All right. Thanks very much. It's my pleasure to be moderating this panel with Lexeo Therapeutics. I'm going to ask founder CEO Nolan Townsend to just give a quick overview and update on the company, maybe talk about the recent regulatory update you had with the FA-CM program, and then we can just do Q&A. So thanks so much.
All right. Well, thanks for having us. It's great to be here, and great to be joined by my colleague Eric, our Chief Medical Officer and Head of Research. So Lexeo is a clinical stage gene therapy company focused in genetic cardiac disease and a gene variant associated with Alzheimer's disease. We have three clinical stage programs, the first one treating the cardiac pathology of Friedreich's ataxia. We've read out interim data for this program in July of this year. We've also reached alignment with the FDA on a plan for our pivotal study associated with this program, which we'll talk more about today. We think it's an exciting opportunity set and a disease of high unmet need with no existing therapy treating the cardiac pathology of this disease. Our next most advanced cardiac program is treating arrhythmogenic cardiomyopathy, and here we're focused on the PKP2 mutation.
This is a 60,000-patient rare disease in the U.S. So, for example, it's twice the size of Duchenne's muscular dystrophy, making this one of the largest commercial opportunities in clinical stage gene therapy. This program is also in the clinic. We'll be reading out data associated with this program at the end of Q1 or beginning of Q2 of 2025. Our third program is treating APOE4 homozygous Alzheimer's disease. We have an interim data readout associated with this program a few weeks ago showing an improvement in certain biomarkers that are correlated to cognitive decline in Alzheimer's disease. So, in total, three clinical stage programs, two of which have important data readouts coming in 2025, which we're excited to talk about today.
Great. Sounds good. So maybe can you set the stage and talk about the recent regulatory update you had for the FA-CM program and what the endpoints could look like in a pivotal study?
Yeah. Maybe I'll pass it to Eric to give some comments on that, and I'll add some thoughts as well.
Yeah. I think it's pretty clear to us now that we are centered around two co-primary endpoints, which are LVMI and biopsy expression of frataxin. The biopsy expression of frataxin will be measured by immunohistochemistry, and LVMI will be measured by MRI. I think the regulatory precedent is there, and I think the preliminary data strongly supports that even in our current doses that we are hitting that mark. So we're excited about that.
Yeah. Yeah. Yeah. So maybe let's talk a little bit about the data you've generated so far, right? So I think there's sort of two questions. One, the frataxin data, the degree to which you're consistently meeting the bar that you've agreed upon with the FDA and the agreed-upon methodologies, maybe talk about that. And then two, LVMI, right? And I think in LVMI, the discussion on the street is, should we be focused on the subgroup data? Should we be focused on all the data? So maybe just talk about these endpoints in the context of your data and why you're confident that these endpoints are well-suited to show the drug effect.
Sure. So let me take both of those on. So in terms of the biopsy data, we're using immunohistochemistry. The rationale for that is that that was a couple of things. One, the regulatory precedent is stronger. So just for making a validated study, immunohistochemistry has a strong track record for that with the FDA. So it's easier to make something that we know right away is going to go quickly through the FDA. The other reason is that we're seeing a biological effect by the immunohistochemistry that ties with our clinical effect, right? So when we look at a response, we're seeing a response when we have an immunohistochemical effect. So we think it's a good measurement for us to tie to something that's clinically relevant. It's a reliable study.
It's done by a third party that scans all the slides, looks at an objective way, identifies areas of positive frataxin stain with a positive control, and then just blindly using imaging software gives you a percentage. So we think it's a reliable, objective measurement for us to use.
On that specifically, how did you agree upon what the right magnitude of change was?
The 40%?
Yeah.
So I think the 40% is a number where we, A, we can look at where we were getting in our earlier doses to see a response from patients as a primary thing to see that. And we think that the 40% aligns with the amount of the distribution where we think we're seeing a biologically relevant effect in our preclinical models as well. So you have preclinical evidence, and as well as obviously it's early, but early.
Was all of that kind of presented and discussed with the FDA for them to sort of convey that they're receptive to this number?
Yes. The preclinical data combined with the clinical data was a substrate discussed. It was also natural history data that was part of the picture. So I think this debate about how much frataxin you need to correct the disease is one, you may have a different answer depending on which assay you look at. And frankly, it can go kind of in circles a bit. But I think what's important here is the FDA has stated now has a position on how much frataxin you need to correct the disease. And the answer is it's 40% via immunohistochemistry. And that's the position that they've now taken. So while the debate can continue, I think we now have a line in the sand on the levels at which we can see an approved drug here.
At those levels, for example, in our cohort one dose, we're seeing these very meaningful improvements in left ventricular mass. You asked this question about the bar set: can we comfortably exceed the bar? On IHC today, the bar is 40%. Our patients are on average 44%. This is without the cohort three dose. We've since dose escalated. We've completed enrollment of cohort three. We've seen a dose response between cohorts one and two via IHC. We would expect to see a dose response for cohort three, which would have us more comfortably clearing this 40% bar. I look at the same on the LVMI side. The bar that was set via cardiac MRI was a 10% improvement. We're at 11.4% with the data that we showed in July. That 11.4% is in this elevated LVMI population, which includes two types of patients.
One type of patient is what's called high normal, so this is one standard deviation above the norm for LVMI. The other set of patients involved there are abnormal, which is two standard deviations above the norm in LVMI, so if you look at that set of four patients, three of the four were abnormal. Those three all cleared the 10% bar. The one patient that did not clear the 10% bar was the high normal, so I think that gives some guide into where things are headed from the pivotal study perspective because the two standard deviations population is the focus of the pivotal study, and those three patients all cleared the 10% bar.
So I think that's the picture that we're excited about because we have IHC where we're clearing the bar across the patients that have been reported to date, and we have cleared the bar also on the LVMI side across the patients there. Whether you look at it via responder or you look at it via averages, we're clearing the bar on the LVMI side.
Yeah. Yeah. How should we think about the signal-to-noise ratio in a small sample of LVMI?
Sure. So I think another way to think about it is why is the FDA so comfortable with LVMI? So when you look historically at magnetic resonance imaging studies that evaluate LVMI, the error is somewhere between 2% and 6%. And that's in congenital heart disease, hypertrophic cardiomyopathy. That's interuser variability, interuser variability. You make that even better when you start to use really good cores. When you use cores dedicated like we're doing, we're using Duke and Cornell, places that do very standardized that are part of clinical trials. The FDA is comfortable around the 10% threshold because that's beyond this 2%-6% noise that they estimate. And I think even more importantly, if you look across HCM trials, now we have a lot between two different companies. What you never see is a placebo change in LVMI. Placebo can change quality of life metrics.
Placebo can do all these other things, but placebo does not have an effect on it, so that's why the FDA, there's this regulatory precedent that we've seen now. Now it's two different programs, right? FDA said yes on that metric because they think it's beyond the noise. For us, I'm comfortable about it, not just because of the signal and noise, but also because of the biological basis that we're looking at and the preclinical data, which also showed this reliable improvement, so all those things together make me pretty comfortable about it.
Yeah. Okay. So as we think about the setup into cohort three, what magnitude of dose response would you expect on frataxin increase?
I think maybe early to kind of define a specific number, but what I would say is we'd like to see an increase versus the cohort two patients that were explored. If we do see that, it would have us comfortably clearing this 40% bar. I think that's what we're looking for out of cohort three is to have ourselves comfortably clearing the 40% bar. We've not a priori set a target dose response that we'd expect to see.
What percent of patients are you clearing it in so far?
We're clearing it in two-thirds of the patients so far.
Okay. Got it. But the focus with the FDA endpoint is going to be on what the mean increase, the median increase, the responder?
Yeah. So we're not guiding on the specific statistical methods for the study, but I would say, again, if you look at it in averages, we're clearing the bar. If you look at it via a responder at above 50%, we're clearing the bar. So we're not guiding on the.
But in either way, you're comfortable?
Yeah. Either way, we're comfortable.
Okay. Okay. I guess how much, to what degree is the magnitude of frataxin increase and your ability to show a dose response in a study like this a function of baseline, right? Because I think you're seeing that a wide variability in baseline, and is that actually a correlate of response?
It does not appear to be directly, but what I would say is the reason we've gone to cohort three, and I think why the cohort three dose is important here, is because it can allow us to clear the bar even when patients present with lower starting baselines. So today we're clearing the bar, but we have, as you described, we have patients at single-digit % of frataxin baselines up to the high 20s and 30%. I think we'd like to ensure or try to find a way that we're comfortably clearing the bar across even the patients that present with the low baselines at the beginning of the study. So that's the point of the cohort three dose is to make sure we have that room. And I think the idea that the 40% is relevant is a function of the protein itself.
Fra taxin as an enzymatic or catalytic function, we've not seen that you need a lot of Frat axin to correct the disease. So once you're hitting a certain threshold and you've catalyzed that reaction between iron and sulfur, you may be having an impact on the disease pathology. So it may not be that there's a substantial difference between 40% and 50% and 60%. It's that you need to clear the bar. And I think that's potentially the logic behind the FDA's agreement on that single bar as the point for IHC.
Do you have anything to add, Eric?
Yeah. I would just remind people. I think Nolan, you mentioned the iron-sulfur cluster. This is an enzymatic reaction. And if you look across biology, even for example, in Fabry disease, you need 1% of enzymatic function to get patients feeling better and diseases like that. So that's why, as opposed to setting these bars as a specific amount of normal, looking at the distribution we think is more relevant because we can have very low amounts, but if we're distributing it across the myocardium, that's where we think we can have the biggest. Well, that's why it explains the impact we're seeing, which is namely when you get a little bit of Frataxin in all these cells, you reduce hypertrophy, you reduce mitochondrial hyperplasia, you reduce sarcomeric expansion, and that's why we are seeing the biological response.
Yeah. Okay. Makes sense. And then I guess the other question on the cohort three side is, is this going to be a better enriched population to more consistently show an LVMI decrease? Can you select patients for this FDA benefit?
Yeah. So we have focused on patients with elevated LVMI in terms of.
For Cohort 3.
Yeah. Just the Cohort 3 and the incremental patients on the Cornell side that are Cohort 2 as well. So we've tried to focus on patients with elevated LVMI in order to demonstrate the effect that.
So can you clarify that? There's more patients being enrolled at the Cornell site?
That's correct. Yeah. So as part of our update, we mentioned there that 16 patients have been enrolled in the study to date. As you're aware, in July, there were 11 patients at that point in time. So there have since been five additional patients enrolled. And then Cornell has two additional patients that will be enrolled to conclude that part of the study. So a total of 18 patients, of which eight are the SUNRISE- FA patients that will all have cardiac biopsies. And the balance, the 10, will be at Cornell. There's five patients in cohort one there, five patients in cohort two, and then our study with the profile that we've discussed.
Essentially, when we think about the data we have, we have four elevated LVMI patients now. At some point, we're going to have more than double that sample size.
At some point, we will have a substantially higher number of elevated LVMI patients to evaluate. That's correct, and I think that will be the primary focus of the data next year will be that population because if the question is, can you comfortably clear the threshold set for the pivotal study, there's a read into that by virtue of the data coming out of the phase one, and that, I think, is the data that will be most important to focus on in 2025.
Yep. Yep. Okay. So can you talk about the data flow in 2025 and what we should expect when?
We will give an update on our milestones at the beginning of the year, likely in our presentation at the J.P. Morgan Healthcare Conference. What I would say, there will be an FA readout with cohort three biopsies and incremental LVMI data that will be in the picture next year.
Okay. How long do you need to follow these patients before it makes sense to expect an LVMI change that's clinically significant?
Yeah. I mean, I think we can start to see things at six months and we start to see, but then we think it continues through 12 months, and then we even have 18-month data where it continues to improve.
Okay. So six months is the minimum timeframe. We're probably talking about second half of next year for more data. Is that right? Or maybe some data earlier?
I think we'll give guidance on the specific timeframes soon. But I think we'll have the biopsies in the first half of next year. So I think we'll give guidance on the timing in the future.
Yeah. Okay. All right. Makes sense. Anything else to add on FA?
No. I think we made a lot of progress in this program. I think going to completing enrollment of the study, all three dose cohorts, the drug's been well tolerated to date. We've seen these material reductions in LVMI. We've aligned with the FDA on the endpoints relevant for a pivotal study. So I think we see a pretty clear path from here through a pivotal study at effect sizes that we've already achieved in our phase one. So I realize there's still some questions about comfortably clearing the bar, but I think also we still have the data from our highest dose yet to read out here. So I think it's a very interesting story for a program and a disease where we've already seen companies like Reata that have gotten pretty substantial strategic interest. We've seen a nice uptake of Skyclarys.
So I think there's a pretty good picture here just from a disease context, what we've achieved in our phase one study, the regulatory alignment that we've achieved, and the broader interest in cardiac genetic medicines and cardiac gene therapy, we think are all great tailwinds for the program.
Yep. Yep. Okay. Makes sense. Maybe one last question on FA before we move on, and that's just on the safety side. I know historically over the past couple of years, you guys had had a conversation internally about whether it did make sense to go up to cohort three, of course, whether you needed to or not, or on the risk-benefit piece because there is greater exposure of this in the liver like most or all AAVs. So what's your comfort level there in the margin?
Yeah. I think you can look at. There's actually pretty good preclinical data on this. And what I think you're specifically talking about is the risk of frataxin toxicity because you're expressing more frataxins. Taking the capsid and all that stuff.
Idiosyncratic. The tox risk seems low given the doses.
Yeah. You and I are totally on the same page or align with that. But I would say for the liver, where you will have more transduction of virus, thankfully, the liver is a regenerative organ. And the way that translates is that you need like 40 times wild type in liver in preclinical models before you start to see frataxin toxicity causing hepatitis. So the kind of threshold is quite high. And we don't think we're anywhere close to that. So again, I think our therapeutic index, even accounting for liver transduction, is quite large, the therapeutic index. So I think we're in a good spot from that. And we haven't seen any, obviously, we're moving into another cohort, but at least at the current cohorts, we haven't seen even blips of transaminitis yet.
Yeah. I'd just add our safety update that we provided last week is inclusive of cohort three. Cohort three began enrollment in the summer. So we're talking about a safety update that includes patients in the high-dose cohort at many, many months of follow-up. And again, the safety update has not identified a drug-related safety issue in that cohort. So I think that's as good of evidence as any of the drug. And typically in gene therapy, you would see those kind of safety responses very soon after dosing.
Yep. Yep. Okay. All right. Great. Look forward to more data there. Maybe we can talk just a bit about PKP2 since that's going to be the next data update.
Sure. Sure.
I guess if we think about this population, it's more heterogeneous. Talk about who are these patients phenotypically that are ending up in an investigational gene therapy trial?
Sure. So level setting on this a little bit. The patients with the arrhythmogenic cardiomyopathy, I think they're recognized as having one of the worst quality of life of patients in genetic myopathies in general because they have this constant risk and concern about getting shocks from their defibrillators. So they're always walking around worried if they're all of a sudden going to get electrocuted. And that leads to extremely high levels of anxiety and depression among these patients. They're often younger when they're diagnosed in their early adulthood, I should say, or teenage years. And what do we do? We restrict them from exercising because that's a key part of the current treatment landscape. And we put in defibrillators and tell them, "You may get shocked at any time, day or night." So they live in this constant high anxiety.
How often are these patients getting shocked?
It really depends, and that's what actually the unpredictability is what makes it so anxiety-provoking. Some patients might not have a shock for a year or two years, and some patients might get 20 in a day, and so it depends. There's obviously medical therapy and surgical therapies that we try. But there's this constant fear of you don't know when it's going to happen that I think plays into the mindset of these patients. Then in terms of the patients in the trial, what we're seeing are patients that have elevated premature ventricular contractions, I should say. These are patients that have extra heartbeats. They can feel those extra heartbeats generally, not always, but they generally can feel them as sometimes we'll just feel fluttering in our chest. But they're happening all the time in these patients.
Approximately above 500, approximately around 1,000 or more. Most of these patients will have even more than that per day. And those are the kind of patients that want to go in the trial. And that's part of our enrollment criteria.
I would add a lot of these patients go on to require transplants. So they're living with the anxiety and all of the issues Eric just described, but kind of all roads for them are leading towards transplants. So I think that's a picture for a patient to say they would consider a gene therapy given all the context of what we just described.
Yeah. I mean, we are proud to talk about our enrollment, and we're seeing robust excitement about this from the patient communities. There's a lot of excitement among the sudden cardiac death community among these patients with plakophilin-2 cardiomyopathy about being part of this because they want therapies beyond just the medicines they have, which tend to have a lot of side effects and having to have ablations and having to get shocks from their defibrillator.
Yeah. Okay. So as it relates to a clinical trial, I mean, you can obviously look at expression of the missing protein, but beyond that, I mean, is it as simple as looking at defibrillator shocks? It sounds like actually a pretty easy endpoint.
From a long-term standpoint, I think, and what we talked about before, the defibrillator shocks over a year period would be hard to enroll a small trial in, but something like premature ventricular contractions as a potential endpoint. On the other hand, because you can measure it on a week-to-week basis, so we can have patients wear a Holter monitor and look at the averages, that's a compelling endpoint. We don't have the data yet to determine with confidence whether that would be the only primary endpoint. There's multiple other things that we can look at. For example, these patients have abnormal right ventricular function and structure. So we can do imaging analysis to evaluate RV volumes, RV size, scar burden in the RV, all kinds of things. We have new sophisticated imaging techniques that we can do that even in the presence of defibrillators.
Then we can look at a host of quality of life metrics. We're heavily phenotyping these patients. So we're going to learn a lot across, and frankly, across multiple trials would be great. We're going to learn a lot about this disease. So I think the endpoints will become pretty clear.
I'd add this disease has a very robust set of natural history studies associated with it. There's a few in Canada. There's one that's out of Johns Hopkins. I think we have a disease that's pretty well understood from a natural history point of view, which can obviously support the future development of the program beyond the phase one.
Yep. Okay. So as it relates to the first dose cohort, what would constitute a win?
Yeah. So I think this is a first-in-human study. We've obviously seen certain effects of the drug in the preclinical setting. But for the clinical setting, our focus is on both safety. So here we have 1E13 per kg dose in adult patients. So we think safety is an important read. Secondarily, we're focused on seeing an increase in plakophilin-2 protein versus the pretreatment baseline. And I think that that can give a good read into the clinical effect that could be observed at later time points on some of the endpoints that we are focused on. So the early read, the early data will be focused on biodistribution and safety at the end of Q1.
Yep. Okay. Okay. And do we have a good understanding of the level of gene correction that may translate into benefit here?
Right now, we just want to show that we're increasing over baseline. It's a little bit difficult to assess exactly how much, getting into that game of how much you need, how much you have. It's a little bit difficult at this time to go beyond that.
Is that just because the preclinical models aren't that good? The literature is not that good?
No. Also, the patients have baseline plakophilin-2, so it's not a knockout, so you're going to have patients presenting with different baseline levels, and so I think it's going to be difficult just to set a single bar for that, so I think that's not probably the most productive direction to go with the readout. What we want to see is that the therapy is having an effect. It's expressing protein, and we've seen that in the preclinical studies. It has resulted in an effect on the disease, so I think the first step is just to see that we're expressing a sufficient amount of protein and then later look at some of the clinical endpoints that are relevant for the disease.
So what are those clinical endpoints specifically?
Yeah, so premature ventricular contraction is something we're going to measure by Holter monitoring.
You're going to have these patients wear Holter monitors.
Yeah. They have ICDs, and you can also get a sense of that from the ICDs, but because the brands gets into the weeds, it's much more reliable to have them wear Holters. They're going to do magnetic resonance. We're going to be doing a whole host of patient-reported outcomes as well.
Right. Yep. Yep. Okay.
You get a lot of data from each one of those things. Also, to your point, and I forgot, we're going to be measuring sudden death. Do these patients have defibrillator shocks? Whether that will play out in a small number of patients, a small time period, is unlikely, but we'll be capturing that as well.
I mean, maybe it's way easier said than done, but why couldn't you just enrich for a population that has a pretty rough one- to two-month history of that?
Yeah. That would be hard. And to do that and enroll quickly would be.
It would be hard.
Yeah. I think to me, I think that the actual sudden death risk is a full approval because you're actually literally preventing death. It's called sudden death. So it's going to be a great full approval endpoint, but I think it will take patients and time to statistically power for it. I'd like to see it happen, but to actually truly statistically power in a non-placebo.
There are other endpoints that are highly clinically meaningful, like premature ventricular contractions, T-wave inversions, and others to look at ahead of the ICD shocks that could yield things that actually could be accelerated approval endpoints.
So there's an FA. Well, and thanks for reminding me, Nolan, the MBA correcting the MD. So I always feel a little hesitant. But yeah, so there's EKG changes as well that we see on these patients that's pathognomonic for disease. There's a lot of interesting debate about whether a gene therapy can actually how long it will take and will it change your EKG, right? That's going to be a game-changing, amazing thing to see in practice. And I think it can. As you see remodeling in the heart, you can start to see T-waves go back to normal. So excited to see that as well.
Yeah. So I think the point is that there are a number of endpoints here beyond just ICD shocks. And so we're talking through a number of them. I don't think we've set the mark on which one is most appropriate long-term, but I think there are a range of them.
I will say that there's a currently ongoing trial of a medical therapy that's using PVCs as a primary endpoint, so the FDA does have some Flecainide trial that's ongoing where there's some regulatory agreement because we know, and maybe if not, PVCs predict sudden death. The more PVCs you have, the more likely you are to have a life-threatening arrhythmia, and the FDA understands that.
Right. Okay. Okay. Makes sense. We only have a minute and a half left. I know we've talked about the APOE4 or APOE2 program a lot, but you recently had some data at CTAD. What's the next step there, Nolan?
Yeah. So we saw an effect in reduction of tau that's similar to the commercially approved therapies within, in particular, in the moderate population. And we saw this effect with no reports of ARIA. So that was an interesting signal that we think is worth further investigation. However, this is a program that we would not take forward as a company on our own. It would be a program that would be taken forward alongside a partner there. So I think our core focus as a company will be in the genetic cardiac space, FA cardiomyopathy, plakophilin-2 cardiomyopathy. And I think we'd see the APOE program progressing alongside a partner in the future.
Okay. Great. Well, thank you guys very much. Great conversation. Appreciate it.
Thank you as well. Appreciate it, Paul.