'll have proof of concept for that probably early next year. We'll have PK data later this year. We have a TNF receptor one antagonist in development. We have STAT6 inhibitors, CCR6, small molecule inhibitors as well. So a very robust, you know, I put it up there against any company, large or small, insofar as an early INI portfolio. So that's moving along in parallel to the CasDataFan development. So, final point I would just make is our cash position is strong. We are funded into the second half of 2028 with everything we'll talk about today, and that gets us onto the other side of data from what's the first registrational trial that's ongoing and enrolling right now, forecast data fan in the second line. So, I'll stop there and I'm sure to get to your questions.

Yeah. A lot going on in the company. There's a lot of readouts. Can you walk us through the timing for some of these readouts, including the PDAC, what's going on in the INI program? I know you guys started that phase three, the peak one study, what's the status for that? So maybe just kind of give us the status on these programs and the timing for the readout.

Sure. So you mentioned PDAC. So we have an ongoing, fully enrolled as of September of 2025, a study in pancreatic cancer with our small molecule CD73 inhibitor QEMLI. And we're anticipating that to read out with data in the earlier part of 2027. That's purely event-driven And so that's running QEMLI-GEMMA-BRAXANE Versus GEMMA-BRAXANE You'll see when we talk about CasDatafan We like to go on top of standard of care Very clean study Very clean data readout And that was driven by very strong phase 2 data And so that's a 2027 readout CasDatafan, the first registrational study We started enrolling that basically at the beginning of this year. We expect it, and we're on track to be fully enrolled by the end of this year. When you think about that's CasDatafan plus Cabozantinib versus Cabozantinib, when you think about Cabo is having a PFS on the order of 10 months or so, while we haven't given guidance to the readout, Clearly that just doing the mathematics takes you, if it's fully enrolled at the end of this year, it takes you probably into the earlier part of 2028 for data. In terms of the I&I portfolio, the first program heading into the clinic, which is imminent, is our X2 antagonist. We'll generate PK data in healthy volunteers quickly. We'll do a cabinet challenge. We'll have those data likely by early next year. And with all the other data going on in the field, proof of concept, you know, we'll likely be entering a urticaria study, catching up to others because we kind of are able to skip a step and go directly to the real study there. And then there's a steady cadence of INDs that will flow in that immunology program. All those molecules I mentioned, STAT6, CCR6, TNF receptor 1, are IND candidates in the next 6 to 12-plus months. I'll also just note on timing because it's probably the most important study that we'll be doing. We anticipate getting our frontline first registrational study in kidney cancer with castataphan up and going by the end of this year. And that'll be looking at castataphan anti-PD-1 IPI versus ipinevo. And I'll just remind people that in the frontline, that's the most commonly used regimen right now, ipinevo. and we'll talk a little bit more about our plans and the data we're generating in support of that as well.

Okay, yeah, a lot of questions on that and others. So the big data readout is obviously the one coming mid-year, I think you guys mentioned, could be more closer to ESMO timing. Maybe just give us an idea of, you know, obviously that CAS-CABL combination, that update to that data, we're going to get that. Besides that, what other cast data are we going to see along the different lines of treatment? How many patients should we expect to see with these various updates?

Great. So I think this is probably the most important question. So in terms of data flow for this year, we've recognized basically with the failure of Belzutafan and LightSpark 012 the front line is there for the taking, and we think it's with a great drug and also a great development strategy. And so we have ongoing studies or imminently starting studies that read on all lines of therapy. And our strategy is actually very simple. So we want to make sure that every patient can experience and have the opportunity to get CasDatafan, and basically that it becomes a backbone therapy across all lines of therapy for clear cell RCC. So let me just go in that, and then I'll talk about the data that we'll support. In the front line, if you want to think about the brass ring in this whole area, that's thinking about ipinevo, which is used in about 35% of front line patients, It has a limitation that's associated with its rate of primary progression, which is about 20%. If not for that, it would be used more. So we think with what we've already shown with CasDatafan, even on top of anti-PD-1, we can reduce that rate of primary progression. And we think that we can take that ipinevo from 35 to roughly 50-plus percent of the market. Now, with that said, the rest of that front line is fragmented so that other 60-some-odd percent is broken into multiple TKI-inclusive regimens. So while individually they're smaller and fragmented as a whole, it's a big opportunity. And we recognized, particularly as we had ad board discussions with both U.S. and European advisory boards, that we should have a TKI inclusive regimen available in addition to that IOIO cast. So that covers the first line. The second line, as you mentioned, very straightforward. The most commonly used, the largest part of that market is monotherapy cabozantinib, and we have an ongoing study that's looking at cabozantinib plus CasDatafan. Finally, in the later lines, we're going to be looking at another TKI. We'll announce that shortly, and we'll be combining that with CasDatafan, running a randomized study versus the TKI alone. And importantly, we'll be doing that in belzutafam experienced patients. So we're killing a couple birds with two stones, generating the data to support the registrational study, but also showing that castatafam works. And finally, and this will be, you know, we haven't talked about this much, but I want to comment on this now, and we'll be talking about it more as the year goes along. So we've really used our monotherapy single-aging late-line study to demonstrate that dramatic difference between castatafan and belzutafan, and obviously you don't have to parse out other agents in the combination. And so we're reaching a point where we should have OS data later this year as well. So I'll remind everybody that while in ClearCell RCC, the provable endpoint is PFS, it's been thought that a dramatic differentiation could be an OS benefit. And we feel that later this year, the OS will be mature enough from that late line to demonstrate a benefit that we think will read through all lines of therapy. So what will those data sets look like? Okay, so first of all, we have the 120 patients of monotherapy data that will update not only all the other parameters, but a first look at OS. In the front line, we have an ongoing cohort that's generating the safety data that will support the phase three study. That's looking at CasDatafam plus our anti-PD-1 Zim plus IPI. We've already spoken with the FDA We know the amount of safety data that we'll need Keep in mind when people talk about ipinevo as a regimen That's actually you get four cycles of ipinevo And then you get nevo So it's 12 weeks of ipinevo therapy So we'll have double digit patients later this year That have had ipi anti-PD-1 casdatafan that will form that safety component. Now, we've also been running a 30-patient, 31-patient cohort that looks at anti-PD-1 plus CasDatafan. We reported in January of this year that there was only a 7% rate of primary progression. Keep in mind that Keytruda alone in this setting would show 25% to 30% rate of primary progression. Ipenevo shows maybe 18%, 19%, 20%. So with castatafan, anti-PD-1 alone, we're already addressing what is believed to be the most limiting factor with the use of that Ipenevo. So we'll show not only the rate of primary progression data, but you'll get a good look with probably about a nine months of median follow-up of the waterfall and the early ORR, and you'll be able to squeeze together the IPI-ANTI-PD-1-CAS plus the anti-PD-1-CAS, and that basically looks like the therapy that we think is most important. And then we'll have up and running our study with that late-line cast data fan plus other TKI, And then we'll have roughly 40-some-odd patient data set that's looking at CasDatafan plus Cabo in the second-line setting. We may even have early data on OS for that. We'll certainly have Kaplan-Meier curves for that as well on PFS. And we're going to make sure that we compare to all the other data sets that are out there, whether they be Bell's Cabo, Cabo alone, or Bell's Lemba. So we think we'll be able to compare across lines of therapy in cohorts ranging from 30 to 40 to the 120 patients that have late-line, so very large data set of castataphan later this year.

So that's a lot. So let me just kind of list it to make sure I capture everything. So obviously, I think the cast couple in that second line plus, I think you mentioned about 40 patients. So that's the update from the previous brochure that you guys already gave. And then the mono update as well to the 120, 120 making that later line, second line setting for mono.

Third line plus.

Third line plus for CAS. And then you also mentioned there's two, the frontline regimens, the PD-1 plus CAS. And how many patients should we...

That's just over 30.

About 30 patients. And then there's another CAS, PD-1, and EP also in the front line. And how many I want?

That'll be about 30.

Yeah, I mean, I think so. We'll have the 31 patients for the CAS, ZIM. The median follow-up will probably be in the 8- to 10-month range. So we'll have, like, an early look at efficacy there. Whereas, like Terry was saying, the CAS, ZIM, IPI cohort will have double-digit number of patients that will have made it through the 12 weeks with the IPI-containing portion. So that's going to be mostly a read on the ability to tolerate that triplet. The safety, okay.

And then I remember in the past you also mentioned a monotherapy cohort in the front line for that favorable risk group. Is that also going to be part of it?

We have that whether or not we update on it. We're trying to be very focused on what we're doing insofar as the data we're generating that supports the registrational studies. So we have those data. We'll probably comment on them. But as you said, it sounds like a lot. We're trying to make it like simple, not confusing instance that those cohorts essentially read on rates of primary progression, which look really good, but they're not really adding to the data set that support those registrational studies. So whether we mix that in or do it at a different point is TBD.

And then you guys also mentioned that there was another TKI-free regimen that you were exploring. And then later on, we saw this PR on this bispecific from Bristol. Is that the one that you guys are – or is there another one?

So what Rich is referring to is we just announced a collaboration with BMS where basically that BMS BioNTech, Bispecific, VEGF, PD-L1, Bispecific, we're combining two arms of their platform study that BMS is executing. We're very excited about that. I think it's a high priority for them. That is another part of our strategy. It's very compartmentalized, so nothing else we're doing is dependent on that, but that is going to be starting pretty soon. I think Bob kind of has described that nicely, is it's kind of in between TKI sparing and IOIO where you have that VEGF component there. I should also say, though, that's part of a strategy in of itself. So it's a sub-strategy. So you'll see us announcing additional collaborations. We've set expectations that we would have more than one. And they're basically going to cover three different important aspects to us. So one, we're working with a partner, for example, like BMS. They have such strong experience, expertise, and commitment in ClearSell RCC. So for a number of reasons, they're a great partner. They've got their platform study up and running. There'll be another study where you'll see that we also weave in an HCC component, which is, again, a very smart place for those bispecifics to go. It's a very smart place for CasDataFan to go. And then there'll be a third where we'll actually be the one executing the study. So while we value the strategic elements of these collaborations, we also like the idea that within ARC-20, we know we'll get one done very fast. And if one of these other partners can beat us, great, but we feel like we'll execute very efficiently. So we're covering strategic partners, HCC as well as RCC, a couple of interesting combinations, and then a study where we all executed ourselves. I see. Okay.

Do you believe that there's still a debate if CAS is the BELF compared to BELF? Because one argument is that in the mono and the combo data that you guys have generated so far, it's still early, right? small sample sizes, there's no control arm. What is giving you confidence that CATS is better?

Yeah, so I don't think there's much of a debate, and certainly when we talk to people, that's not something out of all the questions that people raise. And I'll explain why. So first of all, we have such a huge amount of data as single agent. So the 120-patient monotherapy, you know, that's actually a fairly large end. Keep in mind that involves, you know, over 40 sites. Then we've got the multiple different settings, combinations, instead of where we generated data. I think the most compelling data, though, so if you first just take the clinical efficacy data, you know, we're so much better than Belzutafan on every endpoint. whether it's ORR, PFS, our durability, our rate of primary progression. But to me, the thing, and we have, you know, within a very short period of time, you're going to see a publication in this, you know, high-impact journal that there is that ties together the basic biology and science of HIF2 and CasDatafan, the biomarker data and the clinical efficacy. So that will all be out there. But the piece that's intrinsic in that that gives the most confidence, and to me it enhances that confidence by orders of magnitude, is that you've got the biomarker correlative data from monotherapy patients that shows dramatic. And I state that, I'll emphasize with all caps, dramatic differentiation from belzutafan. So if you just saw those data alone, you would think these two molecules absolutely have to be different. Now keep in mind, everything's rooted in this. This is highly reproducible and has nothing to do with patient population. you have a dramatic PK-PD difference between the molecules, and that's not something that has to do with patient demographics. We did our initial work even in healthy volunteers. But then when you look at the biomarker data, and that's very well defined, that's suppression of erythropoietin production, Not only do you see a deeper inhibition by Casdatafan than you do with Belzutafan, but the time course is what's really dramatic. And this is Merck's own data and our data, and these are, again, not sort of dependent on patient characteristics, or you don't need to randomize to do this. You go do another 50,000 patients, and they're still going to look the same. And then I think the piece that not only do we see that dramatic difference in durability of effect, but we've also demonstrated that the ability to suppress erythropoietin production is correlative with clinical outcome. So if you combine all those factors, the fact that you're seeing better clinical efficacy and you've shown that your suppression of erythropoietin production is deeper and longer and you've shown the correlation of ability to suppress that production, you've got too many dots connecting to not have strong confidence in that differentiation of the molecules. I see. Okay.

And then, so you guys, I mean, you laid out earlier, so your vision for CAS in that 1L to 3L plus setting. So basically, it really dominated that entire treatment paradigm. So looking at that, I mean, what is your partnership strategy in general for CAS? Because you guys have been traditionally been very friendly to partnerships. You have a lot of ongoing partnership. So for CAS itself, how are you going to think about that partnership strategy, or is there one?

Yeah, let me state that now if Bob wants to add anything to that. So what's nice, because this is two-horse race, there's only two molecules out there. And, you know, Belzutafan, Merck is developing that. We think we have a better molecule. We think strategics that might want to combine or collaborate with us recognize that as well. And we're in a position where the collaboration strategy is do things like what you've just seen, where we're combining with something that makes a whole lot of sense, and both companies maintain the rights to their own molecule. And we have every intent of commercializing CasDataFan and not partnering anything that relates to the commercial rights. Is there anything else you would add?

Yeah, underlining the latter part for sure. I think we view whole ownership of CasDataFan as providing a ton of strategic optionality, and we think that the molecule is very attractive from a clinical collaboration perspective, And that enables us in a very capital-efficient way to go places that we wouldn't be able to go by ourselves.

You know, I'd almost take that, and that's how I think we should want people to think about looking at the entire portfolio. So with Cast Data Fan, you know, we own 100%. There's no reason to want to give up any of that. We think we can manage the commercialization. Even in the INI, basically the only program right now where there's even option rights is that our TNF Receptor 1 program, Gilead, maintains an option to that that we think will probably get to a decision on that by the latter part of this year. But X2, CCR6, STAT6, actually CD40 ligands, CD89 antibody, those are all wholly owned ARCIS programs. And that gives us another angle of strategic optionality to complement what we're doing with CasDataFan.

Fantastic. We have a couple more minutes left. There's still a lot for me to go through. It's just a very interesting time for you guys. So let's go into that second-line setting, right? You guys have data coming out for that CAS-CABOLS data. So Merck read out the Lightbark 11 in that second-line setting as well for the Bell's-Lanva combination and achieved a hazard ratio of 0.7 and then about a 15-month PFS against CABOLS 11 months. You guys already showed that 15 months for that monotherapy cohort as well in that data line setting. What do you think is meaningfully better compared to that 15-month for Bells that would clearly show that the strategy that you guys have with that CAF-COPL is a more superior approach versus Lenva and Bells?

Yeah, we think we're going to be able to illustrate that by sharing Kaplan-Meier curve. So we're not going to put out some number that, as you know, median PFS gets criticized as for what it means because it's a single point, particularly for a program where durability is driving hazard ratio. It's also going to drive the commercial aspects of this. I think also when you think about the Merck study, it's a fairly different study in terms of the design. we're both running against CABO, which we think is great because that hazard ratio will let you compare directly. When you think about the Merck study, they had to exclude both Lenva-experienced patients as well as CABO-experienced patients. We don't have that limitation because we're running CABO as the control, so we can include Lenva patients as well as IO patients. So what you'll see from us that I think will give you very strong confidence is Kaplan-Meier curves. We'll compare to all the corresponding Kaplan-Meier curves from the other combinations that you mentioned. And, like as I mentioned before, we may even have some initial reads on OS, which for us, you know, is sort of like, well, this field hasn't been around long enough with HIF2 inhibition, we feel that can be the holy grail from a commercial standpoint. because while OS hasn't been a required registrational endpoint, it's a huge point of differentiation, and I think it also reflects on that durability where we'll have an opportunity to maybe show advantages to their paradigm. So keep in mind, Merck has essentially had three opportunities now and has not shown yet an OS advantage. Okay, got it.

So you mentioned you're not that concerned about PFS. However, the market is going to be very, very concerned about PFS. So if they should not be concerned about PFS, people like numbers. They like one number that they can easily compare things to. How would you think about it? Is that hazard ratio that people should be more focusing on than PFS? What's that one number for you?

So to me, absolutely, it is the hazard ratio. And I think from while you'll be able to look at those curves, you'll be able to extrapolate to how you think that looks. But I also think, you know, people can pick whatever number they want. We're not going to twist anybody's arm out of a particular number, but we feel it would be silly to sit here now and pull a number out of the air that we think represents some sort of a milestone on any of our endpoints. Even when you look at our late-line data, what we would emphasize is the shape of our curve, and we talk about it all the time. The number of patients, even stable disease patients, that are out past two years of therapy, that's really going to drive both the utilization and then the commercial opportunity. So we recognize that people will look at the number. We think our numbers will be good. but we can get asked as much as people want. We're not going to look to throw out any single number for any single endpoint, nor an ORR, you name it. We're not just going to throw a random number out, especially when we don't know the data yet. Right, okay, got it.

And then, so for this year, or ask more timing readout, can you walk through what are the endpoints that you would show? I think you talk about the Kaplan-Meier curve. There's about 40 patients or so. What other endpoints should we see? The DOR, I think you mentioned potentially OS. Maybe just walk through all that.

Yeah, I mean, some pieces of that will be a little bit more mature than others for sure. But I think Terry said, you know, the Kaplan-Meier curve for PFS is probably the piece that we would most underscore in terms of telling us the most about the future of the program. I think we'll have various landmark numbers, so you'll look at kind of that time course of what's happening with patients over time. We do think that, especially just given the pharmacodynamic profile of castataphan, we think the tail effect for that molecule could look really good. So landmark numbers will be important in addition to just a median. And then I think, as Terry also mentioned, the less mature pieces, but the pieces that we may begin to look at would be OS and then duration of response, as you mentioned. I see. Okay, got it.

And then how do we know that this data is going to be reliable, just given that there's no control for this? And then is there any reason to believe that this number could be somewhat inflated inflated, and then once you go start that, you know, that phase three larger study, that it's a different number.

Yeah, so it's interesting. Ironically, in this particular setting, you saw a little bit of the opposite effect with the Belzutafan data. And hopefully the other piece of it is actually ex-U.S. sites performed from an actual efficacy perspective with the Belzutafan program slightly better than the U.S. sites. So some of the dogma that folks have around, like, deterioration of data in U.S. versus ex-U.S. hasn't played out in their pivotal program for Belzutafan. We think we have a pretty representative footprint at U.S., ex-U.S. ARC-20 is a big platform, if you will, with over 40 sites, and we think there's a good representation there. So, you know, I think we feel as good about it as you could for a study of that size.

I think the data are so robust, and there's so many different studies all pointing to the same, you know, the same answer that, of course, you know, your question is like, it's one of those questions that can never be answered. You haven't run the 1,000-patient study. No one's run the 1,000-patient study before they've run the 1,000-patient study. But I don't think you could have a better, more comprehensive data set that goes from, you know, biological profile, PK, PD, clinical efficacy, large number of sites, large number of settings, as opposed to so many times where, you know, where data doesn't hold up is when someone comes in with like 20 patients' worth of data and there's nothing to calibrate it against. The other thing that we've shown, and, you know, we have the same questions early in the monotherapy. Everything you're asking now, we have the same questions with the monotherapy data and what we've shown consistently, and we actually always put the slide in and we'll do it again, the data have gotten better with time as opposed to degrading. And I think it's a hallmark of this mechanism where you have, you know, something that's, keep in mind, 85 to 90-plus percent of patients have HIF-2 as a driver. So every time when someone would ask a question about, like, how do you think it, you know, whether it would hold up or not, it would almost be the opposite that would be the shock because you're going after something that's fundamentally a part of this tumor setting and you're hitting it robustly. And I think, if anything, the Merck data with a relatively modestly good HIF-2 inhibitor have so validated the target and the idea that you would have a molecule that's hitting the target so much harder and not have it manifest itself would be more of the shock than the other otherwise. Fantastic.

So we're well over time. Thank you so much. And very exciting stuff. I think next year we have to book you guys for two fireside shows.

Thanks. Thanks for giving us an extra three minutes. We appreciate it.