Right, so welcome to this next fireside chat. My name is Michael Schmidt, senior biotech analyst with Guggenheim, and it's my great pleasure to welcome Adrian Gottschalk, President and CEO of Foghorn Therapeutics.
Thanks.
Adrian, welcome.
Thanks, Michael.
So Adrian, Foghorn was really founded based on the premise of targeting the chromatin regulatory system and the BAF complex, and it's very scientific. What makes this such an interesting group of targets within oncology?
So, you know, if you go back, 10 or 15 or so years, maybe a little longer, and you look at the era of, genomic sequencing, and cancer, as folks were starting to sequence various tumors, and look deeper in the biology, it turned out that, you know, upwards of, you know, 20+% in this one remodeling complex, was mutated across all these different cancers. And then when you broadened that out a little bit and looked at, things like transcription factors and other elements related to the regulation of, you know, the three-dimensional architecture of your DNA, it turned out to be that 50+% of cancers had some sort of mutation or alteration as it related to this biology.
For the longest period of time, actually, people thought this was just sort of housekeeping, maintenance, machinery, that wasn't very relevant in disease, let alone in cancer. And actually, that's sort of the farthest thing from the truth. So when you take a step back now, and hindsight, as we all know, is 20/20, turns out a tremendous number of these molecular machines, proteins, transcription factors, complexes, are implicated across a wide range of diseases, certainly cancer being at the forefront of that. So, the challenge has been, how the heck do you drug these things? Because in a lot of cases, these alterations are actually loss of function, meaning the protein or the gene has actually disappeared from the cancer in this case.
Unlike some other things in life, you can't drug something that's not there. You can't do something that's not there. So that's been the historical challenge of drugging something that's not there, and in many cases, drugging something that looks very, very similar in healthy cells to other parts of the machinery. So that's where we started as a company. We've come a long way over the better part of eight or nine years since then, but that's what makes these sort of really interesting targets, you know, breadth of application across cancer. And frankly, the challenge of drugging them, because that means there's not a lot of competition to go after a lot of these targets.
Yeah. No, definitely a, you know, large opportunity. It seems like high barriers to entry. How have you approached that from a target selection process? Maybe talk about the pipeline and the products that you're developing.
Sure. So, you know, if you look at our current pipeline as it stands today, you know, chief amongst those are programs that we have actually in collaboration with Eli Lilly. This was on the basis of a deal that we did with them back in December of 2021, where one of the first things that we've been working on with them is this target SMARCA2. It is a synthetic lethal pairing with another protein of a similar name called SMARCA4. They're very, very similar proteins. Turns out in a wide range of cancers, and certainly of interest to us and Lilly in non-small cell lung cancer, the SMARCA4 protein is mutated or lost, and so the strategy there is really to drug the SMARCA2 protein.
I was told many years ago by some very senior executives at other pharma companies that it was impossible to drug this selectively. This company had tried and given up. I think we've proven them and many others wrong on that, and hence sort of the deal that we did with Lilly. When you look more broadly into the rest of the pipeline, at least stuff that is not partnered with Lilly, we have three other interesting targets that we're currently working on. A protein called CBP, which looks to be relevant in ER-positive breast cancer, as well as in some other synthetic lethal situations. We have another target called EP300, which in some ways is the sister protein to CBP.
That looks like it's relevant across a range of hematological malignancies, chief amongst those being multiple myeloma, diffuse large B-cell. You know, that's a very exciting program for us. And then ARID1B, and sorry for all the nomenclature here. People will probably forget all of this right after this meeting, but these are all targets that the field at large has wrestled to drug in a selective direct manner. We're doing that in the case of those three programs through targeted protein degradation. So again, the challenge, similar to where I started in the introduction, is these are interesting targets that people have not been able to drug in any selective manner because they are very similar.
In the case of CBP, EP300, those are sort of the, the sister proteins, if you will, for ARID1B, it's ARID1A. Again, you know, the challenge is how do you get these things-
Yeah.
How do you attack them in a very selective manner?
Yeah. So I think in a way, all of your targets are synthetic lethal paralogs. Essentially, you're targeting one of the targets, and then the second one is mutated in the same background, so creating a cell death. Is that the right way to think about it?
I think in general, that's where we started. What's become really interesting, actually, I'll take both CBP and EP300. So, true for our SMARCA2 program that's partnered with Lilly, where we're looking for cancers that have lost that SMARCA4 protein. I always think of it as the bicyclist where, you know, if you think of cancer as a bicycle and it's lost one wheel, you're now trying to get after that last wheel that's left. In the case of CBP and ER-positive breast cancer, this is just a frank dependency. This is irrespective of the sister protein, in this case, EP300 being mutated. So we've seen evidence independent of the mutation.
Same thing, by the way, for all the heme malignancies, when we think about EP300, this is irrespective of the mutational status of the sister protein. So we started off looking at a lot of synthetic lethals, and what one does, obviously, when one creates chemical matter, is you start looking across a wide range of different tumor types, irrespective of the mutation, just in case something else pops up. Lo and behold, we saw that. So that's where it started, but that's certainly not where we're ending up, if that makes sense.
Yeah. Okay, so perhaps a few questions on your SMARCA2 program, which is your most advanced program, at this point. As you said, I think it's quite challenging to make a selective SMARCA2 inhibitor. How have you guys accomplished that, and, how important is it to have, selectivity over SMARCA4?
Sure. So, I was going to joke and say, if I told you, I'd have to kill you. But, look, the historical challenge, by the way, just to set this up, if you look at those two proteins and you look at the primary sequence, the amino acid sequence, they're about 90+% the same. If you look at the ATPase binding pocket, right, the enzymatic pocket of the enzyme, it's almost identical. So, the way that we went about finding chemical matter on this is, and this was in the much earlier days of the company, is we had the unique capability of being able to compare chemical matter against the full machinery. What do I mean by that?
So the SMARCA2 protein and the SMARCA4 protein, think of it as like the engine of a car, right? V6 or V8 turbo, right? Can't have two engines in a car, you have one, and all around it, all the other pieces of the car is this big, you know, assembly of proteins, this rest of this BAF remodeling complex that is rearchitecting your genome. And we were able to take chemical matter that we found and compare it to a machine that had SMARCA2 as the engine, and then compare what happened when we tried it against a machine that had SMARCA4. And so that's something that nobody else in the field was able to do, because no one could make these machines in that pure a way.
These are monster, you know, 1.5 million molecular weight. So we were able to sort of figure out what chemical matter actually was selective for SMARCA2 in its native sort of state versus SMARCA4. So that's roughly how we sorta came across the chemical matters. We were able to figure that out, and it binds in an allosteric pocket, so we're not having to deal with all the off-target stuff that comes with ATPases. Second part of your question.
6 A.M. flight from Boston, so-
Yeah.
Brain's not-
No, I think that covers it, and I guess my other question was the size of the opportunity for SMARCA2 inhibitors or degraders, you know?
Right.
I know your phase I is mostly focused on lung cancer, but how should we think about the overall size of the opportunity?
Yeah. So, so if you look across, tumors, you'll find in about 5% of all cancers, there's a SMARCA4 mutation. But again, in the synthetic lethal setup, we are most interested in where SMARCA4 has mutations that cause the protein to be lost. The one that we and Lilly have been most focused on, as you just mentioned, is non-small cell lung cancer. We know that lung cancer has, in about 10% of all non-small cell lung cancer, you'll find a SMARCA4 mutation. Of that 10%, let's just call it out of 10 patients, you know, anywhere from six to seven of them actually have the relevant mutation, we believe, that will make them responsive to the 909 inhibitor that we have developed. So, you know, that's a pretty sizable population.
I think the math roughly is that there's about 200, 220 or 230 ,000, unfortunately, new patients diagnosed with non-small cell lung cancer every year. So you're talking sort of in the range, if you took 10% of that, it's about 20, 22 ,000. So, you know, 15 or so thousand patients in the U.S. are probably, relevant on an annual basis here. And obviously, you can double or triple that depending on how you think about Europe, Japan, let alone other emerging markets, China, et cetera.
Yeah. And so the phase I study's been ongoing for some time now. Just remind us where you are in the study and how you're tracking towards perhaps an RP2D identification?
Right. So, again, as a reminder, this is obviously in collaboration with Lilly. We share in the cost of this on a 50/50 basis. So Lilly's been operationalizing the study. It opened and started dosing patients roughly in October of 2024. We have sites open, about 16 sites open in the United States, five sites open in Japan. Over the course of December and January, Lilly has now opened sites in France, Germany, Spain, and South Korea. So that was anticipation of potentially going to dose expansion later this year, but also and more immediately and importantly, finishing up the escalation and the backfilling. So, we've obviously been at it a little bit while. We have not yet hit our maximum tolerated dose.
So, to date, we've been pleased with the safety and tolerability that we've been seeing in the study. In the fourth quarter, well, let me just back up a second. The escalation portion accepts any histology, so you can have any solid tumor, but the patient must have some type of SMARCA4 mutation. In the fourth quarter, the collaboration started backfilling cohorts. And that was—could be triggered by one or two mechanisms, which is we achieving the appropriate PK profile, meaning that we believe we got the right exposure to get to or within the range of IC90 coverage of the target... and/or starting to see clinical activity.
We've not commented on whether it's one or both or whatnot, but suffice it to say, we've, we as a collaboration, have been backfilling patients now at a couple of different cohorts or so. That enrollment is going well. And because we haven't hit the maximum tolerated dose, we are continuing to dose escalate. When you play that all out over, you know, the Gantt charts and look at the timing, and this is my guidance, I can't speak for Lilly on this, I believe that the collaboration will have sufficient information come sort of middle part of the year, could be a few months earlier, could be a few months later, to make a decision on whether we go into the dose expansion section.
I should emphasize that within the backfilling, we are prioritizing those non-small cell lung cancer patients, and we're specifically prioritizing those patients that have that loss-of-function SMARCA4 mutation. So that's where we're at. I can't say enough good things about our colleagues at Lilly and how they've operationalized the study. Their global infrastructure is fantastic. I think that's the benefit of having a strategic partner in the case of these very large tumor types. They're able to execute at a global level that a small biotech like ourselves wouldn't be able to do.
Yeah, and is-
Certainly not at the same speed.
Right. Is this a target where one would expect single agent activity that's you know, perhaps you know, clinically meaningful, or is it more of a combination play in terms of the activity that-
Right.
Might be expected?
So, I personally expect to see single agent activity, and in fact, I would go so far as to say, I need to see, if I were the sole decision-maker in this, I need to see single agent monotherapy activity. I want to know that patients can achieve an objective response, and not just a response, but with actual some duration. Maybe just to contextualize this, right? So, if you look at frontline metastatic non-small cell lung, and this is pretty stark, and it's, you know, these are... I always remind myself, you see numbers on a page, and you see these graphs and lines, but there's actually people behind this that are suffering.
So if you look at first-line non-small cell lung cancer, where patients have been given immunochemotherapy, if you don't have a SMARCA4 mutation, you get about a 40% or so response rate, you get about six or so months of median progression-free survival, and you get about a 15-month overall survival. If you're an unfortunate soul who happens to have a SMARCA4 mutation, and this is, in my mind, horrific, you basically cut those numbers, rough justice, in half. So 20% response rate, about three months or so, median PFS, about eight months OS. That's in the front line. We are seeing patients in the escalation/the backfill part of this, who are fourth- and fifth-line patients. So you can measure their survival, unfortunately, in weeks of time.
So, I certainly would like to see monotherapy efficacy in the setting we're in. I think if you start seeing, you know, several PRs and you see a duration of three to four months, keep in mind, fourth, fifth line, given what I just told you about the front line, that's gonna be pretty exciting to me. To your point on combination, the reality, I... and I know there's a few exceptions out there, but the vast majority of cancer is combination. But I'm a firm believer that you have to see monotherapy activity so that you know what the contribution, at least minimally, of that agent will be to what you do in combination. And again, I just told you what it looks like front line.
The thought will be to go as quickly as we can as a collaboration into the front line setting, doing that in a mix of combinations, likely pembrolizumab, plus or minus chemo. KRAS is a whole another area we can talk about as well, given the unmet need there. So, I think the unmet need is unfortunately very high in this population, but I do wanna see monotherapy activity-
Yeah
... in this study.
Any plans to kick off, combination studies?
So I think that will certainly come next in the expansion phase. The current, if you wanna take a step back and look at the broader clinical development plan, right? We have escalation, which you know is 1A, expansion is 1B. You can think about as a phase II if you want. I know my colleagues at Lilly have been on the record saying they run an escalation, they do expansion, and then they go straight into a registrational study. So this expansion phase, which, knock on wood, we'll move into later this year, will include monotherapy cohorts, but will absolutely then include combination cohorts.
Gotcha.
And again, pembrolizumab, given that's sort of standard of care for these patients, in that first-line setting, will absolutely be on the table.
Makes sense. Any learnings from competitor or peer programs targeting SMARCA2 that could further validate the approach, the concept?
Well, so we know, you know, our colleagues at Prelude had a degrader, two degrader efforts, right? The 909 happens to be an enzymatic inhibitor, so it's a little bit of apples and oranges, but nonetheless. So just to recap, our colleagues at Prelude had a VHL-based degrader that was a once-a-week intravenous formulation. We think they did a nice job showing sort of initial proof of concept against the target, right? They ended up getting two partial responses in lung, I believe it was two PRs in esophageal, and I think one in gastric. They stopped their IV program ostensibly because they didn't believe they were covering, at least from what I read in the press releases, they weren't covering the target sufficiently.
Their oral program, which was a cereblon-based degrader, they stopped, you know, late Q4. Don't know what the data were. What we've learned from that, I mean, I think there's sort of two to three important takeaways. One, we've made their clinical candidates in our hands, they're not very great degraders. They don't get more than 80+ degradation. I think that reaffirms sort of the view that we've had, and this is based on a lot of empirical data, is you have to hit this target hard. For an inhibitor, that means being above an IC90. For a degrader, it means being at 90+%, maybe even 95+% on a sustained basis. So that's point one.
Point two, which I think came out more from their IV program, is, you know, there, there is differences, and we suspected this, and it was nice to see, I guess, some of that proven out between these loss of function, these class one versus class two type mutations. So when you're running a study, you are going to want to make sure you've enriched for that class one population, so you get more of a homogeneous signal in what we believe will be the more relevant patient population. So those are two of the things that I would probably comment on.
Great. So hopefully, you know, looking forward to, hopefully, to updates from this SMARCA2 program via Lilly, presumably later this year. Is Lilly driving disclosures from the program?
We'll obviously work in collaboration with them, but from our vantage point, the decision to go into a dose expansion is a very material event, probably not surprisingly, for us. I can't comment on Lilly's materiality, but I think we can suspect that it's not a phase I oncology program. But jokes aside there, you know, if and when that decision is made, we will want to communicate that. We will want to, you know, communicate at least very high level, the safety tolerability, which is very important, obviously, if we want to do combination as well as high-level efficacy. But we're gonna do this in collaboration with Lilly. We'll figure out how we best do that and communicate that to the street.
Yeah. Okay, so then I did wanna touch on your CBP and EP300 programs, which are your most advanced wholly owned programs. And I know you've been able to generate selective degraders for each of those targets. And, yeah, maybe speak a bit about, you know, learnings from your preclinical work and how you're tracking towards initiating phase one studies.
Sure. So I'll chat briefly about CBP. So, as I mentioned, we've generated some really interesting monotherapy data in ER-positive breast cancer. We had done some in vitro work where we showed some nice combinations with CDK4/6 and with the fulvestrant as well. We recently generated a CDX and two PDX models in CDK4/6-resistant lines and showed actually pretty striking monotherapy data. I would say that are at least as good, if not better, than what you're seeing with the CDK4/6 drugs. So we're pretty excited by that. We're actually doing a bunch more in vivo work, monotherapy and combination. We literally just finished the non-GLP tox studies last week, the in life portion, that is. This was with our once a week subQ long-acting formulation.
As you know, most times you run a two-week non-GLP tox, and you do the four-week in GLP. We actually ran four weeks on this, because we were only gonna give two doses if we did two weeks, so we wanted to see a bit more. Knock on wood, low, medium, high dose, all the animals survived. Obviously, we got to get the more detailed results from that, but we're excited by that. So as you think about next steps here, it's really additional in vivo work in the ER-positive breast setting, scale up of drug, and then preparing for IND. And obviously, you know, the clinic will follow, you know, pending how we think about some capital allocation and how things go over the course of the year.
So excited by that program. Briefly for our EP300 program, that's tracking to be, pardon me, in non-GLP tox studies by the middle part of this year. Again, there, we're gonna go with a long-acting formulation. We also have a molecular glue effort that's been going on in the background. We actually have a cereblon-based degrader that we sort of paused on that, not 'cause it wasn't looking good, but as we prioritize multiple myeloma, IMiDs are cereblon-based drugs. You can't put a cereblon-based degrader in with an IMiD, that doesn't really work well. You don't get the efficacy results. Probably not surprising. So excited by that program. We've got some wickedly selective, that's a Boston term, I suppose, a wickedly selective degrader and highly potent that we're pretty excited about.
I think you're gonna see some movement on that over the course of the year as well.
The CBP program, the phase I study, do you anticipate that being focused on breast cancer or additional other solid cancers?
I think for purpose, if the ER-positive breast hypothesis holds together, which so far it looks like it is, I think we would prioritize in escalation just given the sheer size of that patient population, the unmet need, once people are sort of in the metastatic setting, we'd prioritize that through escalation. I think you could see an expansion or as we backfill, we'd start looking at not only the ER-positive breast stuff, but some of those synthetic lethal tumor types that we previously profiled, you know, some of the genitourinary type cancers, and whatnot. So I think those will come in the fullness of time, but I think from a speed and homogeneity of patient population, we'd probably go straight into the ER-positive breast cancer setting.
Okay. Just wanted to ask one more question. There is some industry peer data as well. I think some of the programs may be less selective, targeting CBP and EP300, but any learnings, and that could read through the approval?
Yeah, no. So, you know, if you look at our colleagues at CellCentric, and I think they've done a great job providing some validation in multiple myeloma, they saw, I thought, very good efficacy in fourth, fifth-line patients. I think the challenge, what they've highlighted clinically, and this is seen preclinically, not only with their compound but other compounds, is if you hit both CBP and EP300, you have a myelosuppressive effect, so you're gonna end up with a range of cytopenias. At least preclinically, we've dialed that out completely for our CBP program, but also for EP300.
We're sitting on a program that, from my vantage point, when it comes to EP300, has already been clinically de-risked in some sense, and we're gonna have, in my opinion, a better safety tolerability profile that will be able to dose higher. We're not gonna have to do a few days on and off like our colleagues are. I think when it comes to combination, which is the name of the game in cancer, we're gonna be much better prepared across not only multiple myeloma, but the range of heme malignancies.
Great. Well, with that, unfortunately, we have to wrap up. I know you also have the ARID1B program, very unique story, very big market opportunity, but, unfortunately, we have to wrap up.
Lots, lots to talk about, but no more time. So-
Such is life. I appreciate the time, and hopefully looking forward to some market updates later this year.
Indeed
... and then, your IND-enabling work on the CBP program.
Yep.
Thank you so much, Adrian.
Michael, thanks.
Appreciate it.
Thanks for having me. Thank you, everyone.