Thank you, Taylor, for joining us. We've got Taylor from Shattuck here with us this morning. Thanks for coming down to Miami.
Thank you. Nice to be here.
So let's dive right into the new program, which is DR3. This is on the TL1A axis. People are very familiar with TL1A. Everybody's very, very focused there, but nobody but you yet on the receptor side of that same axis. So can you elaborate a little bit about how targeting DR3 might differentiate mechanistically from TL1A, where the data is maybe a little more familiar to folks?
Sure. So for many TNF receptors and ligands, they are promiscuous, meaning that one ligand binds multiple receptors or one receptor binds multiple ligands. And TL1A and DR3 is one of the exceptions to that. TL1A is the sole signaling ligand for DR3, and DR3 does not bind any other signaling ligands than TL1A. So from a safety and specificity standpoint, you wouldn't anticipate different effects from blocking DR3 versus blocking TL1A. And when you're developing an inhibitor to a TNF ligand-receptor interaction, it is true that your probability of generating a high-quality antibody antagonist in any given antibody generation campaign is higher if you're trying to find an antagonist to the ligand than the receptor.
Sure. I think that makes sense to me biochemically.
Yeah. And part of the reason for that is that when you're trying to block a TNF receptor, all TNF receptors have to form trimers in order to signal. And so some antibodies can inadvertently end up as being agonists because they can cross-link two, and if you have active Fc domains, sometimes more than two receptors into close proximity to one another. So I think that may have driven some of the early emphasis on the ligand as opposed to the receptor. Our view is that the receptor is a much more appealing target than the ligand. And there's a couple of reasons for that. So first of all, if you look at where and how TL1A is expressed, TL1A is expressed by antigen-presenting cells and certain endothelial cells. And it's not a constitutively expressed ligand.
It gets turned on by different innate immune stimuli: IL-1, different TLR ligands, Fc gamma receptor activation, and when it gets turned on at the mRNA level, you have peak expression within about six hours, and then you have no more expression continuing beyond 24 hours. There's then a secondary means of regulating TL1A, which is that when the protein makes its way to the cell surface, there's a membrane proximal protease cleavage site, and that is what clips the protein from the membrane. It's the reason why you find soluble TL1A in the blood of these patients, and so it's a pulsatile expressed ligand, and you can think we can talk more about the challenges that provides for blocking antibody.
But importantly, this is even the case in patients who have long-duration inflammatory insult, who are struggling with these TL1A-driven IBDs for more extended periods than six hours.
That's correct, and so if you look at, there's a bunch of data out there. If you look at biopsies from, I'll just talk about Crohn's disease because it skips around the bowel, and if you take a biopsy from a patient with active chronic Crohn's disease and you look in an inflamed lesion, that's where you find TL1A being expressed. If you move two centimeters down the bowel to the adjacent non-inflamed tissue in a Crohn's disease patient, you don't find any upregulation of TL1A whatsoever, and if you do the same thing and you ask how much DR3 is present in those two locations, you find, first of all, it's more abundant than TL1A, and it's evenly upregulated in the inflamed and adjacent uninflamed tissue, and so the reason for that is that the cells that express DR3 are primarily lymphocytes and innate-like lymphoid cells.
And so first of all, these are cells that migrate. They don't just reside in tissues, right? Your antibody will have primary access to its target in the central compartment, and those cells will migrate in and out of tissues. And if you have an antibody which does not trigger receptor-mediated endocytosis of DR3, then you would expect that blockade to be durable in these patients. So one of the things that stuck out to us when we looked at the Prometheus phase II data recently published in the New England Journal is that if you look, there's a really useful figure in the data supplement that breaks down who benefited most and how did they benefit. And if you look at these patients, the patients who benefited most were the ones that had pan colitis. There was much less benefit.
It actually crossed no benefit in the error bar for patients that had left-sided colitis and patients that had proctosigmoiditis. That's exactly the opposite of what you would expect because pancolitic patients are typically more resistant and harder to treat. When you ask why might that be, one hypothesis is that in a pancolitic patient, you actually had upregulation of TL1A throughout the bowel. So then there was a means of retaining that TL1A blocking antibody throughout the entire inflamed area.
Whereas for patients with either geographically limited or temporarily limited TL1A expression, you essentially lose antibody control because of the natural cycle of TL1A.
That's right. And these are migratory. Both you see in Crohn's are migratory diseases. And so as inflammation spreads, the game that you're playing with a TL1A blocker is that as the inflammation moves, now there is going to be an inflammatory stimulus that's going to tell those tissue-resident APCs to turn on TL1A. And what happens first? A TL1A blocking antibody permeates that tissue and immediately absorbs that newly expressed TL1A, or the TL1A has the opportunity to bind the tissue-resident DR3 positive cells that are present in higher abundance. Of course, it's not going to be a black-and-white thing, but it's hard to imagine that there is no activation of DR3 in that setting.
So here, getting back to what you were saying about the mechanistic advantages of DR3, really the trouble with the TL1A maybe is that you have to catch that ligand before it does its job when it's newly expressed versus maintaining blockade both in and out of an inflammatory context.
That's right.
Makes sense. Makes sense. But the flip side of that is, despite a ton of interest in this axis and a lot of work being done on these mechanisms, nobody has gone after DR3 as opposed to the TL1A. We've started to see TL1A bispecifics. We've seen next-gen TL1A antibodies starting to come into development. Why hasn't anybody gone after the receptor instead?
Yeah. I think it's driven by a couple of different variables. So first of all, if you go into PubMed today and you search for DR3 or TL1A and you filter out HLA-DR3, you get about 330 hits. If you do the same thing for CD40, you get 17,000. Do the same thing for OX40 or 4-1BB or GITR, you get about 2,000, right? So now everybody's talking about TL1A, but historically, there have been a very small number of labs across the world that have studied it for a prolonged period of time. Steph Targan was one of them. Eckhard Podack, who was my scientific mentor, was another. Fabio Cominelli at Case Western was another. Richard Siegel at the NIH. And that's about it. I'm in Amersham County in the U.K.
One of the first companies to build a TL1A blocking antibody was Human Genome Sciences way back when. They sent some of that TL1A inhibitor to Steph, and they sent some to Fabio. They did work with the antibody, and that sort of initiated some of the understanding of how this axis works. It was an okay antibody, but not a great one. So Steph set out to make his own. I've never asked him the question, but was it solely because he'd had the experience with the HGS antibody that he sort of said, "This is the right way to go"? Was it also influenced by his knowledge that it was slightly harder to build a receptor-blocking antibody than a ligand-blocking antibody? I don't know, but it's probably a mixture of those two variables.
It really wasn't until, obviously, the Prometheus phase II data that this axis declared itself as being clinically important. Before that, Teva had failed a trial in asthmatic patients. The jury was still very much out on what this could mean as a product.
Makes sense. Let's talk about durability or clinical differentiation, I guess. Durability comes to my mind at the top there. Just because the half-life extended TL1As are shaping up to have multi-month durability, but we've just spent some time discussing from a biological mechanism perspective some of the challenges of going after the ligand versus the receptor. What would you expect to see from a DR3 antibody program in the clinic on the durability front, especially, but in early data? How would you be able to declare yourself as differentiated from a TL1A?
Yeah. There's a couple of parts to the answer. So first of all, on the dosing and exposure side, one of the benefits of going after a stably expressed receptor that's on lymphocytes is that you can measure receptor occupancy. And we just finished our GLP tox study, and we expected to see durable receptor occupancy on circulating lymphocytes. We knew preclinically that it didn't cause receptor-mediated endocytosis, and we know clinically that a drug like Keytruda, for example, binds to PD-1, doesn't cause internalization. And if you look at receptor occupancy, you see full PD-1 occupancy six months after the patient's last dose of Keytruda. And so I think what we are seeing already is that sort of durable occupancy profile. And I expect that seeing that in non-human primates will very closely predict what we see in phase I in humans.
Are you using a half-life extended antibody?
We have both. The one that we just finished the GLP tox is the non-half-life extended version. And so when you're looking at dosing with this sort of agent, you're looking both at it will be a Cmin-driven dosing paradigm, but influenced by when receptor occupancy begins to drop.
And you expect to see when you say Keytruda shows full receptor occupancy six months after the last dose, that's obviously competitive with half-life in serum there, right? You see Keytruda levels in serum dropping well before that. So would you expect to see a similar impact of receptor occupancy being stable even as serum concentrations of drug drop below the levels that you would expect to induce that sort of receptor occupancy with a single dose?
We do. Yeah, we do. And so we'll see. We've got the half-life extended version as well. That one will go into the chronic tox studies in 2025, and we'll make a call whether it makes more sense to stick with the parental non-half-life extended or the YTE. The other area of differentiation that we are anticipating is that because of the more durable blockade of the axis that going after DR3 may provide, that that may close the gap between in the Prometheus data set, the roughly 25% of patients that had clinical remission and the two-thirds of patients that had a clinical response. So we'll be interested to look at that. And then the third area of differentiation relates to the utility of building bispecifics off of 32 5.
There are three well-validated biologic targets now for IBD: TNF, alpha-4 beta-7, IL-23, and now the DR3 TL1A axis. IBD docs don't love combining multibiologics with an anti-TNF because it carries a higher infection risk. And so that's the reason why folks are focused more on combining one or more of IL-23, alpha-4 beta-7, and TL1A/ DR3. And it's not widely known, but Amgen built a TL1A bispecific TNF antibody, ran a phase I clinical trial, published the data last year, and killed the program because every patient had high-titer neutralizing ADA. And the reason for that is that with a TL1A containing bispecific, you have a molecule that will circulate in the serum and slow the degradation of TL1A and, in that case, TNF. So it's binding two soluble molecules. That, by definition, is an immune complex.
If you look at the Roche data with just a sole TL1A blocking antibody, over 80% of patients had ADA. The Amgen data shows that TL1A containing bispecifics have very high, tighter ADA. When you're building a DR3-directed bispecific, you're binding a membrane-bound molecule. You will not generate those sort of circulating immune complexes that you get with anti-TL1As.
Now, one of the things that you were sort of embedded in that answer that you just gave is a talk about the clearance rate of TL1A bound by antibody versus not. And this is something that we've also heard from, for instance, Teva when they talk about their "selective" TL1A blockade, which doesn't block interaction with the decoy receptor, which they say is a driver for degradation of the TL1A and clearance of that complex. Would you expect to see antibody-mediated stabilization of TL1A to be dependent on that complex's ability to bind the decoy receptor or not?
No.
This is a half-life. This is an antibody-driven stabilization.
Correct.
The decoy itself, you don't expect to be driving clearance of the complex?
I do not.
All right. Great.
Yeah.
Let's talk about a few more aspects of potential clinical data that I'd love to get your take on. Obviously, we're a ways away from seeing DR3 in the clinic, but you mentioned moving response to remission. We talked a little bit about durability. What about time to response? Because that's something that also seems mechanistically relevant given, as you say, a TL1A antibody potentially having leakage through to receptor before the antibody binds that ligand. Would you expect to see an impact on time to response going after the receptor versus ligand?
I'm not sure, and I'm not sure partly because the phase II clinical designs are pretty set for these patient populations. And your main endpoint looking for clinical remission is going to be at 12 weeks, endoscopic remission at 12 weeks, and we will not be scoping patients earlier than that. Now, do we start to get information in terms of patients' performance in the modified Mayo score about how they're doing early on? Maybe, but we certainly won't get the endoscopic data.
All right. In our last few minutes, you mentioned getting through GLP talks and that lead asset, the non-half-life extended one. But what's the current status of the program from there? How soon do you expect to be in the clinic, and when are those key first readouts?
Yeah. So we're all systems go for IND with the GLP tox behind us. IND timing is scheduled for mid-2025, and we're expecting about a 70-patient single- and multiple-ascending dose study enrollment beginning hopefully late Q3 and enrollment complete by Q2 or so 2026.
With a 12-week endpoint for readout second half.
Yeah. Spyre just had obviously some early data from their SAD study with their alpha-4 beta-7. You could always consider releasing data early from the SAD patients, which you'll have pretty quickly. But the full data set will be available toward the end of the first half in 2026. And here, we're clearly looking for safety, PK. We're looking for receptor occupancy, as I said. And that is one of the differentiating features in a Shattuck Labs.
Before clinical data, you'll know when you can stop dosing it.
Correct. And I think will give us a better ability to pick the right dose than you get when you can't measure something like that. The other output that some folks are occupied by, that we now have data from the non-human primate study on and will confirm in humans, is that, as what I've mentioned early on, we don't have any evidence preclinically that this antibody has any risk whatsoever of residual agonism, and can now definitively be confident about that based on data that we have from monkeys.
Fabulous. Beyond UC and IBD, which obviously are going to be the near-term focus, but beyond those indications where we've already seen the TL1As take people's attention, do you think there are additional advantages in other autoimmune indications where maybe the TL1As haven't been as effective, like that asthma readout that Teva failed? Going after DR3, could you have access to some of those indications or broader indications in the autoimmune space?
I think we will, but we will initially stay focused on IBD. Part of the reason this is a more technical discussion than we have time for, but part of the reason that many folks were hesitant to interfere with this axis earlier is that it has both immune regulatory and autoimmune properties to it, and what dictates the outcome is in large part driven by whether you have control over the antigen that's causing disease or not, and in IBD, you clearly do. In asthma, you don't necessarily know when a patient's going to be exposed to an environmental antigen that's causing disease, and so that is a variable that needs to be considered when you're looking at how to advance this axis outside.
You don't want to be perpetually interfering in an immune axis if you don't have to be.
Yeah.
Yeah. Makes sense. All right. Well, thank you. We're out of time, Taylor, but thank you so much for joining us.
Thank you.
I appreciate all the discussion on this very exciting axis.
Thank you.