Good afternoon, everyone, and thank you for joining us at Needham & Company's 24th Annual Healthcare Conference. My name is Ethan Markowski, and I'm a member of the Biotech Research Team here at Needham. Joining me today from Shattuck Labs is Taylor Schreiber, Chief Executive Officer, who will provide a presentation on the company, followed by a Q&A session with the remaining time. With that, I'll go ahead and turn it over to Taylor.
Great. Thanks, Ethan, and thank you to Needham for the opportunity to present at the healthcare conference this year. These are my forward-looking statements. I'll start with providing you all with an overview of TL1A and DR3 biology and outline the reasons why we think that targeting DR3 may lead to higher rates of complete remission than what has currently been seen with TL1A. We'll talk through the antibody candidates that we're developing themselves. Most people are probably aware, but this is a snapshot of the now published phase 2 data from Prometheus and developing tulisokibart that Merck is now in control of. The data presented here and showing complete placebo-adjusted complete remission rates in the range of 25% was incredible to see, was the first data point that started to provide clinical validation that blocking TL1A in IBD patients could be clinically efficacious.
What's been even more exciting has been seeing that since that time, the results have really been replicated by both the Roche antibody that came from Roivant and now, of course, with the Teva antibody as well. You know, it's relatively uncommon to have three independent placebo-controlled trials with different antibodies going against a similar mechanism like this, with no clear leader in terms of getting toward approval. That provides really great optimism for patients that are suffering from IBD for what this might yield in terms of benefit in the future. To go through the axis itself, the TL1A-DR3 axis is fairly straightforward. For many, TL1A and DR3 are both members of the TNF superfamily. Many TNF superfamily receptors are promiscuous, meaning that one ligand binds multiple receptors or one receptor binds multiple ligands.
When that's the case, there can be a very obvious reason to target either the ligand or the receptor to maintain specificity and reduce off-target effects. That's not the case with this axis. TL1A is the sole known signaling ligand for DR3, and it does not cross-bind any other signaling receptors. The only other protein that TL1A is known to bind is a soluble molecule called decoy receptor 3 or DcR3. This was a protein that evolved in humans. It doesn't exist in non-primates. Whenever that happens, it tends to give you an indication that this is an evolutionarily important pathway. The role of DcR3 is to neutralize serum TL1A.
Because of the simplicity of the axis, we believe that the clinical data to date showing the safety and specificity of TL1A blocking antibodies is highly likely to predict the safety profile of a DR3 blocking antibody. The fundamental biology around this axis began to be published in the early 2000s. TL1A was first described as the signaling ligand for DR3 in 2003. Since that time, a limited number of labs in the world, including my academic mentor, Eckhard Podack, at the University of Miami, have been the folks that really have defined the rules of engagement for how this axis works. This figure on the right here came from Richard Siegel, who was another one of those early investigators that was important in defining how this axis works.
What he showed in this publication here was that if you have mice that express high amounts of TL1A, those mice become highly prone to developing spontaneous inflammatory bowel disease. You could fully neutralize that phenotype if you cross those TL1A overexpressing animals onto a background of animals that do not express DR3. This is one of many data points from those days that proved the monogamous nature of the axis and that TL1A was the sole signaling ligand for DR3. If you look across all of that early literature, you will not find examples of where either genetic deletion of DR3 is inferior than genetic deletion of TL1A or where pharmacologic inhibition of DR3 is inferior to pharmacologic inhibition of TL1A.
You will find a few examples, including this one, where this came from Fabio Cominelli's group at Case Western Reserve University, where he has a mouse strain here that is highly prone to developing spontaneous Crohn's disease-like ileitis. What he then did is he crossed those animals onto animals that either lacked DR3 or animals that lacked TL1A. What he found was that the animals that were crossed onto a DR3 knockout background had a much higher degree of protection from inflammation than the animals that were crossed onto the TL1A knockout background. These are some of the preclinical data that get us excited about the prospects of neutralizing DR3. The other data comes from translational data. What these authors were doing in this figure here was taking biopsies from either healthy volunteers or patients with Crohn's disease.
They were biopsying the GI tract either in parts of the bowel that were not actively inflamed, or they were taking biopsies a few centimeters down the bowel in places that were actively inflamed. What you can see here is that in a normal non-Crohn's disease patient, about 3.5% of the cells in the GI tract will stain positive for TL1A, whereas about 8% will stain positive for DR3. If you then look in the uninvolved part of the bowel, you'll see that there is no upregulation of TL1A whatsoever. You only see upregulation of TL1A in the actively inflamed part of a Crohn's disease patient bowel. Whereas when you look at DR3, you see that DR3 is both more abundant and evenly upregulated, both in the inflamed and adjacent uninflamed parts of a Crohn's disease patient's GI tract.
This could be an important difference in expression in terms of targeting one side of the axis versus the other to fully neutralize inflammation. The data from that publication is one of several that showed the same thing at the protein level, but obviously there are a limited number of patients that are included in those manuscripts. What we then did is we pulled data from the Mount Sinai Crohn's and Colitis registries, and we asked, do those data confirm now in a cohort of about 2,500 ulcerative colitis and Crohn's disease patients that DR3 is indeed more abundant in the GI tract than TL1A? What you can see in red is that those data from a large cohort of patients do confirm that conclusion. TL1A is much less abundant.
The reason for that is that TL1A is a tissue-restricted ligand primarily induced in antigen-presenting cells. As I'll show you in another slide, it's expressed in a pulsatile manner. In contrast, DR3 is a stably expressed target, and the cells that express DR3 are primarily lymphocytes. That's the reason why if you look in the blood of those very same patients, you find lots of DR3 present and really no TL1A being expressed. This, again, these are two figures from a series of older publications where these authors were asking, what is it that turns on TL1A and how is it regulated? NT in this figure stands for no treatment. What you can see is that the non-treated groups of antigen-presenting cells here do not express TL1A.
When you stimulate those antigen-presenting cells with a variety of different innate immune signals, though, you see that transcription of TL1A gets turned on and then gets turned off fairly quickly. This is a figure from a different publication that was looking at the same thing. Here in human antigen-presenting cells, you can see that there is no expression of TL1A at baseline. When you expose those antigen-presenting cells to bacterial cell wall components like lipopolysaccharide, which is probably a predominant driver of TL1A expression in patients with IBD when you have barrier disruption, you have rapid induction of TL1A expression that reaches a maximum in six hours or so. Importantly, you'll notice that there is no ongoing expression of TL1A 24 hours later. It is a moving target on a transcriptional level.
There is a secondary means of regulating TL1A on the protein level because once TL1A goes to the cell membrane, there is a membrane proximal protease cleavage site that cleaves the extracellular domain, and that leads to shedding of TL1A from cells. That is the reason why you have soluble TL1A that you can detect in patients. That is also the reason why humans evolved a decoy receptor to neutralize shed TL1A. These basic biological parameters of DR3 and TL1A expression and localization are what drove generation of this figure here. In the actively inflamed parts of an IBD patient bowel, you have TL1A being expressed in red, and you also have DR3, and DR3 is more abundant than TL1A in the actively inflamed parts of the gut.
You then go to the adjacent not yet inflamed tissue, you still have lots of DR3 present, but you do not yet have TL1A. What that means is that there is no means of retaining an anti-TL1A antibody at that margin of inflammation. The reason that is a problem is that as inflammation spreads in these patients, it will then become a race, whether that newly expressed TL1A has the opportunity to interact first with locally abundant DR3, or whether that newly expressed TL1A is immediately neutralized by an anti-TL1A antibody that is passively diffusing through that tissue site. This is the reason why we believe that targeting DR3 may lead to higher rates of complete remission than what has already been seen with anti-TL1A antibodies. This sort of biology is not unique to the DR3-TL1A axis.
If you look at the differing expression patterns of PD1 and PDL1, you walk away with the same conclusions, where in that case, PDL1 is the tissue-restricted inducible ligand. PD1 is constitutively expressed by circulating lymphocytes. This is one reason why the efficacy of anti-PD1 antibodies may be superior to the efficacy of anti-PDL1 antibodies. Turning now to our lead antibody program, SL325. This is a humanized monoclonal antibody that is going into phase one in just a few months. We also have a similar antibody. It has the same DR3 binding regions, but we have now added a series of half-life extending mutations. That's SL425. At the end, I'll show you what the interaction between SL325 and 425 looks like.
I'll also touch on what we think some of the advantages of developing DR3 bispecifics might be relative to TL1A bispecifics. SL325 was designed with a few parameters in mind. First of all, we wanted a super high-affinit y antibody, and we found one that binds human DR3 with a 1.3 picomolar binding affinity. We designed the antibody to spare the decoy receptor so that the natural degradation mechanism of TL1A would remain intact. We found an antibody that binds an epitope that both interferes with trimerization of DR3 and also interferes with trimer-to-trimer binding between TL1A and DR3. Finally, we found an antibody that does not trigger DR3 internalization on cells that express it. That comes through clearly in the receptor occupancy data from our monkey studies that I'll show you in a minute.
There is a lot more information available on our website than I can cover in a short presentation. This is one example of the sorts of assays that we have run in developing SL325. Here, what we are looking at is whether SL325 can interfere with TL1A binding to DR3 and how the potency of that inhibition compares to the potency of the Merck antibody or the Roche antibody. We have subsequently tested the Teva antibody. In all cases, there is about a tenfold improvement in blocking TL1A binding to DR3 relative to those leading antibodies.
Ultimately, the gold standard functional assay for whether you have a good inhibitor of the TL1A DR3 axis is taking lymphocytes from patients with either ulcerative colitis or Crohn's disease, stimulating them with an artificial T cell receptor agonist, in this case, CD3, CD28 beads, and then titrating in TL1A in the presence of your antibody and asking whether your antibody can fully neutralize the co-stimulatory effect of TL1A on these IBD patient lymphocytes. First of all, in orange here, you can see that we can add in our antibody, and it, in and of itself, does not cause any induction of interferon gamma from these T cell receptor-stimulated IBD patient lymphocytes. When you bring in TL1A, you can see a dramatic induction of cytokine production from TL1A. As you titrate in your antibody, you can fully neutralize that TL1A-mediated co-stimulatory effect in patient samples.
We took our lead antibody into GLP toxicology studies in cynomolgus macaques. This study was completed last year. What we did is we gave the antibody at one, 10, or 100 milligrams per kilogram doses. We gave three doses over a month-long period separated by two weeks each, and then had a four-week recovery period. As is typical in GLP tox studies, we're first and foremost looking for signs of any safety events. Fortunately, there were no infusion-related reactions observed in any animal at any point in time. There were no changes in clinical chemistry parameters, gross path, or histopath. The NoAL was defined as 100 milligrams per kilogram, the top dose that we administered to these monkeys.
One of the benefits of targeting a receptor like this is that, because it is expressed by lymphocytes in the peripheral blood, you can draw blood samples from these animals and measure receptor occupancy on those lymphocytes. This is a particularly important question for a first-in-class antibody like this because, as I showed you before, DR3 is more abundant than TL1A. One question was, were we going to have to go to a fundamentally higher dose of antibody to fully block DR3 than what has been seen so far with TL1A? What the receptor occupancy data show is that even at the lowest dose of 1 mg per kg, we achieved full receptor occupancy immediately following the first dose.
We could then look at each dose, whether the antibody remained bound to the receptors following the two-week interdose intervals or the four-week interdose interval in the recovery group. What those data showed is that full occupancy was maintained through at least those intervals and probably longer. This speaks to the fact that this antibody does not get internalized by cells that it binds. Now, 325 is a humanized antibody, and you do expect to see some evidence of immunogenicity in non-human primates. We did. We saw just a couple of animals that developed antidrug antibodies. This actually became a very helpful data point for us because we also noticed that in those two animals, there was a drop in receptor occupancy when the animals became ADA positive.
At that same time, the concentration of free drug in the blood fell below one microgram per mil. In all other animals who remained ADA negative, they retained full receptor occupancy, and the concentration of free drug in the blood remained over one microgram per mil. These data now become extremely helpful in forecasting what trough concentrations might need to look like in a phase one trial. We use that in our POPPK modeling that I'll show you in a few minutes. I think a common question that we get is, why hasn't anybody else gone after DR3? Why is everybody focused on TL1A? One reason for that is that Stephan Targan and the group at Prometheus started with TL1A. Once something works, there tends to be a herd mentality to stick with what works.
As I've outlined, I don't think there's a good biological reason to do that. Another reason folks probably preferred TL1A versus DR3 is that when you're going into any given antibody generation campaign, your chances of success are higher if you want to have a ligand blocker versus a receptor blocker. The reason for that is that all the TL1A blocking antibodies really need to do is interfere with primer TL1A binding to DR3. When you're building a receptor blocking antibody, you also have to make sure that there is no evidence of untoward receptor agonism because your antibody can always potentially bring together at least two subunits of DR3. I think that's the bigger reason that has caused folks to shy away from this or perhaps not be successful.
I showed you some of the preclinical data that convinced us that we did not have any evidence of agonism. Ultimately, the best test of that is looking in non-human primates to see whether there is any evidence of lymphocyte proliferation, activation, or cytokine production a week after administering your antibody. We looked carefully for that in non-human primates. There was no evidence of lymphocyte proliferation or activation of any kind. These data give us great confidence that this antibody is a pure play antagonist. These figures here are looking at the preliminary POPPK models from the non-human primate data. We have highlighted in black the one microgram per milliliter trough concentration that we believe is necessary to maintain full receptor occupancy.
We have looked at what absolute doses would exceed that trough concentration, either at a Q4 week or a Q8 week maintenance dose. What these data suggest is that at a Q4 week maintenance, we will exceed that trough concentration at either one or three milligrams, per kilogram dosing. If we want to go to a Q8 weekly maintenance regimen, we will likely need either a three or six milligram per kilogram maintenance regimen. A caveat to these models as they currently stand, is that we do not yet have a complete assessment of how long the receptor occupancy may actually last in humans. If you look, for example, at patients that were treated with the anti-PD1 antibody pembrolizumab, which also does not get internalized, folks remain functionally blocked for PD1 six months after their last dose.
If that ends up being the case here, then these POPPK models could underestimate what the true absolute dose is to maintain full RO. We certainly do not want to be wrong about that. This is the reason why we are developing SL425 in parallel. As I said, it is the exact same antibody as SL325, except that it contains some half-life extending mutations. As the Healthy Volunteer phase one study initiates for 325, we will also be initiating chronic toxicology studies for both 325 and 425. Chronic tox is necessary no matter what before initiating phase two trials in IBD patients for 325.
What this will set us up to know is in around the second quarter of next year, look at the actual PK data from our single and multi-ascending dose study in human patients, or human volunteers, I should say, and then look at both the safety and the POPPK modeling data from the chronic tox study comparing 325 and 425, and ask ourselves, do we have evidence that 325 can move forward and ultimately be administered no more often than a Q8 week regimen subcutaneously, or are we better off bridging to SL425 to achieve that maintenance dosing goal? That is the reason for aligning these two programs so that 425 could bridge in to the phase two studies at the necessary time if that is the profile that we need.
We're in a unique position in a now clinically validated axis like this where we are the first to be developing an agent to the receptor side, and hopefully, I've convinced you that there are very good reasons to go after the receptor side. In addition to the fundamental biological reasons to target DR3 as opposed to TL1A, one of the things that we also now know is that all of the TL1A antibodies have fairly high rates of immunogenicity. The reason for this is that the TL1A antibodies bind, stabilize, and slow the degradation of serum TL1A. What many companies who have shared data focus on is what happens to the concentration of free TL1A after administration of an antibody. That qualifier free TL1A is likely used because something different is happening to the concentration of total TL1A.
Pfizer actually published this with the antibody that Roche is now developing. What they published is that the concentration of total TL1A goes up two logs after treatment with that anti-TL1A antibody. What you're actually measuring there are immune complexes between the anti-TL1A antibody and serum TL1A, and that is a prominent source of immunogenicity. The secondary benefit of going after DR3 from an immunogenicity perspective relates to the opportunities to develop bispecific antibodies. Amgen was the first to develop a TL1A-directed bispecific. They made a drug called AMG966, which was a TL1A by TNF alpha bispecific. They brought it through a phase one trial and published a really nice paper in Frontiers in Immunology. If you search for AMG966, you'll find it.
What they showed in that phase one study is that every single patient developed rapid, high-titer neutralizing ADA, and they defined the reason, which was that in that case, you had an antibody that with one arm was binding and stabilizing TL1A, with the other arm was binding and stabilizing TNF alpha, slowing the degradation of both. That led to even larger immune complex formation, and thus every patient developed those rapid, high-titer neutralizing ADA. That is a cautionary tale for the future development of TL1A-directed bispecifics that you do not have to tangle with if you are instead going after a membrane-restricted receptor like DR3. We are excited about that opportunity, and we will have more to say about the bispecifics that we are developing soon. Here we are. We will be in phase one in just a few months' time.
We expect to look at initial data from the single ascending dose cohorts by the end of the year, have completed the single and multi-ascending dose study by around Q2 of next year, and be in a position to start those phase two studies in IBD patients in the second half of next year. We currently have cash through our phase one study and into 2027, and so we're certainly looking forward to generating some of that phase one data in humans soon. That's all I wanted to go through for today, and I welcome any questions that people have. Thanks for the chance to speak.
Yeah, thanks for the presentation, Taylor. As a reminder to any viewers who are watching, you can ask a question through our conference portal, ask a question feature. While that list is compiling, maybe I can go ahead and start us off.
I think from our research into IBD, administration convenience has definitely become a priority of late, as in the less frequent dosing you can achieve, the better. You talked about how you have a half-life extended version in the pipeline. Curious, what dosing regimen do you think you need to create in order to be competitive? Are you guys also considering subcutaneous dosing formulation?
Yeah. A few thoughts. Yes, we are developing a subcutaneous formulation, and that is the plan for phase three study. It would be to move to subcu dosing minimally for the maintenance portion of the study. The approved patients with IBD are used to getting treatments every four or eight weeks. You do not want to have a drug that you have to give any more frequent than that.
I think as time goes on, even the Q4 weekly will be less appealing than the Q8 weekly. Those are the goals, and those are the reasons why we're developing both 325 and 425 in parallel. The programs, as hopefully I outlined, are aligned in a manner so that if the data tell us that 425 is the direction to go, we're not looking at a delay there. That being said, there are biological reasons in the TL1A DR3 axis that cause you to question whether blocking the axis permanently may not be a no-brainer goal. It may be a no-brainer goal in IBD, but in lots of the other indications where TL1A DR3 biology is implicated, you have less control over when the offending antigen that's driving disease is present.
In diseases where the offending antigen is not ever present, you may not want to always block TL1A DR3 signaling. I am happy to follow up with anybody who wants to learn more about that. That could be the reason why the same Teva antibody that looks wonderful in IBD failed in asthma. I think folks need to be thoughtful about more than just what is convenient for patients when making these decisions.
That is a good point. When we do, I guess, look at the space, I think another thing that jumps out is there is typically a very high placebo response rate in clinical trials. If you actually adjust clinical remission rates, you usually get about 20-30% is the rule that you usually see.
This is kind of more of a curiosity question, but is there anything that can be done to either reduce the placebo rate, or are you really just relying on boosting the efficacy to kind of counteract that effect?
If you look at, so first of all, you're correct. There's a huge range in placebo rates. Just if you look at the Teva and Prometheus study, it can vary by 20%. One thing that you should do is make sure that you make conservative estimates for what the placebo rate should be and power your studies accordingly. Alimentiv had a really nice poster at Echo this year where they did a retrospective analysis across multiple studies over time, and they zeroed in on about a 9% placebo complete remission rate in UC as the centering point.
You can think about that as an anchoring place for power calculations, but you need to think about the upper and lower bound of that as well. We are working with our statisticians to do that. The other variable that has a large impact on where your placebo rates end up is looking at how the split of patients that are advanced therapy experienced versus advanced therapy naive lands for the folks who enroll in your study. All of the studies that have been done have, I think, put in reasonable criteria to try to put guardrails around not having too many of those folks, not having too few of those folks, and not having too many of those folks is what drives lots of enrollment outside of the U.S.
You still have to just go into this eyes wide open that you could plan for 9% and end up with 18%.
Yeah. Obviously, one topic, or I guess maybe two topics, we've been asking a lot of the companies this week, sort of broader macro. Obviously, there's a lot of uncertainty in the market right now, particularly around two key topics, which are tariffs and recent changes at the FDA. I won't quiz you too hard on tariffs because I know that's evolving by the minute, I believe. Just in general, what sort of impact, if any, do you expect from tariffs as well as recent leadership changes at the FDA?
Yeah. All you have to do is open your eyes today to know that even companies that may not be directly impacted by tariffs like us are still impacted by tariffs.
That can relate to various facets of your business, in our case, raw materials that you're purchasing from different places, wherever you're conducting your clinical trials, how the pharma wallet gets pushed one way or the other to adjust tariffs, and how that affects the M&A appetite around the industry. Everybody's affected by tariffs. In terms of changes at the FDA, we have some recent data points. Given that we're in the middle of submitting our first IND, I can tell you that we've been in the pre-IND interaction process for the last couple of months. I'm happy to report that we weren't expecting a response on our pre-IND questions until later this month, and we actually got them today. That's great. Does that stay around forever?
I hope so.
So far, so good.
Yeah.
I think having those data points is reassuring to some investors out there because it's really the best you could hope for at this point is just things continue as normal, and the more recent interactions you have to justify that, the better. I am looking at the chat. I do not see any further questions at this time. I will give you the floor. I think you already kind of summarized where the company stands and what you're looking for. Any final thoughts here?
Yeah. I mean, look, we are heads down, glad to be close to the clinic, excited to get those first data with a first-in-class DR3 antibody, and look forward to engaging with folks who are watching and our investors moving forward. I think we've got something really special here, and that's going to unfold over the course of the year.
Great. Thanks, Taylor.
Thanks for attending the conference. It's always good to have the conversation. Thanks.
You too.