All right, Welcome once again to the 45th Annual TD Cowen Healthcare Conference. I'm Yaron Werber from the Biotech team, and it's a great pleasure to introduce Foghorn Therapeutics with us today, Adrian Gottschalk, CEO. We have Karin as well, and I know Cristian is coming in as well. Adrian, I think you have about 20-25 minutes of slides, and then we'll do Q&A. Thanks for coming.
Thank you, Yaron, for having me, especially with the Phil Collins soporific music that was playing. Very relaxing. Hopefully, I'll be a little bit more stimulating than the music. If I catch you nodding off, I'll try and speak more loudly or more interestingly. I'm going to tell you a little bit about Foghorn today. I've got some slides. I'll try and make sure we leave some time, as Yaron said, for Q&A. Certainly welcome questions at the end. If you have thoughts, please don't hesitate to ask when we get to that portion. Obviously, making some forward-looking statements today. You guys can read the Safe Harbor statement at your leisure. The way to think about Foghorn is an entirely new way to get at gene expression within the cancer space and potentially beyond that.
We've been focused for the past eight or so years on the area of chromatin-related biology, with the notion that tackling this relatively underserved, underexploited area would lead to new therapeutics and new ways to impact disease. We've been doing this through small molecule drug design and development. As you'll see throughout the presentation, we've actually been very focused on protein degradation as one of the key mechanisms to get at this particular area. It's also allowed us then to unlock some of the more difficult drugs that we believe the industry has failed time and again to actually selectively drug. It's also leading us into new areas. I won't be speaking too much about this today, but other areas of induced proximity, which is an expansion to our platform.
Maybe just to level set then in terms of what this biology is doing, here on the slide, you see a very oversimplification of regulation of gene expression. For those of you not familiar with the space, chromatin is simply the compacted form of your DNA. You take about three meters, all those ATGCs in the appropriate sequence. It gets smushed into the nucleus of the cell. It is about a million times compression ratio. You need some sort of molecular machinery to unpack that DNA, to expose it to transcription. When all that is working well, you get beautiful gene expression. You get healthy cells. Your cardiomyocytes remain cardiomyocytes. Your brain cells are brain cells, et cetera, et cetera. Unfortunately, when this goes awry or mutations in the system occur, you end up with disease. That is indeed what happens in the case of cancer.
Cancer hijacks this machinery for its survival, and consequently, you end up with some serious disease. You can see here some of the different pieces and parts that we're working on. These include the modeling complexes, transcription factors, helicases, and other sorts of elements. What we've been doing with Foghorn is we've gone after very novel targets. We're not a me too or trying to improve on stuff that's being commercialized. These are all targets that previously have been relatively undruggable. We've done this now not once and not twice, not even three times, but multiple times, whether that's SMARCA2, which I'll spend some time talking about, CBP, EP300, and ARID1B. These are all meaningful synthetic lethal relationships in cancer and ones that have basically been not well drugged previously.
I won't spend a ton of time on our platform other than to say what has allowed us to be successful is really the integration of three different domains: a deep understanding of the chromatin biology and what creates these cancer dependencies, a whole series of assays that we've constructed, be it biochemical, biophysical, functional, that allows us to study this different machinery in the right biological context. That word context, I would underscore, is really important. That's what allows us to find this very selective chemical matter. Lastly, integrating this, and it's a circular sort of system, as we learn as we go, but really allowing us to then find the right chemical matter. As I mentioned, we've been very exquisitely focused on the graders for the recent past of the company. This slide is not meant to be read very quickly.
Folks can get it online. This is really highlighting the principal components of what we've been doing in the field of protein degradation. Quite frankly, I would put us up against any of the pure play degraders in terms of what our capabilities are here. I'll touch on some of the unique things that we've done when we talk about the programs. There have been really three chapters, if you will, to the story of Foghorn. The first was sort of as we were just validating, could we even drug some of this underlying biology? Good news, we've done that. We've done that in spades. The second chapter really was around trying to move things into the clinic. We did a very large collaboration with Eli Lilly that was back at the end of 2021.
This was a $380 million upfront, 300 in cash, $80 million in equity for what were preclinical targets at the time. I think it was the largest preclinical oncology deal done at that time. It may even still be the largest deal done to date. Now we're obviously in the show me or the execution phase of putting things into the clinic, whether that's our 909 program or the things that are coming down the pike. This is really the resulting pipeline then that has come from our labor over the last several years. I'm going to spend today talking about SMARCA2, CBP, EP300, and ARID. I'll try and cover this briefly again so we have some time at the end. As you can see here, Lilly features very prominently in terms of our partnership. We're currently cost sharing 50/50 on all the SMARCA2 programs.
That's both the inhibitor that you see at the very top, as well as the protein degrader that's listed in that second section. We're hopeful that the second target that's presently undisclosed, we're hopeful at some point this year that that could be a disclosed target. Obviously, that's going to be up to our colleagues at Lilly. Lastly, again, just to reiterate on the deal structure that we did, we do enjoy 50/50 economics on SMARCA2. That's basically in the United States, we'll have a share of the profits, knock on wood, once the programs make it that far. We have three undisclosed discovery targets that we're committed to working on with Lilly. Lastly, in terms of what to expect from this year, we have 909 in the clinic.
I think it's reasonable to think about data from that program towards the end of this year into the first half of next year. We'll have a meaningful disclosure at AACR around preclinical combination data. There, the collaboration will be showing the 909 program in combination studies with pembrolizumab, with KRAS, as well as with chemotherapy. We look forward to disclosing that. CBP has the potential to be in IND and into the clinic by the time that we get to the end of this year, into the first half of next year. EP300, you should think about that as being about six to nine months plus minus behind where CBP is. ARED1B, which is a really important target, akin to what SMARCA2 is, really no progress from others there that we can tell. We've achieved selective degradation.
We're going to be talking more about that program and the preclinical data and path forward at some point this year. I'll now shift to talking a bit about the programs. I'll move through this relatively quickly. For SMARCA2, this is an ongoing phase one dose escalation study. We have two approaches here. One is inhibitor that I'll spend most of the time talking about, and the other is the degrader, both of which are obviously in partnership with Lilly. This is a major synthetic lethal relationship. I recently looked just to compare it to BRAC and PARP. This is equally as synthetic lethal, if not more so than that relationship. Obviously, PARP inhibitors have become very successful drugs in their own right.
Here, what we're looking for are patients who have mutations in SMARCA4 and who have lost that protein, therefore conferring a dependency on the SMARCA2 protein that we're aiming to drug. This mutation is found in about 5% of all solid tumors. It's mutated in about 10% of all non-small cell lung cancer patients. It's about 22 or so thousand patients per year on an incident basis in the United States alone. While we're doing a phase one that's taking all Histologies, the purpose will ultimately be to enrich for non-small cell lung cancer patients as we get into expansion. We know that these patients who harbor SMARCA4 mutations have a worse prognosis than those who don't. We also know that these are patients who don't respond as well to immune checkpoint inhibitors. The unmet need here, unfortunately, remains relatively high.
These patients, as you can see from the chart on the right, this is from Memorial Sloan Kettering Cancer Center, have minimal overlap with most actionable non-small cell lung cancer mutations, the exception being roughly three in ten patients who have a SMARCA4 mutation also harbor some sort of KRAS mutation. The molecule that we've developed that, again, is in the phase I is roughly a 33-fold selective inhibitor, meaning that it's preferentially inhibiting SMARCA2 over SMARCA4. That's what's shown on the slide based on comparing median IC50s between wild type and mutant cell lines. The next few slides are just the preclinical data that we had disclosed last year at AACR. This was the basis for the excitement for us and for Lilly in wanting to move this forward. In this particular model, we're showing a regression. We show that the drug is actually well tolerated.
These body weights, by the way, I do not show the body weights for the next three models on this slide, but it is basically equally and well tolerated for all of these. We see a mix of tumor stasis as well as regression. Again, this is as a monotherapy agent. Very briefly, just on the trial design, as I have mentioned, we are doing all Histologies, but you are required to have a SMARCA4 mutation or demonstrated loss of the SMARCA4 protein. There is an opportunity to backfill patients as we get up into higher cohorts where we would enrich for non-small cell and we would enrich for specific loss of function mutations. This study presently has about 20 or so sites open. Five or six of those are in Japan. The remainder are in the United States.
Lilly is responsible for the day-to-day execution of this, and they're doing a fantastic job. Just as a reminder, when we get to late April at AACR in Chicago, the collaboration will present the combination preclinical data. We'll have an analyst investor call at that same time to go through what those are. The last thing I'll say then, just the other part of this program is the degrader. These data actually predate the collaboration for confidentiality reasons. I can't share the progress other than to say we've maintained the selectivity profile that you see on this page in terms of almost infinite selectivity for SMARCA2 over SMARCA4. We're very excited by the inhibitor, but we've got another great molecule in the wings with the degrader and hope to be able to say more about that in the future.
I'll shift now briefly to talk about CBP, then EP300 and ARED. CBP and EP300 are both histone acetyltransferases, fancy scientific way of saying that they're also regulating gene expression based on moving acetyl groups around. You can see that sort of in the right-hand diagram. The field has been able to successfully drug these targets, but in so much as they're doing it, it's in a dual fashion. We have very compelling and interesting data from our colleagues at CellCentric, a privately backed company. They showed some very interesting data at ASH this past December in multiple myeloma. Unfortunately, the liability that comes with drugging both CBP and EP300 is an impact on the heme compartment. Specifically, you see impacts on platelets or the ability to clot. You see thrombocytopenia. You see an impact on neutrophils.
You see neutropenia and in some cases, anemia as well. What we've been able to do with both of these targets is selectively drug them and degrade them. I'll show you some of that data and what we find interesting here. I'll start with CBP. This is setting up a synthetic lethal relationship. Again, think about the same thing that's happening with SMARCA2, where EP300 is lost and now the cancer is dependent on CBP. You can see a range of the different cancers that may be applicable here. We've been able to demonstrate, this is all in vivo data across lung, bladder, and some gastric models here. You can see that we're basically able to flatline the tumor or get regression depending on the context here with this degrader. This is a 50 picomolar potent degrader.
It's got over a thousand-fold selectivity for CBP over EP300. More importantly, we've applied, and this goes back to the investment we've made in our degrader platform, we've actually applied long-acting formulation technology. Here you can see you're getting well out to over 500 hours on a single dose for this in an intramuscular formulation and then also a 10 mg compared to a daily 10 mg subQ. Our goal for this program is to get it into subQ formulation. That's currently where we are. We certainly imagine we could start off with a once-a-week subQ, potentially get to something that's less frequent and very compatible for patients. Importantly, as I mentioned, the safety and tolerability side effects that come with dual inhibition. Here we're showing our selective degrader, FHT-CBPD9, as compared to two dual bromodomain inhibitors. One is a Genentech molecule.
The other is an OverDib, which is the CellCentric molecule. They are shown in light blue and gray respectively. We know we are sparing megakaryocytes here. We are avoiding those limitations. I am now going to shift and realize I am going quickly, but we will make up some time for the Q&A. EP300, again, the reciprocal protein here to CBP. Here we see not only the synthetic lethal approach where we are losing CBP now, but we also see a frank dependency in some of these tumors like AR-positive castration-resistant prostate cancer, as well as heme malignancies like multiple myeloma and diffuse large B-cell lymphoma. Indeed, when we look at this molecule across a range of heme malignancies, as you can see here, all the way from multiple myeloma to T-cell lymphomas and Hodgkin's lymphoma, et cetera, you can see a pretty profound effect.
I would highlight multiple myeloma here is basically sensitive across the board. Here we show some of the preclinical work again where we're seeing nice regression or stasis, generally very well tolerated across these models. I call out here a comparison to the CellCentric molecule that I just mentioned. We've expanded and blown up the scale here. We sort of cheated in this model. The CellCentric molecule is dosed on a four-day on, three-day off cycle in part, as I understand it, to manage some of the side effects. Here we cheated and we actually dosed it every day. We dosed it on a complete regimen. We still outperform that molecule, but again, without the tolerability safety liabilities.
The other interesting piece for us with this program, we've actually come across a potential biomarker for diffuse large B-cell where we see about 60% of diffuse large B-cell lines being responsive to an EP300 degrader. We haven't yet disclosed the protein's genetics that we're looking at here, but we think this gives us an interesting path forward into that competitive space. Lastly, as I mentioned here, this is an important enough approach to us that we've also been able to develop an oral degrader. I was showing you data from a VHL-based degrader. We actually have an oral version of this molecule as well and compares very favorably again as we look at the PK, but also on the safety-related profile.
The last program that I'm going to talk about before Yaron, and I do some Q&A and are happy to take questions from any of you is our ARID1B protein degrader platform. This is a target that is very, again, similar to SMARCA2 in the sense that the industry for a long time has been banging their head against the wall trying to find a way to selectively perturb this particular protein. The opportunity here is similar to SMARCA2. It's about 5% of all solid tumors. I would dare to say that it could be potentially larger than SMARCA2 given the implications of this. This is one where every once in a while I get a call from a friend or a colleague asking us where we are because they've had a family member or friend diagnosed with a cancer that happens to have an ARID1B mutation.
This is very, very real and the unmet need here is very real as well. What we've been able to do, and I apologize, this may not be satisfying because we're not showing any of the cool degradation at this point, but we've been able to find very selective binders to ARID1B. That may sound like that's not a meaningful accomplishment, but ARID1A and 1B are very, very similar proteins. They're very large proteins. Just the fact that we've been able to find that through our platform was a major breakthrough. We've subsequently turned those actually into degraders. The last slide I'll show you here on this program is we've progressed those, and this isn't even the most recent, but basically from 100 micromolar hits to now we're actually with sub-200 nanomolar and increasing the potency there. We're really excited by this program.
This is one that has routinely garnered attention from strategics over the course of the life of the company. It's a program that I think will be suited to a partnership at the right point in time, similar to our SMARCA2 program. I'll just close by saying we're in a great position these days anyway, relatively speaking. We've got roughly $243, I'll round up to $244 million in terms of cash on the balance sheet pro forma as of the end of the year. We'll obviously report that more formally in our upcoming 10K. We're well positioned to advance a lot of these things by ourselves and obviously with our collaboration with Lilly. I think we've got some meaningful updates coming over the course of the next 6-15 or so months.
With that, I'll pause there and I think we can take questions for another five or ten minutes. Yaron, I'll turn it over to you or anyone in the audience.
Yeah, maybe I'll take Adrian the first question. On the slides, I'm talking about the CBP degrader. It says IND in 2026. I think in the past you've said maybe sort of early. I might have misheard you, but I thought maybe you said even late this year.
Yeah, I think there's a potential for an IND later this year, but I always think about these things in terms of windows as you're running DRFs and GLP talks and trying to optimize stuff. I like to push the team hard to see what we can do. I think of this as sort of Q4 into the first half of next year.
That program is under the broader collaboration with Lilly.
No, CBP is, sorry, I'll just go back to the pipeline chart here. That is a wholly owned program at this point in time. Sorry, pull up the, probably stick on the pipeline chart. Yeah, so that's wholly owned.
That's right. It was the last we were in discovery. Okay. EP300, you said is about six to nine months behind.
Yeah, that's the way I sort of think about that program depending on the path that we pick, whether we do a long-acting VHL or we go with an oral Cereblon-based degrader, just sort of will depend on some of the timing factors. That's roughly how I think about the sequencing. EP300 will come after the CBP program.
CBP is injectable no matter what. You think availability, EP300, it sounds like it's going to be oral.
That's certainly a route that we'd like to go there. I tend to be for the right frequency with subcutaneous, I think that can be a very viable alternative. I think we have oral given sort of the space that EP300 is playing in, whether that's with multiple myeloma or others, I think oral could be a very viable route there.
How are you able to perhaps, maybe if we go back to the upstream sort of lineage between CBP and EP300, how are you able to then selectively degrade one or the other?
Right. The short version without getting terribly technical here is we have found with heterobifunctional degraders with these projects that you're often able to engineer selectivity, which has everything to do with the fact that you're not just taking a binder, you're taking a binder, you're finding a linker, and then you obviously have your E3 ligase. The oversimplification CEO version of the science is you're dealing with the geometry there, which gives you more degrees of freedom. It sort of intuitively makes sense because as cells are trying to remove wanted proteins, they're having to do that in a very, very specific way, right? Like what gets disposed of by the cells, rubbish bin by the proteasome. It's almost you're just taking advantage of nature's way. Now, in the case of CBP and EP300, we do have bromodomain binders here.
What's interesting is the actual degraders are able to take advantage of some of that geometry I mentioned for lack of a better term. We've frankly just through a lot of hard work, medicinal chemistry with the right assays and the right establish of those ternary complexes, we were able to engineer in that selectivity. We've got a thousand-fold for CBP with VHL. We've got a thousand-fold for EP300 with VHL. I think we're north of 250 or whatever that was in the slide for Cereblon with that. That's something that others haven't yet fully achieved. I know one of the large pharma companies had an EP300 program and unfortunately hadn't progressed that due to selectivity issues.
With Lilly, you're targeting essentially SMARCA2 two different ways, right? One is an inhibitor, the other one is a degrader. The degrader is still preclinical. As you think about the difference, I mean, how deep can you go with a degrader? Because I imagine a degrader would be potentially preferential depending how potent your inhibitor is.
Yeah, you know, essentially I've been getting a lot of those questions in our investor meetings today. What I can say is at least when we look at an inhibitor and degrader and a range of in vitro cell lines and look at impact on gene signatures, there's not a tremendous amount of difference. With our inhibitor, we were able to get regressions and stasis in some very difficult lung cancer models. It's not clear that a degrader gets that. Now, what is obvious from our slide deck is the degrader, at least where we were at the end of 2021, and I'd say we've sort of maintained the selectivity profile as we've worked on that with Lilly, you are getting improved selectivity with the degrader. That is going to be a question of does that translate into something meaningful in the clinic?
Now, with 909, I think it's really important to note that the goal there has been to hit that target with an IC90 or better. It is all a question of covering SMARCA2 at that IC90 without meaningfully perturbing SMARCA4 at, let's say, an IC50. If you're able to achieve that, which we believe we are with a 33-fold and with the PK that we have, that may be sufficient. I don't know that. We will see how the clinical trial plays out. I personally remain very optimistic about it. I fully anticipate, given what we've seen from a competitor, that we're going to have monotherapy efficacy. That is why we run the study. I think the degrader, I think about in sort of two or three ways, ultimately it's going to be Lilly's decision, but I see it very much as a backup.
I see it as an opportunity to potentially solve for selectivity if we need that down the road. Third, you can imagine that if this all plays out, you're going to want to have multiple shots for what's going to be a multi-billion dollar potential franchise. I think a degrader allows you for a combination. It may allow you to overcome resistance. You can sequence these things. We will just have to see how that plays out.
When do you think we might see data for 909? Is that going to be more likely from the lung cohort or the bladder endometrial colorectal cohort?
Yeah, the way the dose escalation is set up is all Histologies at this point, but requiring a SMARCA4 mutation. That is in part to ensure that we get rapidly through the dose escalation cohort. The potential for data, as I may have mentioned, would be towards the end of this year into the first half of next year. That is all going to be predicated on how many cohorts we actually have to go through, the decision by the collaboration on how much backfilling to do. The backfilling is set up to focus on non-small cell and specific loss of function mutations. I would anticipate, but obviously we will see when we get there, that the data will include a healthy number of non-small cell lung cancer patients with the appropriate mutations.
What I would say just as a general matter is the epidemiology of this is biased anyway to non-small cell. The sites that have been picked as part of the phase one have a very healthy lung cancer component to it. There is both a natural epidemiological enrichment as well as the fact that we've picked sites that have a meaningful lung cancer service as part of it.
Maybe based on the preclinical models, how fast of an answer of action do you think you can see?
You know, I don't know that I've got a great answer for that. I would think that we would be seeing stuff within the first few months. I think Prelude in their data, you know, not appropriate to go into their molecule, that's for them. I think even with some of the limitations that they have, they were able to achieve some meaningful reductions within a few months. I would expect that you would start seeing that. Certainly the goal is not just response, but also duration.
With Prelude, they did see PRs, right, in the lung, but they were short duration. Can you maybe talk about their data and maybe at least how your molecule is better?
Yeah. I think what I would say from our vantage point, and obviously they're still in dose escalation, I think they've committed to some data at some point this year. Our oversimplified view of it is, as I mentioned, you've got to hit this target really hard. I think given that they've got a once-a-week intravenous delivered drug, they're just not hitting the target hard enough. I think that was demonstrated in their preclinical data. You saw a rebound of the SMARCA2 protein, I think at day two or three, came up another 20-30%. I think you saw that in their cycle one day eight biopsies of peripheral blood monocytes. You saw that they were getting to about 60-70% degradation. We just don't believe that's enough. We've worked on this target for about seven plus years.
We know we've got to hit it really hard, whether that's an enzymatic inhibitor or whether you're degrading it, you need to be at about a DC90 on a relatively constant basis. Maybe that will improve for them as they get further into escalation. I hope it does. The patients need immense. I hope it does for their sake as well. I think we've got the better drugs.
I think we are at time. So Adrian, thanks so much for coming. We really appreciate it.
Yeah, thank you guys.
Good to see you.
Yeah, likewise. Thanks guys.