Good afternoon, everyone, and thank you for joining me at the first day of the Needham Healthcare Conference. My name is Gil Blum, and I am a senior biotech analyst here at Needham & Company, and I cover the immuno- oncology and gene therapy subsectors. It is my pleasure to have with me today Taylor Schreiber, the CEO of Shattuck Labs. And as a reminder for everyone watching, viewers can ask questions through the conference portal, and they're able to submit questions at any given time. So without further ado, Taylor, I think the best place to start is maybe the broader I/O picture and kind of, you know, given general investor disappointment with anti-CD47 and other assets in the I/O space. How do you think that data from Shattuck and others could revitalize interest here?
Sure. Well, Gil, thanks for including us in the conference. Happy to be here. The I/O space has changed in a few different ways, in some ways generally, and then in some target-specific ways as well. You know, I think one of the takeaways, if you think back over the last 5-10 years since PD-1s were first approved, is that blocking other adaptive immune checkpoints, TIM3, IDO, TIGIT, LAG3, you know, the list goes on with at least six other targets, you know, really either hasn't worked at all or has provided incremental benefit at best. Those are, you know, any of those targets really don't have monotherapy activity, and some of them seem to have some incremental activity when added to a PD-1 backbone.
If you look at how the landscape has changed outside of checkpoint targeting over the same interval, you've seen the success of different T-cell engagers. You've seen the success of certain CAR-T approaches. And I think those two clinical learnings set side by side, paint a picture that, at least as it relates to adaptive immune responses, PD-1 is unique. But then when you look at the CD3-based CAR-T cell engagers and CAR-T data, it paints a very clear picture that, in patients where checkpoint mechanisms are in place, that those mechanisms can be overcome with an appropriately activated immune response. And activating CD3 is one way of doing that, but there are a number of other agonist receptors that might be able to do the same thing.
And that's really where Shattuck's platform arose, was through the desire to pair a checkpoint inhibitory domain with an immune agonist domain. And, you know, then I'll get back to that in a minute, but then there are CD47-specific learnings that have really been a roller coaster over the last five years where the early data from the company Forty Seven and Trillium ran into difficulties. And, you know, we can get into this as well in a few minutes. But Magro was discontinued because of heme tox, plain and simple. And, unfortunately, you know, I think from a long-term development perspective, those were toxicities that could be easily managed in small phase 1 trials with dedicated investigators. But once that molecule found itself in a more real-world setting, those dose-limiting anemias were much more difficult to manage.
You know, that's the setup that we and other companies are stepping into, with maturing data with our lead molecule in the CD47 landscape, SL-172154. You know, this molecule is different than other CD47 inhibitors, both by virtue of the fact that it has an inactive Fc domain to avoid dose-limiting cytopenias, but then it also has this immune agonist domain, CD40 ligand, that it really may differentiate from any prior CD47 inhibitors if we can demonstrate that agonist potential of the molecule does something more than CD47 inhibition on its own. So we've got data coming soon where we hope that that will become clear to folks on the outside.
Okay, excellent. Maybe this is a good starting point to kind of discuss who even remains in this space now with magrolimab kind of being discontinued. How many of your peers are still operating at this point?
Yeah, so, I mean, the ALX Oncology is, I think, the one that is doing best at this point in time aside from us. And we and ALX were the 2 companies that made the design decision to engineer out Fc gamma receptor binding for our molecules. And that came with the downside that in early clinical development, there was no expectation of monotherapy efficacy.
Mm-hmm.
but it came based on the decision was made based on the conviction that we both had clearly that if you engineer out that Fc gamma receptor binding function, you would avoid the dose-limiting cytopenias that have led to the discontinuation of so many other molecules in the space, including magrolimab. The other, so there's ALX. There's also now Pfizer with the Trillium assets. And they've been very Pfizer's been very quiet about what's been happening with those assets since they acquired Trillium a few years ago. And the first update, if you can call it that, that we've received from them was in their R&D day last month where they announced that they discontinued TTI-621. So that was the IgG1 SIRPα fusion protein.
Makes sense.
But the TTI-622, which was their IgG4 SIRPα fusion protein, was being prepared for a phase 2 study in AML and MDS. So we don't know what that means yet, but that was what they disclosed. And then it's unclear exactly what I-Mab might be doing with lemzoparlimab in China. The AbbVie collaboration clearly unwound there. But that's it at this point. And you know, I remember the slides from 3 years ago where there were 6 or 7 companies with Fc active molecules and 2 with Fc inactive. And you know, it's interesting that the 2 with Fc inactive molecules are really the 2 that remain today.
So this is a great segue. What do we know about what happened with magrolimab as it relates to the safety data? And as a, you know, as a good proxy to what could happen with any Fc active, CD47 targeting agent?
Yeah. So there's been an interesting set of disclosures here. And, on February 7th, Gilead had a company statement where they announced that magrolimab was placed on a full clinical hold and that the ENHANCE trials were being discontinued as a result of that. And they said in that initial company statement that they would soon be sharing the data from the ENHANCE trials in a publication or a conference. And so anybody who was following magrolimab was clued to that company statement and read it and had no reason to go back and look at that document again. What Gilead did is, 3 weeks later, in late February, they updated that company statement without reannouncing that anything was added to that statement. And what they updated it with was a link to the top-line data for ENHANCE- 2 and ENHANCE- 3.
What those data show are that the 60-day mortality in the Magro arm on ENHANCE and ENHANCE-3 was about double the 60-day mortality of the control arms. In ENHANCE-2, so ENHANCE is in high-risk MDS, where the control is AZA. ENHANCE-2 was in TP53-mutant AML where the control was AZA plus venetoclax. So in ENHANCE-2 and TP53-mutant AML patients, these patients are sicker anyway. And there's an expected higher early mortality rate for those patients, sadly. So in the first 60 days, the Magro arm was pretty similar in terms of the 60-day mortality to the AZA ven arm. But then there was this precipitous drop-off after about four months on the Magro arm but not on the AZA ven arm. That drop-off is not yet explained by the data that's there.
It suggests that there must have been a high rate of discontinuations or dose interruptions early on study for the magrolimab arm and that somehow those patients weren't even receiving the appropriate chemo standard of care. That's speculation, but it's hard for me to figure out why else there would be that late drop-off in the TP53 mutant arm. When you look at those data, there's also a description of the adverse events, which, and grade 5 events. And, there's a number of cardiac events that are listed. There's a number of, you know, general inflammatory events that are listed as reasons for death. There's bowel ischemia, macrophage activation syndrome in some patients. What's missing is anemia.
But the reason it's missing in the description of causes of death is that anemia is not a cause of death. It's the things that are secondary to anemia which are causes of death. And the other piece of information that's available is what appears to be a patient safety document that may have been provided to patients who are enrolled in the study that tabulates the related adverse events on the Magro plus chemo arms versus the AZA ven arms on these studies. And when you tick through the various columns and try to ask, you know, what is it, what is either disease-related or AZA or AZA ven-related and what's Magro-related, the only categories that are where the adverse event profile is different are in anemia and anemia-related events, and infusion-related reactions.
And that's those are the things that Magro added that you don't see in the AZA or AZA ven arms. So, it's been an odd pattern of information release, but the data are out there, and it confirms what has been long suspected that, you know, unfortunately, Magro caused very significant anemias, and those anemias seem to be the cause of early mortality, and ultimately the demise of Magro once that drug found itself in a more real-world setting. I hope that they share more at EHA or some other meeting. I think that would still be helpful. But for us, it paints a very clear picture that, you know, the reason that 154 now has a first-to-market opportunity in AML and MDS, is because Magro was discontinued not due to the target but due to the molecule being toxic.
At the end of the day, I think that will be helpful for us and helps us understand why enrollment on our studies has been so brisk.
Mm-hmm. Maybe a bit of a hypothetical there. And given the strong signal for anemia and anemia management, could this be driving some of the toxicity in the magrolimab studies?
Yeah, I'm speculating here, but, you know, we know from the phase 1B study publications that came out in the Journal of Clinical Oncology, in the 3rd quarter of last year that there was a very high rate of transfusions during the 1st cycle of patients who received magrolimab. And, you know, and also, we know that one of the clinical protocol changes that was made after the very 1st partial clinical hold in January of 2022 was a requirement for patients to have a hemoglobin of 9 g/dL to be eligible for enrollment. And so, you know, there's potentially some hypertransfusion before the 1st dose, and then there's transfusion after the 1st dose. And so that's a lot of blood products that you're giving to patients.
You know, it potentially that can create some fluid overload, types of things that, you know, are always difficult to manage both from an infection and a pulmonary, pathophysiological, perspective that I don't know. I can't say whether those things played a role, but you'd think about it.
Okay, that all makes sense. Now, maybe kind of focusing on, you know, 154 and that class of immune oncology agents. You know, it's the lack of, you know, upfront toxicity and, you know, Fc activity is a little bit of a double-edged sword when it comes to investors, right? It's hard to drive responses. Have you been able to as a single agent, I mean, have you been able to kind of overcome some of that skepticism with your data?
Maybe a little bit. I think we're about a third of the way there. And, you know, we had in the data we shared at the end of last year, we had one really nice monotherapy response in a treatment refractory TP53 mutant AML patient who was bridged to transplant and remains disease-free. It's a great outcome for that patient. And I don't believe any of the other CD47 inhibitors had a monotherapy response in a patient like that. But, you know, the data that caught people's attention was the front-line expansion data in both TP53 mutant AML and high-risk MDS. And, you know, those were two of the three arms that we built into that clinical protocol. The third arm is an AZA venetoclax combo arm in wild-type AML.
The reason that we began with the first two AZA-only arms is that our primary challenge a year ago was demonstrating that 154 is adding something to AZA alone. That's harder to do in an AZA ven combo where the AZA ven backbone leads to very high response rates on its own and longer response durability. But in the TP53-mutant AML population, it's expected that AZA gives you a complete response in less than 10% of patients. One good aspect of the VIALE-A data now being available is that it is now the largest controlled dataset for what you would expect AZA plus venetoclax to do in a TP53-mutant AML patient. The AZA ven complete response rate was 14% in that study of 250 or so patients.
And so that aligns with the historical benchmark of less than 10% if all the patient gets his AZA. And, you know, in our early data, we were trending about double that. And it was an interim look at the dataset. And there were a number of patients that had stable disease but where the blast counts had gone from 29% to 5% or 28% to 5%, and heme counts were recovering. And, so, you know, if we capture a few more of those patients, then it would become even more distinct from, the expectations for AZA or AZA ven. And I think, you know, we're trying to set high benchmarks to proceed, because of the historical challenge of, you know, needing to be stringent about differentiating contribution of components in studies like this.
We're trying to do the same thing in the high-risk MDS population where, you know, first of all, out of the 24 patients we've enrolled, 20 of the 24 have TP53 mutations, and 19 of those 20 have TP53 mutations with complex karyotypes. So this is a group of patients that is biased strongly for negative prognostic factors. And so there's nothing about these patients where you could look at it and say they got lucky, in terms of the patients that were enrolled. And similarly, on the enhanced trial, again, now the largest dataset for what you'd expect, AZA to do in a high-risk MDS population, the CR rate, it was 24%. And that's about what we've thought it would be.
And it's higher than, you know, what 47 and other companies had been citing in the past of AZA alone and giving you a CR in the high teens of patients. And, you know, again, we need to see CR rates well north of that. And, you know, that's part 1. And part 2 in both indications is seeing that CR rate translate to response durability. So, that's all the context. And the reason why I think we're, you know, maybe a third of the way there is because we haven't shared that durability data yet or the full CR rates. And that will be coming at EHA in June.
Maybe to kind of set expectations for EHA, any assessment of the number of patients and, you know, in relationship to what has been disclosed so far?
Yeah, so we have amended the TP53 mutant AML study back in December, to double the size of that cohort. And those additional patients were enrolled very quickly. We finished enrollment at over 20 patients, in early February. And so at EHA, we'll have 20+ TP53 mutant AML patients. We'll have the CR rates for all of those patients and maturing duration of response. TP53 mutant patients have a median overall survival of somewhere around 7 months treated with AZA plus or minus venetoclax. And somewhere around 14% of those patients are effectively bridged to transplant, and so we'll be able to compare against all of those metrics, in June. In the high-risk MDS, we have 24 patients on study that we also over-enrolled that cohort. The original target was 20.
We'll have the full CR rates and maturing response durability for those patients in June as well. Then, the other study that has just gotten underway, many people listening have probably heard of Project Optimus at FDA. The heme division in particular at FDA is asking most companies like us to address Project Optimus. They have not asked us yet, but we suspect they will at the end of a phase 1 meeting. Rather than wait for that meeting, we've decided to just proactively address it. So we're enrolling patients now. We'll enroll about 20 patients with high-risk MDS to an AZA-only control arm, about 20 patients to a lower dose of our drug plus AZA, and then 20 patients to the current dose of our drug plus AZA.
We expect that we'll finish enrollment, you know, around the end of this year and that that will be the last significant box we have to check before beginning the 1st registration-directed study.
Speaking of a registration-directed study, have you started considering some of the study designs? Are we talking about a relapse refractory single-arm sort of study or defined populations? I mean, you have your TP53 mutants, et cetera.
Yeah, we have. And we will be having conversations with regulators over the course of this year to help better refine this. But, we believe, and many of our investigators are sharing that it may be possible to enroll TP53 mutant MDS and TP53 mutant AML patients on the same protocol. And, the expectation there is that the primary endpoint is overall survival. However, that's one of the topics that we'll have to discuss with regulators this year is, you know, if based on our current study data, we're seeing, you know, particularly high response rates or particularly high percentages of patients being bridged to transplant, you know, would there be a surrogate endpoint that could be considered there? The assumption based on recent experience with Takeda's pevonedistat study and Aprea's study is that it will be OS, but we'll see.
anyway, so.
Yeah, OS isn't that bad in this indication. It's not those studies are not wrong.
That's right. That's right. So the working design is that we'll enroll TP53 mutant MDS and AML onto one protocol, and it will be controlled against a physician's choice azacitidine plus or minus venetoclax. And in TP53 mutant AML and MDS, venetoclax can incrementally improve CR rate, but it does not meaningfully move overall survival. And so that's the reason from a trial efficiency standpoint, it would be fair to give docs the option of using AZA or AZA ven, there.
All right, makes sense.
So we'll see. But that's the idea.
Maybe an additional item here. Outside of the very clearly important efficacy data, would there be other presentations this year for biomarker and immune activation? Personally, we're really interested in seeing additional evidence for the contribution of CD40 agonism.
Yeah. So we've been collecting bone marrows on our current studies, and we've shared a little bit of data as to the fact that we've observed, first of all, lots of our drug in the bone marrow, but also signs that myeloid cells have been activated in the bone marrow and that their polarization state has shifted from an immunosuppressive M2 to an M1 phenotype. And we expect to build on that data at the EHA presentation. And that's another one of the nice aspects of now running the controlled study with an AZA-only control arm and then 2 dose levels of 154 plus AZA. We'll be able to look in those patients also at a number of both peripheral and bone marrow biomarkers to help paint a clearer picture of what CD40 agonism is adding in those folks.
That won't be the dose optimization study data won't be available at EHA, but we expect it will be available over the course of the year.
It'd be interesting to see if there's a dose dependence given, you know, immune oncology. That's not always the case. Sometimes it will take.
There has, you know, in the periphery, there clearly was a dose dependence to receptor occupancy, to the magnitude of peripheral cytokine increases, to the magnitude of margination of CD40 expressing cells.
That's right.
So we believe that we'll, there's a good chance we'll be able to see those findings carry through.
Excellent. I do want to spend some time on the ovarian cancer program. I mean, the other big portion of data updates that we're waiting for, towards mid-year. So your initial results that were reported back in November, you had an OR of about 27% with three out of 11, and those have since then been confirmed PRs. Where would you say the bar lies for upcoming data update, with more patients and more follow-up?
Yeah, so also in this study, we're trying to set stringent benchmarks. And what we've been guiding toward is that if we don't see a response rate of at least 25% and a duration of response of at least five months, that it's not an interesting result or it's not a medically meaningful result in the PROC setting. You know, you could have a lower response rate and say, "Well, it looks like your drug's doing something," and that statement to be true. But it wouldn't be medically meaningful when you compare, you know, the other treatments that patients with platinum-resistant ovarian cancer can get. So those are the benchmarks we have to hit. And if we hit it, I think there's a path to develop that Doxil combination further.
And the same is true in the mirvetuximab combinations, where we have, of course, patients that are being enrolled across the spectrum of folate receptor alpha expression and, where, you know, the response rate for mirvetuximab alone in the FR-alpha high patients is 42%, with about a 5.9-month duration. We need to be meaningfully higher than that, in the FR-alpha high population for that result to be interesting and worth further development. And same thing in the FR-alpha low and medium patients, where, of course, the benefit of MIRVA is much lower and much shorter duration than it is in the highs.
Maybe probing the hypothetical situation, which does happen in immune oncology, you could maybe not see a ton of activity upfront, but you may have that, you know, profile that's very typical in immune oncology. Would, you know, long duration of tumor-free intervals for PROC patients support continued development even without a strong OR?
Probably not for me.
Okay.
I think, you know, if you're a Roche, she can wait around for a long time. But I think for a small company, that's a tough sell.
Okay, that's very clear, actually. As to the Elahere collaboration, I mean, how should the data be interpreted, you know, given the patient enrollment includes a variety of FR-alpha expression levels? I mean, you're hoping to see some added activity on top of what the drug has in its label, I'm assuming.
Yeah, we are. And, you know, it's it's relatively straightforward using Mirasol as the comparator in the FR-alpha high patients. ImmunoGen and now AbbVie have, you know, done a really nice job looking back at some of the older study data, the Soraya data and other studies, using the PS2 assay and benchmarks have used that to benchmark the expected response rates in the FR-alpha low and mid patients. And, you know, these are their data, and it's not my business to share it before they do. But I can just say that those data are available, and those are what we're using as benchmarks.
Do you expect the data will be stratified by FR-alpha expression?
We do. Yeah, we have patients clearly bucketed by 25%-50% as low expressors, 50%-75% as medium, and 75%+ as highs.
The same question I had for the heme programs. Do we expect to see additional biomarker data also in the ovarian cancer in the next data update?
We will. So there too, we've been collecting biopsies and peripheral blood on these patients, and we will update those data packages as well.
Okay. Maybe a broader overall overarching question here. Is there a potential for a chemo-free solution given how your asset works, meaning, let's say, you combine it with a, you know, cytotoxic antibody?
Yeah, you know, I, that's long been part of the CD47 mechanism. There are certain chemotherapies that cause.
More presentation.
When tumor cells are exposed to certain chemotherapies, they cause those cells to upregulate stress signals that provide a prophagocytic signal to a macrophage. Another clear way of doing that is painting the membrane of a tumor cell with an antibody, a tumor antigen-targeted antibody that has an active Fc domain. That's really what the thesis is behind ALX's trial in gastric cancer. That will provide a prophagocytic signal as well. So there's a number of other tumor types, particularly solid tumors, where that would be the desired combination mechanism. But it's important to recognize that there's a number of different ways of causing different tumor cells to be decorated with a prophagocytic signal.
All right. I know you guys are presenting this week at AACR. Maybe it's worth kind of mentioning a little bit of your earlier-stage programs. This one's specifically about the setting of acquired PD-1 resistance. Well, maybe give us a bit of the highlights here.
Sure. Yeah, I haven't been this proud of our team in a while, and they routinely do really good work. So this is. We had a paper come out in Cancer Cell in January that was a collaboration between Matt Hellmann's group at Sloan Kettering and subsequently AstraZeneca, where Matt had been collecting lung cancer biopsies from patients at baseline and then when they developed acquired resistance to a PD1 inhibitor. We had, in parallel with that work that Matt was doing, developed a number of preclinical tumor models that were also PD1 acquired resistant. We had been sequencing the heck out of those tumors and looking at them every way that we could think to try to define the phenotype and ultimately the drivers of PD1 acquired resistance.
And it led, it showed that these tumors were, had a transcriptionally hyperactive interferon signature. And it perplexed us because that's exactly the opposite signature you would expect to see in the setting of PD-1 acquired resistance. And lo and behold, that was exactly the signature that Matt had been observing in lung cancer patients that were acquired resistant to PD-1. And so the Cancer Cell paper pulls together those two data sets and demonstrates really clearly the interferon hyperreactive biology that seems to characterize PD-1 acquired resistance. Since those two strands came together, because we had this preclinical model, we could continue to interrogate it to try to identify the drivers of PD-1 acquired resistance. And it led us to an E3 ligase called TRIM7. And TRIM7 is; there's not much known about it.
It's a very, very interesting protein that ubiquitinates at least three different client proteins. One is a protein called RACO1. The way that TRIM7 was first described was with a U.K.-based group. What they were looking at is to try to figure out what were the proteins that were essential to drive proliferation downstream of mutant KRAS. Lo and behold, TRIM7 was the key protein that downstream of KRAS had to ubiquitinate RACO1, to enable RACO1 to remain in the cytosol longer and drive activation of JUN and AP1. The other two proteins that TRIM7 ubiquitinates are MAVS, which is the mitochondrial antiviral signaling protein, and STING. MAVS and STING are both key modulators in the interferon pathway.
And, you know, remembering that we were led to this protein through PD1 acquired resistant phenotype, that signature makes all the sense in the world where a tumor cell that upregulates TRIM7 is going to have a stronger proliferative ability due to stabilization of RACO1, and it will become resistant to interferon through degradation of STING and MAVS. And so the presentation that George Fromm, one of our CSOs, will give tomorrow at AACR describes, you know, some of that biology of how we were led to TRIM7, but then importantly, that we've now developed small molecule inhibitors that bind the E3 ligase binding pocket in TRIM7 and are very, very potent inhibitors in nanomolar potency, covalent and noncovalent versions, and directly inhibit proliferation in both KRAS mutant and KRAS wild type human tumors.
Inhibition of TRIM7 in some of those KRAS mutant tumors seems to override the need to inhibit KRAS in those tumors we've compared to the Mirati and a few other compounds. Then when you put some of these tumors back in vivo and inhibit TRIM7, it restores PD-1 sensitivity in these PD-1 acquired resistant tumors. This is, I think, a beautiful story of discovery that has now led to a development candidate. I'm really excited for folks to see the data.
Yeah, so super, super interesting. It would be interesting if by disrupting this, I mean, it seems like the cytokine picture for a PD1 resistant is, shall we say, more immunogenic than one might expect. So if you remove the blockers there, maybe you get even, you know, a resensitization.
Yeah. Yeah, it's going to be great. And this is a molecule that you expect will have clear monotherapy efficacy, even in KRAS mutant tumors, even in a PD-1 resistant patient. So that there's a number of development directions that something like this could take.
Okay. So kind of a last point for investors, just remind us of, you know, cash position and runway.
Sure. So we had, as of year-end, about $130 million in cash. The guidance is that that will last into 2026. And that includes the funding of the dose optimization study that I discussed in high-risk MDS. And it also includes some of the long lead manufacturing items. We've developed a commercial manufacturing process for 154, and it includes tech transferring that process to our CDMOs, so that we're ready to begin those registration-directed studies next year.
Excellent. We're at time. Thank you very much for attending today, Taylor.
Thanks, Gil. It's great talking to you. Take care.