Let's start. So welcome back, everyone, to the panel track. I'm Yigal Nochomovitz. I'm one of the biotech analysts at Citi. So this is the immuno-oncology panel. I think this is the first time I've done an IO panel recently, so welcome everyone. We have the CEO of Arcus, Terry Rosen. Welcome, Terry. Mark Lanasa, the CMO of Solid Tumors at BeiGene, welcome. And Theresa LaVallee, the Chief Development Officer at Coherus. Welcome, Theresa. We just recently launched coverage on BeiGene and Coherus, and of course, Terry and Arcus, we've... It's been many years.
Forever.
So, for those less familiar, why don't we just kind of run down the line and just... I know there's a lot to talk about, but just give a very high-level, two-minutes elevator pitch on the, on the pipeline, and just what are the key focus with, you know, trying to emphasize the IO angle, given that's the topic of the day. So, okay, Mark.
Great. Happy to start. Happy to be here today. So at BeiGene, we're very committed to development in solid tumors, with a specific focus in immuno-oncology. Our cornerstone medicine is tislelizumab, our PD-1 antibody. It's been a really positive year for tislelizumab, with three more positive Phase III readouts in ITT and frontline gastric cancer, extensive stage small cell lung cancer, and in resectable non-small cell lung cancer. We have a wave of NextGen or Next Wave IO molecules that are coming through, which include co-checkpoint inhibitors. I'm sure we'll talk about TIGIT today, but also LAG-3, TIM-3, and one of our newly disclosed targets is PVRIG. And then we have a group of newly emerging immune targets, which include HPK1, DGKzeta, which are inhibitory kinases downstream of the T cell receptor.
We're quite excited about those, as well as CCR8, which is a Treg-depleting monoclonal antibody.
Great. Theresa?
I'm Theresa LaVallee from Coherus BioSciences. Firstly, thank you for the invitation. Great to be on this panel with the colleagues here. Coherus is a commercial-stage biotech company in the San Francisco Bay Area. It started in biosimilars. Over an 18-month period, we're halfway through launching five products. Importantly, one of those is our first immuno-oncology product, toripalimab, a next-generation PD-1 inhibitor, that we're partnered with Junshi Biosciences. We're the US and Canadian partner, and it's really shown profound overall survival effects in nasopharyngeal carcinoma, rare cancer. Then our pipeline is focused on extending patient survival, whether that be with a TIGIT, really maximizing the T cell, or a big focus of ours, like many of the folks here, including Arcus, is on the TME. So looking at the effects of the immunosuppressive mechanisms to try to bring PD-1 checkpoints immunotherapy to additional patients.
We're in the process of looking forward to near-term close with Surface Oncology, where we'll have a clinical stage IL-27 antibody that's shown single-agent responses and immune activation in patients, a CCR8 ADCC-enhanced antibody targeting the Tregs, and then our own homegrown discovered ILT-4 antibody against the immunosuppressive cells.
So Arcus is a little earlier stage. We've been around for eight years. We started from a blank piece of paper. We're looking to build a long-term independent company. We've got seven molecules in the clinic right now. We've got a big discovery group, a growing and pretty large development group. On this panel today, probably a big focus would be the anti-TIGIT program. We got involved early in anti-TIGIT. Our anti-TIGIT's differentiated from the other molecules in late-stage development, and the fact that it's an Fc-silent anti-TIGIT, which we believe may confer certain advantages. We have four ongoing registrational trials, three in various lung settings, one in GI. The one in GI, we're well-positioned to be first in that setting. We're in a, as I said, a phase III trial. Our pipeline is broad, we're...
Both in terms of molecules and studies. I'll highlight a couple other molecules just because of the near-term readouts and to bring some focus to what we're doing in the immuno-oncology and the things we have. A pretty unique small molecule, CD73 inhibitor. We're gonna be sharing OS data in pancreatic cancer later this year or the latest, early next year. We think this could be one of the biggest movements in pancreatic cancer in the better part of a decade. We also think that their biology, which I would say reflects really the aspirations of immuno-oncology, enhancing the immune system where you already have a T cell response, but with essentially no added toxicity to improve long-term and overall survival. I think that that program is very prototypical of that.
Then the other program that I would just highlight, that's really outside of the immuno-oncology, and we are a little bit more agnostic in, in not just thinking about immuno-oncology, as I guess others are as well, is our HIF-2 inhibitor program, where we'll have some data from phase I later this year, that really clarifies its advantages relative to the one commercial agent, belzutifan. Belzutifan has, you know, demonstrate a proof of concept, but it has a very clear and notable blemish in that it has absorption-limited pharmacokinetics, so they can't hit the target well enough.
We'll be sharing data that basically PK, PD, and safety that show that we can hit HIF-2α, probably three to four fold the exposure of the Merck molecule with an incrementally more potent molecule, but with the PK profile that enables us to do that in an equally safe manner.
So if you could each talk a little bit more about how you're building your IO portfolio. You've taken different strategies, there's internal development, there's in-licensing. Just expand on the strategy. How do you build an IO portfolio? What do you look for in a target? What kind of balance of targets do you look for, mix of small molecule sometimes versus a versus biologic? What's some of the high-level thinking? And then we can get into an important topic, which has already been mentioned, TIGIT. So.
Well, yeah, if we can follow the same sequence.
Sure.
So the way that we have approached it is to think about mechanism fundamentally, that we're looking for combination partners for tislelizumab, our PD-1 targeting antibody. Understanding that there are multiple facets to the immune system that are dysregulated in human cancers, and we believe that PD-1 is the beginning of the story, but it's unlikely to be the whole story. So what we wanted to do was to have tools that would allow us to move these different components of the immune system, whether that's the tumor microenvironment, whether it's T cells, whether it's NK cells, which we're also looking at with some of our molecules and bispecific antibodies. A number of the molecules that you see in our portfolio now are co-checkpoints that are targeting other checkpoint molecules, either as doublets with PD-1 or higher order triplets.
As we're looking at PD-1 plus TIM-3, LAG-3, we have intent to look at PD-1 plus TIGIT, plus PVRIG. Again, we're very interested in these small molecule kinase inhibitors for the molecules downstream of the T cell receptor. We're interested in T reg biology with CCR8 as well, also PI3K delta. So it, it's about having mechanistic diversity to test different hypotheses, which also gives us an entry into different solid tumor types.
I think for Coherus, we are biologically focused, so biologics, not small molecules, and using the data that's really come out over the last 10 years of work in the clinic, looking at mechanisms of PD-1 resistance. And so that goes with the TIGIT and how it really was initially thought to be part of PD-1 resistance. I think the clinical data is telling us maybe that wasn't as true in patients, but seeing really maximizing antitumor immunity through the T cell, and then importantly, a number of mechanisms of known PD-1 resistance, and how to modulate the tumor microenvironment to promote T cells getting in and being active, and that's where our focus has been. We have our own program with ILT-4. We had an internal program on CCR8.
The acquisition of Surface Oncology accelerates our program three years and saves a lot of money, so we're clinical stage instead of being behind on that one, and really looked at that asset as being highly selective and potent, so prioritizing it in a development sense.
Similar, we look again to have a fairly broad approach. We're a little bit agnostic in exactly the type of target, but we're looking for places where we think we can meaningfully differentiate. Ironically, anti-TIGIT was something when we first got into it, we got into it early, and we felt it was more an enabler than something upon which to differentiate. So we felt we were starting a company from scratch. We felt we were going to be big in oncology. We brought in our own anti-PD-1. We felt that there was possibility that anti-TIGIT could prove to be very important. And so anti-TIGIT has turned out to be a big deal for us for two reasons. We happened to get in early, and then we had this differentiation.
But strategically, we tend to look at targets, whether they're tumor cell intrinsic or they are immuno-oncology, where we see that basically, there's an opportunity, because based upon the biology that's already out there, there's a high degree of validation. But clinically, it may be that there's neither a good molecule nor actually have that validation. So our sweet spot technically tends to be in the small molecule arena, where we believe that the group we've built, you know, not just in the medicinal chemistry, but all aspects of that, enables us to... if there's a target, we're going to get a best-of-class molecule. And then thinking very carefully would be about the combination partners, recognizing that you can change the setting, you can change the combination, but you can't change the molecule.
So anything you see that we do in that small molecule arena, we're going to feel that it's, it's highly differentiated based upon the properties of that molecule, but biology drives everything.
Okay, so let's just have the discussion around TIGIT, and then we can move on to other topics. So obviously, we saw the recently leaked data from Roche, but there are other data sets. There's your data set, the ARC-7 data set. There's Roche has the hepatocellular data, there's the CITYSCAPE data, there's early data from Beijing as well for your TIGIT. So just kind of talk about the pros and the cons. You know, how do you see this in terms of odds of success, that the target is going to be validated and sort of reach the pedestal of the PD-1 or not? Just kind of go through your thinking, pro and con.
Sure. Thank you. So, to tell perhaps our narrative as it relates to TIGIT, we also were relatively early to the clinic with our TIGIT molecule. Our decision about having an Fc-enabled molecule derives from the preclinical work that we did. That, in preclinical experiments, we found that the Fc effector function was necessary for efficacy in preclinical models. So if we engineered that out with the same epitope, we lost the activity, so that was a decision that we made, a deliberate decision to retain the effector function. I think that the way that this played out over time was the CITYSCAPE data came out. There was a lot of excitement. That looked to be very, very strong proof of concept.
The data that was coming out of our phase I study, the 900105 study, in both tisle combo, as well as tisle plus chemo combo, was consistent with the data that we saw in CITYSCAPE, and therefore, we quickly moved into phase III in that PD-L1 high subgroup, similar to actually almost identical to the Skyscraper-01 study. In other tumor types, however, because we have built an in-house delivery mechanism, we think that we can generate phase II data quickly and at favorable cost. We made a decision to generate randomized datasets in hepatocellular carcinoma, esophageal cancer, cervical cancer, lung chemo combo, and small cell lung cancer. We started all of those studies at around the same time, and actually, we're gonna present three of those studies at the upcoming ESMO.
So we're excited to share those data, talk about next steps. So as it relates to Skyline, we were surprised to see those data in the public domain. I think it's encouraging to see the magnitude of effect that's been observed, even if it's not yet statistically significant. The concern with the data that has been voiced is that the control arm underperformed, which makes interpretation a little bit difficult. So I think the decision that we've made with our study to have pembrolizumab as the control arm, I think, was the right design decision to be able to say, "Yes, we have beaten the objective standard of care for PD-L1 high non-small cell lung cancer." So exciting to see how these data continue to play out.
Our study should complete accrual or is on target to complete accrual this year, with a top-line readout late next year.
Okay.
Yeah, and I think for me, TIGIT is a great example of the drama in science. It's, like, so much up and down and reactionary, instead of just being really vigilant with the data. So the devil in the details, Terry mentioned Fc-silent, Fc-active. There's also the mechanistic rationale behind TIGIT, really is the crosstalk between PD-1 and, and TIGIT. So the Skyscraper data, while important, may not tell you the whole story 'cause it's a PD-L1 and an Fc-inactive. So it may... What it tells you, though, I think, is that there is biologic activity. The next is, do you get a P value to have clinically meaningful difference to get to market? So the way that we've looked at it across the board is really thinking about TIGIT and PD-1 crosstalk and where that pathway would be relevant.
The tumor type that we had prioritized based on all the disease linkage studies that we've looked at is HCC. We were very pleased to see activity in that disease type, where PVR is the highest, and really have focused our efforts on really trying to get to proof of principle in combination with our asset, which is also Fc-silent. I think that's an interesting argument that I've spent a career... I mean, it even goes back to panitumumab versus cetuximab, right? Should it be IgG1 or IgG2? And we had this whole debate with OX40 and so this Fc-silent, Fc-active, the mouse is different than the human.
I think particularly as the translational work comes out and we see the effect on Tregs, because that is the one could be the differentiator, where we see that cross-linking with the Fc active, depleting TIGIT-positive Tregs, and then trying to figure out what that contribution is to any efficacy in the clinic.
Yeah, so the first point I'd elevate to your first question about the importance and the role, and I think it's unambiguous from the multiple datasets and without running through all of them, and multiple experiments, different companies being data-driven, again, not driven by narrative, that anti-TIGIT is important. And it's a target that's clearly going to find its place in the armamentarium. And I'd like to talk about one piece of data that's hardly gets discussed, but I think does point to the breadth of the opportunity in that when we talked about the CD155 biomarker work from ARC-7, we weren't talking about that at all as a patient selection strategy, but more as a molecular smoking gun, if you will, for this being real. Now, the importance of that wasn't just for lung, and it wasn't just for ARC-7.
So what was known in the literature before any of us did an anti-TIGIT experiment, is that if you have high CD155 and you're gonna get anti-PD-1 alone, high versus low is a very negative prognostic. And in fact, you know, one of the things we validated in our studies on anti-PD-1 alone, if you have high CD155 versus low and just picking the median as a cutoff, the hazard ratio is like 0.38. So you're gonna do worse on if you have high CD155. That's not surprising 'cause you talk about that linkage between anti-TIGIT and anti-PD-1. So clearly, patients that have this other you know ligand for that when it engages its target, TIGIT on immune cells is immunosuppressive, it's not surprising that they do worse.
The second component of that, though, was that the doublet actually reversed that. So patients that had high CD155 now performed as if they had low CD155. The importance of that isn't just to that study, but it points to, that tie between biology and clinical activity that's so correlative that it, it's not a surprise. And so, like you say, HCC, if you look in GI, for example, you're gonna see that same negative prognostic. If you look at, PD-L1 all comers, where you have chemo plus anti-PD-1, that high CD155 is even more a negative prognostic. So it bodes very well in addition to that, TIGIT PD-1 linkage as important in, in that immunology. One of the things that we're gonna share later this year is from a study called, Edge-Gastric or ARC-21, and that's in the same GI setting.
That's the adenocarcinoma histology. It's a registrational study where it's the one where we are well positioned to be first. I think what's gonna be interesting about that is not only the efficacy, it's a 40-patient data set. It's single arm, so you can only interpret it so much. It was really designed to facilitate getting you know broad sites up and going around the world, but the data set is meaningful. It's 40 patients, you'll see efficacy, and you will get a sense of safety. I think that the Treg piece that you talk about is something that will play out more.
The one thing that is very clear from clinical studies in humans is that if you have an Fc-enabled anti-TIGIT, you are gonna see reductions in Tregs, and that certainly can and is, you know, leading to some immune-associated AEs. So you get a sense of that doublet, which in the end, as we move forward, is gonna become probably a, you know, a core therapy, anti-TIGIT plus anti-PD-1, plus then whatever gets combined with that, that I think you'll be able to start to get a sense of that efficacy, but with essentially no safety baggage coming from that anti-TIGIT.
Microphone. Do you have a microphone for Andrew? Thank you.
That's a small one.
Thank you. So Andrew Baum, Citi. TIGIT is a molecule near and dear to my heart, so, as you all guys know. So couple of questions. So firstly, given the curves from Sky One were very different from what we saw from CITYSCAPE, there's clearly questions about, you know, the thesis may be different from what we initially thought it was, because this looks like an undefined subgroup of patients that are responding to a late treatment effect. So how does that inform trial design, particularly the utility of PFS versus OS? And when you look at your existing trials, whether you're tempted to take out PFS in order to maximize the power for OS.
And then the second question relates to the chemo combination trials, and it would be easy to argue, given the intrinsic mechanism of TIGIT around PD-1, which we think, because we have to be humble, because I think there's seven different mechanisms for PD-1, so all this is hypothetical, and we may be revising this in a few years. But let's just say that it is, right? When we look at the chemo combination trials, should we assume that the probability of success is low because obviously the PD-L1 expression is low, or do we believe in the sort of magical concept of immunogenic cell death, and therefore, we can expect something different here?
Obviously, we're all hoping it's the latter, not the former, but I'm just interested in your views, given we have that phase II CITYSCAPE data, where all the benefit was driven by the PD-L1 high patient population. Whoever wants to go.
Well, so, I would actually say on the trial design, if you start, obviously it's, it's gonna be setting dependent, but in lung, OS is the only approvable endpoint, as you know, and so we do think about that effect on OS, probably being that, you know, what people talk sort of prototypical about immunotherapy. I'm actually gonna use something from ARC-8 because it gets to that chemo OS concept that I think you when we, when you share those data, it'll reinforce when people talk about immunotherapy, mostly what they're talking about is anti-PD-1. It's not like there's 100 different things.
I think what you're gonna see, that's an example where the removal of adenosine or, you know, blockade of adenosine formation, where essentially what you're doing is not creating a T cell response, but if you have a T cell response, you're going to enhance it. And so what's interesting in that, and I think that plays out where you see dramatic effects on OS, but you're not seeing the same sort of dramatic effects on response rate or PFS, is that if you think about and this gets to your point about chemo, and I think why there should be high optimism about that, essentially think about that vaccination concept that you're creating, an immune response.
So historically, when you think about response rate, and if you think about chemo, really what you're getting is that debulking, and that's, you know, that, that's a response. So 27% tumor reduction is a stable disease, 39% is a partial response. Now, it may actually be that the 27% response during that debulking has a better immune response. So if you can do something that enhances that immune response playing out better, that patient may even do better than the 42% reduction.
That's why I think you see this really profound tale, is that if you can get the immune system to do what it's supposed to and then enhance it, the early part of the whole process may not be the most critical, and you may see benefit in patients where you can actually generate an immune response, and now you're either enhancing that or enabling it to last with greater durability. And you get that real measure, which is what we're all aspiring for, long-term overall survival, and particularly in a very safety favorable way. That's the other thing that you're starting to see with anti-TIGIT. You're gonna see the same thing with CD73 inhibition, is it's coming with almost no safety baggage at all. So very optimistic about the chemo being able to enhance it with the immune activity.
My question wasn't indicating any strong bias for me. It's just representing the conflicting views-
Yeah
... based on the translational medicine, right? And even in Merck, you know, within the same group, people arguing from both sides on those lines. Thank you.
But I think you have to look at it, too. I mean, even if you go back to the early days of when PD-1 and CTLA-4, PD-L1 went into the clinic, you know, I can remember the arguments about you can't combine it with chemo 'cause you're gonna kill the T cells, or later, even with the MEK inhibitors, that it doesn't make sense mechanistically, yet they've shown efficacy. So I think from... we have to see the data in the humans and also not say that there's all chemo are the same. We know some chemos combine better in the disease setting. Going back to what I think Terry was alluding to, CD155 PVR, thinking about where the pathway is relevant, so it would make sense.
You know, whether you call it immunogenic cell death by the canonical definition, with all the right biomarkers, or you say the chemotherapy is able to stimulate an immune response by releasing antigens or whatever it is. I mean, I think that's been shown that in the context where TIGIT is really potentiating a PD-1 anti-tumor immunity response of the T cell, that in the right context, it should work. So I don't know that there's a magic plug-and-play here, that you really have to do the thoughtful experiment with both which chemo and which disease context are you talking about. But I would be optimistic that in the right settings, that you can see that efficacy.
So I, this is a wonderful and very thoughtful mechanistic conversation. I think as it relates to the chemo combo, I'll make a sort of very pragmatic argument that, we weren't sure. There's a lot of, data that could be interpreted in both directions, so we made the pragmatic decision to, "Let's run a randomized phase II study." So we're enrolling a randomized phase II study, 270 patients. That study is fully enrolled. We're just waiting for the primary data to read out, and we'll share that next year. Again, the idea that ultimately it's the clinical data that can allow you to have good decision-making entering in the phase III, where the majority of the costs are born. I, I would go back to the initial point you made.
I also was surprised by the shape of the curves that we're seeing in Sky One. It is different from what we saw in CITYSCAPE. I was surprised by the non-proportional hazards, the late separation, when they separated the magnitude of separation. I was open to making an argument that PFS, because this is IO versus IO, if you had statistically significant, clinically, you could argue that that's clinically meaningful and therefore approvable. But, certainly OS is the gold standard in lung cancer, and I think it does raise interesting questions related to, the maturity of the data, the information fraction, those sorts of things that we can think about.
How much is known about the differential potential for TIGIT in different solid tumors? I mean, is there a thesis out there that it could work better in lung versus hepatocellular or gastric, or that's still really an open question? And is it related to the CD155 in any way?
Go ahead.
I'll start this time by shaking it up in order here. I mean, I think that's the idea of clinical development should be that it's always a question. You don't do trials because of this; you do it with a clinical hypothesis. So PD-L1, everybody can talk about whether they like it or dislike it as a biomarker. I would argue it's been one of the most successful biomarkers, not for patient selection, but telling you who to treat, what indications to go into, and the prevalence of it. So and I think that's what Terry was alluding to. Their data, they've seen that PVR tells you the prevalence in the population and where the mechanism is relevant. So it's a little bit different for disease positioning versus patient selection. So that you look at, is there enough patients in the population that will benefit from it?
Because the issue in IO with the direct patient selection, it's not a driver mutation, it's an immune system. There's a lot of reasons for PD-1 response and PD-1 resistance. I mean, when you read the reviews, you get the 18 mechanisms of PD-1 resistance. TIGIT may have been one of them, PVR could be a factor, but then there's MHC and the tumor microenvironment and all these other things. So I think that's where it gets really complicated quickly for people. But using the PVR, we do see broad expression in many tumor types, particularly in GI, that could lend itself to having activity. So it wouldn't be like, oh, LAG-3, it worked in melanoma, nobody knows why, and now we don't know what to do with it.
I would just fully agree and echo those points. I think that ultimately, clinical development is challenging, and you have to have an a priori hypothesis, and it needs to come from somewhere. So, looking at the preclinical data, again, we also are very interested in PVR expression for rank ordering tumor types for study. That's how we, in part, came to our list of tumor types that we're looking in phase II studies. The same type of framework, whether we're talking about TIGIT or LAG-3 or CCR8, you need to have some sort of tumor prioritization matrix to place your best bets since there's the human immune system is so complex, and there's so many different facets we're trying to control.
Yeah, and so to not pile on, I'll just, like, say a couple sentences because I agree with everything that was said. I think that hypothesis right now that anti-TIGIT turns anti-PD-1 into super anti-PD-1 is a good working hypothesis based upon biology and based on where PD-1 and TIGIT are, so the immune part of the equation, as opposed to the ligand part of the equation.
And so that's a good starting point, and then thinking about where, you know, again, something that makes a whole lot of sense, if that, if you are getting, you know, PD-L1 is plugging up one of those, CD155 is plugging up the other, that that's a pretty good place to start to think about your priorities, that, that at least are in the clinical realm of things, that then your Venn diagram might include other strategic components unique to your company or unique to where you can execute, opportunities, et cetera. But I think technically, those two pieces are at the, you know, sort of in, down the middle of the fairway based upon what we know now.
All right, let's just do a little more granular discussion of pipeline and upcoming catalysts. So you have the AdvanTIG-105, the phase 1/2 non-small cell lung cancer study. That's Oci and Tisli. Can you just talk a little bit more about the goals of that trial? And what are you planning to show and when?
So we've actually published... That, that's a relatively large phase I study with a number of expansion cohorts. We had 10 different expansion cohorts to explore the opportunity in different tumor types. Again, for us at this point, because it seems with TIGIT that really what is gonna be decision-enabling is the time-to-event data, that we would have more confidence in randomized data sets, which is why we're putting weight on these randomized studies, again, some of which we'll share at ESMO. So, I don't think we'll have a whole lot more to share about the Study 900105 in terms of-
Okay
... longer follow-up on single-arm cohorts.
What about but you will have data at ESMO in the frontline, HCC and esophageal squamous cell carcinoma?
That's right. So-
Talk about those.
Yeah. So we'll, we'll have three abstracts of randomized data. So second-line PD-1 naive cervix cancer, second-line PD-1 naive esophageal cancer, and then frontline HCC in combination with a bev biosimilar. So again, these were rank ordered based upon our understanding of biology. We're very exciting to share those data. They're under embargo, so we can't talk about them yet. But the other thing that we'll be able to talk about is our next steps in those tumor types, subsequent investments into later-stage trials.
Okay. Then can we talk a little bit, Theresa, about the for TORI, the anti-PD-1? Can you just give us an update on the operational side of the coin as far as FDA inspection and timelines, to the extent you can comment on approval there?
Yeah. So, for those that haven't followed the space, Toripalimab has a missed PDUFA date from December of last year. And that was mostly because... that was solely because the FDA couldn't travel to China to do the necessary inspections, both on the manufacturing side and the clinical side. We're glad to say that we are just about cleared those last elements for the inspection, and we anxiously await FDA action. I mean, we do have breakthrough therapy designation and a very profound overall survival advantage as it was showcased at ASCO, and the discussant really got up and talked about practice-changing benefit. There are no approved treatments in the US for NPC, so this would be the first approved agent across all lines of therapy: frontline in combination with chemotherapy, and then PD-1 monotherapy in later lines.
So the challenge, Yigal, is, without a PDUFA date, how do you know what the clock is for the FDA? I mean, we were talking about this. We have a similar-
Shared misery
...misery, urgency. But you know what? Trust me, every day of my life, I think about of ways to engage. So we think it, it will happen soon.
... Okay. Terry, let's talk about the gastric trials, 'cause I don't think that gets enough airtime. The Edge-Gastric study, which you already mentioned, I think you're gonna have some of the data. That's the dom plus zim plus chemo in first line, upper GI, and you're gonna have the data later this year, if I'm not mistaken. So just talk about the rationale for that, and then also the larger trial, the phase III, the STAR-221 phase III trial.
Yeah, we're actually super excited about that study. That's one of the reasons we got into it early. This, you know, in fields like this, you kinda have to parallel process. So we have the phase III up and going. The reason in this platform study, Edge-Gastric, is part of opening as many sites globally. We recognize that in certain geographies, they're gonna wanna see data with the exact combination and the exact setting, particularly from a safety standpoint. So this 40-patient study was really designed with that intent. So I'll actually say that STAR-221, which is the registrational version of that, is enrolling. We've been saying gangbusters. It's really enrolling quite well. And part of the rationale is because there's not only big medical need there, but there's clinical trial need.
So it's a setting in contrast to lung, where, you know, there's such huge competition from so many different agents and so many different studies. These investigators, there's not a whole lot to look at, so it's been enrolling great. The dataset that we'll share is from those 40 patients. It's single arm, but it's meaningful. You're gonna get a good feel for ORR. When we share these data, we won't have mature PFS. We may have six-month PFS, but it may be telling. We think the data are exciting, and you'll also be able to look at a safety profile that does give you a good feel for anti-TIGIT, Fc-silent, anti-PD-1, and chemo, and how that compares...
Because that's gonna be an important part of the equation, as it always is, as we talk about these agents that, that do have better safety profiles and the benefit they bring. So we don't wanna overstate because we recognize it's single arm. We recognize the all the issues that we're always whining about, as a matter of fact, in terms of cross-trial comparisons. But, you know, you'll have a Merck dataset with chemo and anti-PD-1, and you'll have the BMS dataset to look at historical. So I think it'll be a meaningful dataset for us and the field. And the GI setting is an exciting, and I think it's gonna be a big deal for anti-TIGIT.
What about how well those patient populations mirror each other, Edge-Gastric and STAR-221?
So-
How well do they mirror each other, those two?
Oh, I think they're gonna be... It's a very similar patient population. So obviously, the registrational trial is gonna be much more global, and it's larger. But I think that the phase II study will be very, you know, about as good as you can get from a 40-patient study. And that enrolled very well. There was a lot of excitement in the investigator community.
And then, Mark, the AdvanTIG-302 phase III , just, I do get this question from time to time. Just walk through how it's similar or different from the Roche Sky One trial. What are the key differences and key similarities?
Yeah. So I would say the key differences, as we highlighted before, the combination partner for TIGIT is PD-1 rather than PD-L1. And then our comparator arm, our control arm, is pembrolizumab monotherapy, which is the unquestioned standard of care in non-small cell lung cancer. It's actually a three-arm study. We have an arm, underpowered monotherapy arm for tislelizumab to establish contribution and components, but the prior two were the two key differences.
What's the timelines for the... You're gonna finish enrollment by this year?
By the end of this year, with the initial readout scheduled for the end of next year. Event driven.
Okay. You commented on some of the earlier stage programs, the anti-TIM-3, the anti-LAG-3, the OX40, the HPK1. How are you thinking about relative prioritization of those to bring those into the clinic? And how do you see them in terms of relative competitive, competitiveness as IO targets?
So, as far as, shall we say, the safest bet in the portfolio is, beyond TIGIT, probably LAG-3, given that it's been successful in melanoma with BMS.
Mm-hmm.
We have a relatively broad development program, again, with multiple randomized phase II studies getting started for LAG-3. As I mentioned before, just to highlight something else, we're very interested in the small molecules of HPK1 and DGKζ as inhibitory kinases that are blocking complementary pathways downstream of the T cell receptor, with interesting data emerging in HPK1, both in terms of efficacy and safety.
And then, Theresa, just let's hear a little more about the Surface Oncology assets, the IL-27 and the CCR8. What is appealing for those two assets, and why did you bring them into the company?
Yeah, I think when we talk about biology and disease positioning and clinical hypothesis, the IL-27 is about as straightforward as you can get. I mean, I think it's an immunosuppressive cytokine. We know that modulating cytokines can modulate the immune system. So the big surprise with this one, at least for me, was that there was single agent response. So monotherapy treatment, and, you know, everybody can argue with Emmett Schmidt's arguments about what criteria you need to use, but I think everybody will agree single agent activity is a better criteria for probability of success later in development. Really importantly, the places where there was single agent activity in renal cell carcinoma and non-small cell lung cancer, and efficacy as well seen in combination with PD-1 and HCC, is from a disease linkage standpoint and preclinically where they saw activity.
So when they have done a series of mouse models, they get activity, but not universally. So there's a lot of tropism. So they get activity in the lung and in the liver, not sub-Q models, not other models. And that really lent itself well when you looked at where expression was highest for IL-27, as well as prognostic significance. It's highest in HCC and other tumor types where there's activity. So I think it really sets itself up well with, from all of the disease linkage data, the preclinical data, and then early phase clinical activity, to really position it for development and getting that answer to the clinical hypothesis quickly.
Sorry. Just in terms of the IRA, and now you have 13 years to play with, and arguably the Sky One leak from a Bayesian perspective may de-risk the role of TIGIT. Do you think you have enough to initiate an adjuvant trial with the drug, given that, you know, these are long trials to run? Or is this something that, you know, you think is gonna be beholden upon a large cap pharma to do, and it's not something that you necessarily have resources or bandwidth to pursue? You know, obviously, then the other question is mechanistically. I mean, you would imagine it may work better, but who knows?
Yeah. No, and I think, you know, there's a lot of interesting discussion coming on now in terms of the adjuvant space. I mean, we're having active discussions about these types of things, but you can also think from a setting up to time to endpoint in the neoadjuvant space with path CR as a good readout and all the correlatives with circulating tumor DNA. There may be ways to really think about that from a de-risking and positioning for later stage development. How that fits in with IRA, well, that's a much longer conversation, but, yeah.
Okay. Thank you very much. Good discussion.