Maybe give us a quick snapshot of Lexeo, some of the milestones we should be looking forward to over the next 12-18 months.
Yeah. So Lexeo is a clinical-stage gene therapy company focused on genetic cardiac diseases and a gene variant associated with Alzheimer's disease. We have three clinical-stage programs, two on the cardiovascular side, one on the central nervous system side. The most advanced cardiac program is treating the cardiac pathology of Friedreich's ataxia. We've read out interim data for our Phase 1 study associated with FA, which I think we'll spend, you know, a lot of the time discussing today. Our next most advanced program is treating arrhythmogenic cardiomyopathy, and here we're focused on the PKP2 mutation. Here, we're looking to read out data in the coming months associated with the first dose cohort. Our third program is treating APOE4 homozygous Alzheimer's disease.
We're treating these patients with the APOE2 gene, which we believe confers a lot of the protective benefits of APOE2 to these patients. We'll be reading out data associated with this program also this year across all four dose cohorts, likely at the CTAD conference. These are the ongoing clinical studies that we have in the pipeline today, and look forward to discussing them.
Great.
Super
Why don't we start with the Friedreich's ataxia program? And starting with the basic biology here, frataxin is a mitochondrial, I guess, functioning protein, something to do with iron-sulfur clusters, I'm not quite sure. But how do we think about the correlation between that defect and the phenotype of a hypertrophic cardiomyopathy, in addition to some other features?
Sure, sure. Yeah, and you got to the main point. So frataxin allows the binding of iron and sulfur, and iron-sulfur clusters, which occur because of frataxin, are a key catalyst for a whole host of reactions within the mitochondria. So without frataxin, you have mitochondrial dysfunction. The response to the loss of frataxin is mitochondrial replication. So you see this enormous... Because mitochondria, you know, can replicate and start dividing, and so you can see the expansion of mitochondrial content, which is a pathognomonic of this disease, so that's some of the hypertrophy. The other hypertrophy is probably sarcomeric, and that's because if the heart is energetically deprived, it's not getting enough, the response is to make more sarcomeres to maintain cardiac output.
You have this combination of mitochondrial expansion to explain some of the hypertrophy, and then true sarcomeric expansion as well.
Then, in terms of kind of target expression levels, you do have a mouse model that kind of gives a little bit of guidance. Maybe describe the mouse model and what we can infer in terms of the levels of frataxin that you'd need to
Yeah
... to establish.
Well, and I think, you know, this would resonate with you. You can start even at the biology. This is an enzyme, and we know from enzymatic deficiencies, you don't need a lot to resolve the pathobiology. I would say, you know, the best analogy is like hemophilia, where very small, small amounts of protein can reverse the phenotype. In this case, this enzyme, which you mentioned its function, frataxin, we don't think you need a lot. And maybe the best evidence is a mouse model with less than 5% of normal. This model was made to mimic a human mutation, where you have 800 repeats. GAA repeats are inversely proportional to the amount of frataxin. So in this model, which has 5% of normal frataxin, you see normal cardiac output.
You see the ataxia, the other things, but the heart seems to need very little frataxin. In the human condition, we see obviously heterozygous patients, we see patients with shorter numbers of GAA repeats, also with normal cardiac function. So there, there's a lot of supporting evidence that you don't need a lot of frataxin. Given that frataxin can be toxic at very high levels, in this case, we're shooting for the lower end.
Yeah.
Okay.
We've not yet talked about the clinical data, but what's intriguing is that actually, in our cohort one patients that are, you know, likely below 5% of normal, we are seeing improvement in certain biomarkers that are relevant for the disease, so actually, this thesis is playing out, you know, clinically, potentially even in a more advantageous way than what we originally envisioned, because at below 5% of normal, we're beginning to see this improvement in certain key biomarkers.
Why don't we dig into the clinical observations then? Maybe even just starting with what you are finding in terms of the baseline frataxin, and the extent to which that may correlate with disease burden.
Yeah, so this was actually surprising and, you know, for us, exciting, but there was some speculation that patients had, like, 20% of normal frataxin, something like that. That was all based on the skin, where the denominator, you know, buccal cells, these type of things, where you don't make a lot of frataxin because you just don't have a lot of mitochondrial content in those tissues. In the heart, you need a lot of frataxin because, you know, a third of the heart's weight is mitochondria, so you need a lot of frataxin, right? It's a very energetic organ.
In this case, when we did heart biopsies, and this is the first time I think this has been done in FA patients in a real measurable, quantifiable way, we see that almost uniformly thus far, it's a small N, they have less than 2%. They're almost like a knockout. They're not quite a knockout, but they have very low levels of frataxin in at least when we measure it by mass spectrometry, which should be a sensitive marker. And now it's consistent across multiple patients as well.
Maybe we can then review the design of the trials, the dosing and escalation in patient numbers, and what you're looking for.
Yeah, so this is a trial where we have. It's obviously a single-arm, Phase 1 study, looking, you know, obviously, the primary endpoint, like all these studies, is safety. We have, you know, a E11 cohort, which is completed, and now we're in the E twelve range of one point two E twelve. So it's three cohorts in total. The first two cohorts were E eleven, now E twelve. The endpoints that we're measuring that we think are most relevant biomarkers for disease are structural, so we're looking at LV mass and LV wall thickness, measured by MRI. We're also looking at the biomarker, which I think is most relevant in Friedreich's ataxia, which is troponin.
Okay. Maybe we can talk a little bit about the frataxin measurements you're using. You're evaluating both mass spec and IHC. What are you seeing across those?
Sure. So just to level set, mass spec quantifies the amount of frataxin. After you take a piece of tissue, you digest it, and it's a quantification, whereas IHC is looking at the distribution of frataxin across cells and, you know, specifically myocytes in the myocardium, so they tell you different things. What we're seeing is improvements in a dose-responsive way in protein expression as based on... And we've already been, you know, this is part of our, you know, our two kind of our readouts on the biopsies. We're seeing an increase in frataxin. That's by our second cohort that is close to 5% of normal, right on the border, a little bit above in some patients.
So in the second cohort of patients, which is still our E11 dose, we're seeing close to 5% of normal by IHC, and we're also- I'm sorry, by mass spec. And then, by IHC, which is exciting, we're seeing around 50%, more than 40%, around 50% of cells transduced, which we think based on animal studies, which we haven't talked about, but animal studies show that when you get 40% of your cells expressing frataxin, you, in these animal models, we improve cardiac output. So we already think that we're getting there in our second cohort, and now we have a third cohort, E12, a 1.2 E12 dose, that we're waiting on biopsies as well.
Yeah. And actually, the IHC assay results may explain the divergence of the cardiac, you know, treatment effect with the protein in cohort one. So as I described, cohort one is, you know, likely not at 5% of normal, and yet we're seeing these material improvements in LVMI, troponin, you know, wall thickness. However, in IHC, we are getting to a pretty meaningful degree of coverage, as Eric described. So maybe that, you know, you need frataxin in more cells, and the amount of frataxin can actually be below 5% in order to see this treatment effect, and that would be a thesis that's consistent with our with our clinical data.
Sorry, so, so the reason that the IHC numbers are so much higher than the mass spec numbers is because-
They're just telling you different things. IHC might... It detects. You know, the way IHC is detected is if it sees any pixels that are positive, that's a positive, right?
Okay. So it's not quantifying-
It's not quantifying-
The amount
... It's distribution. Yeah.
I think the other helpful aspect for IHC is that it has more regulatory precedent associated with it, so it's the assay that's been used across other gene therapy programs. We designed the LC-MS assay de novo for this particular program, so it's a new one. It does tell us about a % of protein relative to normal, but IHC has more regulatory precedent linked to it.
When we're referencing the mouse model 2% and 5% kind of expressions, that was mass spec, or... Okay.
Yeah.
Just confirming that. Now, Eric, you'd mentioned the LX1001 group dose had two cohorts? What are the two cohorts?
1.86 × 10^11
Okay. And then, I think you've kind of given us a bit of the high-level view of some of the data and observations. Maybe drill in a little bit deeper, you know, the number of patients at each dose that you've shown, the amount of follow-up that you've had thus far.
Yeah. I mean, in terms of the entirety of the study, cohort one is six patients. We presented data from those six patients. Cohort two will have, you know, four patients with cardiac biopsies and then another, you know, five patients without cardiac biopsies, so a total of nine patients in that cohort. Then cohort three will be three patients, and so that's the entirety of the study in terms of what we intend to enroll. In terms of the data itself, well, probably I'll pass to Eric to speak through the effect size we're seeing across LVMI, you know, wall thickness, and troponin.
And maybe if we could even just go back and start with the frataxin levels again, just keeping that 5% threshold in mind and how close you're getting to that at the various doses.
Sure. So in the first cohort, we saw significant percentage increases in frataxin, showing that we're having a dose effect, but below the 5% range. But by the second cohort, in the patients we've. And we've disclosed all this before, right at around 5% of normal frataxin in the patients that have been biopsied, thus far. And as I said before, in that second cohort, this is the 6E11, we're also seeing roughly around 50% of expression by IHC.
Yeah.
Okay. And then, in terms of ongoing dose escalation and where you can go from here?
Yeah. I think the goal would be to clear the 5% bar. That was the rationale for dose escalating. We think we're seeing an effect size across the endpoints that is clinically meaningful and should be acceptable to regulators. So for example, in left ventricular mass index, for the patients that started with elevated LVMI at baseline, we're seeing an 11.4% improvement. This is against what we believe the clinically meaningful threshold is of about 10%. We're seeing about a 14% improvement in lateral wall thickness. We're seeing a 50% improvement in troponin. So these are, you know, highly clinically meaningful effect sizes across these endpoints, but the one where we are potentially below the bar is on protein.
We're sort of sitting below, you know, 5% in frataxin via LC-MS. This was the decision to dose escalate to cohort three, where our goal would be to clear the 5% bar. This is all with the caveat that I described earlier, that it may be that IHC, in terms of the percentage of cells, is actually the right assay to evaluate, and there we are getting, you know, consistently above the 40%-
Right
... which we think is the bar for IHC.
When are the biopsies done?
Three months after treatment.
Would you expect to see pathologic improvement in terms of the number of mitochondria per cell, and what do you see there?
We are collecting other parameters, including electron microscopy, all those things. And so this is, you know, relatively new territory, so we'd see if we see a reduction in mitochondria, all those things. But I think for us, the endpoints clearly are the, you know, the amount of frataxin we would envision is more likely to be part of a endpoint package.
If you're talking about effects over time, I would point to the LVMI data as one example of that. I mean, at 12 months, we saw this 11.4% improvement in LVMI, but at 18 months, it was 18% improvement. So we're seeing a deepening of effect, you know, over time, and an increase and an improvement of the disease. So I think that's, you know, speaks to the effect that the therapy's having over longer time points as well.
You know, in thinking about your question, there are some enzymatic things you can do to look at mitochondrial function, all this stuff, but I think the best biomarker for us is troponin, 'cause we know cells under stress secrete troponin, cells that are dying secrete troponin. Cardiomyocytes specifically make troponin, and in this case, we saw close to a 50% reduction in troponin in all our cohorts. That would be very statistically unlikely to be play of chance, to see across multiple patients, to see this. You know, this is a disease where we know troponin is elevated, and see it reduced, unlikely to be, you know, some kind of. You know, troponin doesn't change with placebo. Troponin doesn't change.
It's a very quantifiable number in my mind, so to see that effect size is something that gives us confidence that we're seeing a biological effect of our therapy.
independent of hypertrophy of wall thickness and LVMI. I mean, it's a different focus that-
Yeah, I almost think of it, and we were talking about this earlier, as earlier in the disease path. You know, if there's a chain of events, troponin, we're seeing a uniformly, almost uniformly, as people start to get cardiomyopathy, elevated troponins, maybe before they manifest with measurable hypertrophy.
Are you seeing a dose response on troponin?
I think it's early to, you know, determine that. You know, I think the Cohort 2 data set needs to reach a certain level of maturity in order to determine any conclusions on dose response from the cardiac biomarkers. You know, we definitely are seeing a dose response in frataxin. So we are seeing higher levels of frataxin expression in Cohort 2 versus Cohort 1, so we'd expect the same, you know, in Cohort 3. But I think the data set is not at a level of maturity in Cohort 2 to kind of draw any formal conclusions on dose response yet.
Clinical parameters.
Yeah, clinical parameters.
When will the next update be, and what should we be looking for incrementally?
Yeah. I think, you know, we'll be guiding to the timing of the next update in the near future. It's likely to be in 2025 on the other side of the Cohort 3 biopsies, but we'll give formal guidance on that in the coming months.
We do have some templates for pivotal trial designs for genetic cardiomyopathies. What are you thinking for frataxin's likely trial size and endpoint or and/or co-primary endpoints, if you want to-
Yeah, so I think, you know, there is some precedent, as you, as you said, and for us, we think of a combined primary endpoint, potentially, that includes a clinical endpoint like LV mass. 10% reduction in LV mass would be something that we think, based on natural history studies of FA, as well as other cardiomyopathies, would be significant. And then, in terms of a biopsy parameter, whether that's IHC or mass spec, and then we've mentioned these kind of numbers we're thinking about, whether that's 5% or more frataxin, 40%-50% IHC distribution. The good news is that, you know, we think we're there or close to there already, so that makes us pretty excited about moving forward.
Do you see any wiggle room on that 5%? It sounds like, in theory, you may be able to lower that to hit a responder.
Yeah, I agree, and that's a discussion that we'll need to have with the agency. But I would also say it appears that IHC is more so correlated to the improvement in these endpoints that we're seeing. If you were to ask me today and look at the two assays, I think IHC is probably the one that makes the most sense because it has the regulatory precedent associated with it, and it appears to be more correlated with the improvement in the disease we're seeing, even at the lower doses. But I think more to come on that as we you know advance those discussions with the regulator.
Maybe we can turn to the PKP2 program and a quick review of the biology, and again, kind of in this context, why might it manifest as an arrhythmogenic cardiomyopathy?
Yeah, sure, so arrhythmogenic cardiomyopathy is one of the most common causes of genetic cardiomyopathy. It's manifest as having life-threatening ventricular arrhythmias. Patients initially present with feeling extra heartbeats or premature ventricular contractions. They go on to develop these arrhythmias, often when they exercise, and then, over a decade or so, go on to develop heart failure. The most common genetic cause of arrhythmogenic cardiomyopathy is the loss of a protein called plakophilin, plakophilin 2, specifically. That allows for the formation of the desmosome. Desmosomes allow myocytes to adhere to each other, so if you don't have plakophilin 2, you don't make desmosomes, myocytes stop adhering to each other, and they start drifting apart, and infiltrated between these myocytes that used to be connected is fat and fibrotic tissue, and that's an nidus for arrhythmia.
The hypothesis is pretty straightforward, and it's been proven out in numerous preclinical studies. If you restore plakophilin 2, you restore the desmosome, myocytes start sticking to each other, you reduce arrhythmia, you increase survival in our mouse model, at least, which is a mimic, which is a human mouse model, right? We took a human mutation, showed that those mice mimic human disease, and that they die prematurely, and they have a lot of arrhythmias. If we restore plakophilin, we can actually reverse disease. We prevent premature death, we reduce premature ventricular contractions, these extra heartbeats.
Yeah. I would also add just from a, you know, disease perspective, I mean, this is a, you know, sixty thousand patient rare disease-
Right.
In the US, so it's, you know, more than twice the size of Duchenne muscular dystrophy. So in addition to all of the aspects that Eric described in terms of the biology of the disease, the severity of the disease, and so on, this is a very substantial commercial potential program, potentially one of, if not the largest in gene therapy today. So I think it's one that over time, as our data set and others read out, it will get more attention from the investment community.
What do we know about the amount of Plakophilin expression required to normalize the... or?
Sure
Significantly reduce arrhythmias in the population? What, what gives you the confidence you can hit that?
Sure. This is a autosomal dominant disease, so we assume patients have roughly 50% of normal. And some autopsy data suggests that over time, they may start at 50%, but as they lose the desmosome, they start going down 40, 30, 'cause they stop even on the good copy of the gene, you could see they stop making it. And consistent with that, they actually stop making all these other desmosomal proteins as well, plakoglobin, desmoplakin. So all the proteins associated with the desmosome start being downregulated. What it looks like. Obviously, this has to come from preclinical models to how much do you need to be in normal, right? Because patients are either heterozygous or they're normal. In preclinical models, it looks roughly around 70%-80%.
So if you start at 50 or 40, you're trying to get to 70 or 80% of normal, probably 70 to 80%, you know, and same in terms of distribution, you probably wanna have a distribution of Plakophilin 2 across the, at least the right ventricle. Good news is, in our preclinical models in primates, we show really good VCN, really good distribution of this protein throughout the myocardium.
Yeah. But you know, that being said, I think in the clinical context, our goal with the work we're doing, especially at the Cohort 1 dose, is to increase plakophilin, you know, versus the pretreatment, you know, baseline. I think the correlation of the actual amounts of protein required and improvements in these, you know, clinically relevant endpoints, like premature ventricular contractions, today is unknown. And I think we'll have to observe that-
That's fair
Through the clinical study. We'll have to observe the level of protein expression we're achieving, and then the correlation of that to things like premature ventricular contraction. So I think the clinical side of it is still, you know, yet to be seen. And frankly, our data will play a role in helping to determine that threshold.
Are those target levels that, at least you theoretically proposed, IHC or mass spec?
Probably IHC again.
Yes
... but we're still... You know, this is first in man studies, where we're getting data and learning as we go. Because it's hard. There's not good biopsy databanks of these patients and things, so we're learning a lot.
How is the dose escalation strategy similar or different to the Friedreich's ataxia program?
I think, you know, there are factors that are very similar from a framework point of view for dose escalation, so the first one is, I think, you know, safety plays an important role in dose escalation. For the FA program, we saw no drug-related serious adverse events, so this allowed us to, you know, consider higher doses, and I think, you know, in this program, safety will be an important component of this disease. Because, you know, we're looking at 8e13 doses times adult patients, you know, the level of overall vector is meaningful, and therefore, safety readouts will be an important aspect of our decision to dose escalate.
Beyond this, I think protein will play a role, and I think, you know, looking at efficacy endpoints such as PVCs will also be a factor. But I would say, you know, safety is probably the most important factor in a dose escalation decision.
So it sounds like there's somewhat meaningful patient-to-patient variability. What about intra-patient variability? Like as we're starting to generate data focused on ECG or Holter monitors, I'm guessing is where you'll capture the PVCs. What kind of signal do you need to discern through what noise there may be in this setting?
Do you wanna-
Yeah, sure. I think in these patients, we know that they, on average, have around 500, at least 500 PVCs per day to be diagnosed with ARVC. That's part. You know, there's this universal statement, universal guidelines on making the diagnosis, and so with that being said, that's the threshold for just being in our trial. But we can see variability among these patients. You can see patients that have 2,000, 10,000, you know, PVCs that would be an extreme per day. The more you start with, the more likely you are to have a meaningful reduction. I think gives you more room to do that.
So I think we'll have to look at each of these patients that we enroll, see where they're starting at with PVCs, and look at the percentage reduction, but I don't anticipate getting more normal than normal. So if they start low, they might not have a percent, you know, as much reduction as someone that starts high.
Yeah, and to your point of variability within the patient, I think as with any endpoint, you want a sustained reduction. So, you know, even if you see PVC reductions at early time points, you wanna see that, you know, continues. And that would be, you know, probably the right way to think about if this were to be an endpoint for a registrational study, you would wanna, you know, see that you're having a sustained effect on the disease. PVCs may be the right way to show that, and you wanna see that over time, so.
Is there a run-in period where you kinda capture that variability per patient prior to enrollment, or is there just one point in time?
You're doing a week... You know, you're getting this over a week, of data before and then after.
Okay, great.
So yeah.
Are we still on track for a data update this year? Maybe you can help kinda characterize what we should be looking for, you know, which signals we should focus on at this stage, and which may just be too early to detect.
Yeah. So we will have data in the coming months. I think we're still looking for the right, you know, venue in which to present that data. We think, you know, this could be the first data presented for this disease or could be one of the first data sets, so we wanna make sure it's in front of the right audience, in the right setting to have an impact. In terms of what to expect, as I've mentioned, you know, we think safety is an important component of this, and we've seen it played out in other gene therapies. The difference in safety profiles made all the difference in the ultimate product profile, so that's an important, you know, readout for this.
We're also gonna be looking at biodistribution, so all the protein expression that we discussed, and I think it would be early at three months to expect an improvement in things like premature ventricular contractions. The gene therapy is just reaching peak expression at three months, so I think it may be early to expect an improvement in PVCs. That would be something that I think would be expected at future time points.
Got it. Would that be the first clinical symptom, or might troponin be an earlier biomarker, or?
Uh-
Any thoughts there?
... you know, troponin's not a—it's not quite the same in this disease as others, so PVCs are usual. You might have a characteristic abnormal EKG and then PVCs, but I think PVCs are probably... 'cause it's very quantifiable, non-invasive. You can do it, as you said, 'cause there's variability, so you can collect it over a period of time. So that's usually, I think, probably the best. The biopsy data, you know, you're collecting protein expression, but you can also look at these other proteins, part of the desmosome, and see they're upright. You know, in the preclinical data, at least, when you improve one protein, you improve the others as well, so that's something that you could quantify.
Maybe we can close with a very quick snapshot of the APOE4 program and the next updates from there.
Yes, sure. So, here we're delivering the APOE2 gene to the CNS of APOE4 homozygotes, with the thesis being that we're, you know, correcting the upstream genetics, and this has a downstream impact on several of the pathogenic mechanisms that are involved with Alzheimer's disease. We are looking at several biomarkers, one cerebrospinal fluid, amyloid beta, tau, and phospho-tau, which are, you know, commonly evaluated in Alzheimer's studies. We're also, we will have amyloid and tau PET scan data. And if those familiar with the area, amyloid PET scan was the accelerated approval endpoint for lecanemab. So, these PET scans are... could be, you know, clinically meaningful data as well.
What we're looking for is, you know, improvement across, you know, several biomarkers associated with the disease, and that's the readout we will be working towards this year, you know, likely at the CTAD conference. We will have cognitive decline data as part of the study, and that will be read out there as well. However, I'd say that the study's not powered to show a benefit in cognitive decline, so it's likely to be directional trends at best for that. The focus will be on the biomarkers that are commonly associated with Alzheimer's.
Got it. Well, we are out of time. So, Nolan and Eric, thank you so much.
Thank you
... for joining. Looking forward to some more updates from Lexeo this year.
Absolutely.
Big ones.
Thank you so much.