Next company I have, that I'm hosting a fireside chat with, to my left is Lexeo Therapeutics. It's a company that I cover, I am a big fan of. They're working in the gene therapy space with a lead program in Friedreich's Ataxia and another phase II program in Alzheimer's disease. So, Nolan, Eric, thank you guys very much for joining today. Maybe to just start, can you give the audience an overview of what Lexeo is focused on and what programs you have, at what stage of development?
Yes. So we're a Gene Therapy Company focused in genetic cardiovascular disease and a gene variant associated with Alzheimer's disease. On the cardiac side, our most advanced program is treating the cardiac pathology of Friedreich's Ataxia. This program's in an ongoing phase I to II study. We read out data in July of this year, which I think we'll spend some time talking about today. Our next most advanced program is treating arrhythmogenic cardiomyopathy .
This is another rare disease, but it's a relatively large one of about 60,000 patients here in the US. And here, we're in an ongoing phase I to II study as well. On the CNS side, our clinical stage program is treating APOE4 homozygous Alzheimer's disease, and so here we're delivering the protective APOE2 gene to the CNS of those patients. This phase I study has completed enrollment, and we're working towards a data readout by the end of the year. Three clinical-stage programs, two on the cardiac side, one on the CNS side, all with data updates expected this year.
Okay, great. I mean, maybe to start, I mean, it seems like maybe the magic sauce at Lexeo and, you know, forgive me if I'm... Or correct me if I'm putting words in your mouth, but it is sort of utilization of the AAVrh10 vector, seems to be what's unique relative to a lot of other companies who are pursuing AAV9 or various other vectors. I guess maybe can you give us a little background on that vector and, you know, why you think it's special, and differentiated compared to some of the other serotypes?
Sure. So, it you know comes from the original ReGenX portfolio. It's already been tested in humans in clinical trials prior to Lexeo using it with no significant safety signals. But I think the magic in that vector, in the selection, a lot of work went into looking at cardiac transduction. And what you see in comparison to AAV9, which had been kind of the gold standard for the cardiac space, is about twice the transduction in the heart. And that two X difference can have a really meaningful change in your ability to dose patients at the lowest dose possible and reduce your risk of adverse events, which we know is a bugaboo of gene therapy. So has this beneficial transduction, has some similarities to rh74, doesn't activate complement as much, it seems like. So I think it has this kind of Goldilocks capsid.
Yep.
I also would add, I mean, from a validation perspective, a lot of the lines of evidence to support rh10 were in preclinical models, multiple large animal studies, small animals, and so on. But I think via the readout we now have in our FA program, we're seeing clinical validation of this vector for use in cardiac.
Mm-hmm.
In that, you know, all of the patients in our FA cardiac study were treated with AAVrh10, and we're beginning to show, you know, early signs of efficacy in that study.
Great, so that's a good segue into LX2006, designed to deliver frataxin to cardiac muscle in FA patients. Maybe you could start by talking a little bit about the disease pathology and why you see this approach making sense for FA.
Sure. Yeah, the hypothesis here is pretty reductionist in that we know the root cause of this disease is a deficiency in the protein frataxin. Frataxin essentially has an enzymatic function. It allows for the binding of iron and sulfur within mitochondria, and without iron-sulfur clusters in mitochondria, they don't function, so you're missing the engine of your cell. Heart cells and nervous system cells rely heavily on energetics, so without frataxin, their energetics is abnormal, so the thesis is pretty straightforward: replace the frataxin, prove that deficiency, improve mitochondrial function, and improve cardiac function in this case.
Okay. I think a lot of people kind of think of FA, to some extent as a neurological condition. Certainly, there's a big neuro component to it. How many patients with FA develop cardiomyopathy?
So-
... what sort of the slice of the market that's-
We think that actually, the majority of the patients with FA, we know now die from the cardiomyopathy, though they obviously have a large impact on their quality of life from the Ataxia. In 70% or approximately 70% of the patients, cardiomyopathy and heart failure is the cause of death, so it's significant. The patients recognize it when we interact with them, that they say their scariest appointment is with the cardiologist, 'cause that's the one that's gonna limit their lifespan.
Okay. So currently, you're running the SUNRISE-FA study. We've seen a number of data points so far in two doses, initially in patients with cardiomyopathy. Maybe you can walk us through, to start the trial design and you know, what are the goals in terms of what you want to find out in this study?
Sure. So obviously, phase I study, looking primarily at safety of this capsid and the construct in our patients, but we're collecting a number of biomarkers, which we think are important for the development of this drug to treat cardiomyopathy. We know, for this is a hypertrophic cardiomyopathy phenocopy. In other words, these patients develop heavy, heavy hearts, so we're evaluating LV mass, and we're looking at other measures of hypertrophy, such as wall thickness.
We're looking at biomarkers. We know that patients with Friedreich's Ataxia have abnormal troponin in their blood, which is a sign that their cardiomyocytes are sick and under stress, and in some cases, dying, and so we can measure that as well. Those are the primary things I should say, and I forgot to mention importantly, we're biopsying these patients, and the biopsy allows us to confirm that we're getting transduction and appropriate amount of frataxin expression, both from a distribution and a quantity.
... Got it. So that's your study. There's also a parallel study running that Weill Cornell is running, and you've licensed the data for that study as well. I guess, can you compare and contrast the two? You know, what are the major differences between the two studies?
I think they're almost identical, except, you know, except that the, the Cornell study is not doing biopsies. But other than that, from a material, in terms of data collection, almost identical. So-
Okay, so I mean, this is-
Partically-
a monogenic disease, you know, frataxin is the primary driver or protein that's absent here. How do you think about measuring frataxin? And, you know, what sort of levels can we anchor to, to say, you know, X frataxin level in X tissue, Y tissue, you know, indicates some sort of phenotypic modification?
Yeah, and we've done a fair amount of work at this, looking at normal frataxin levels and looking at animal models where there's no frataxin in preclinical models, and what we think are two ways to look at it. One is the quantity of frataxin, and that we're measuring by mass spectroscopy.
Mm-hmm.
And then the other is the distribution of frataxin. So are cells expressing it in the heart, And that we're measuring by immunohistochemistry, and we think both of those are important. Preclinical models, some clinical data suggests that as little as 5% of normal is what we think or hypothesizing is what you need to improve cardiac function. That, again, goes back to our fundamental hypothesis, which is this is an enzyme that's allowing this binding of iron and sulfur, and as such, you don't need a ton. And in fact, if you have too much, you might have toxicity. That's been a bugaboo for other programs.
Mm-hmm.
So we don't think you need a lot in terms of quantity. In terms of distribution, we also think you need... You know, the preclinical models suggest, and we're measuring this by, as I said, immunohistochemistry, around 40% distribution of cells expressing this or areas expressing that, we think is sufficient based on preclinical models to improve function. But as Nolan point out, our initial data suggests is supportive that you may not need a lot of this protein to see these clinical endpoints that are relevant, like LV mass. Even in our initial cohort of patients, we saw with relatively low amounts of protein expression, you're seeing clinically relevant improvements in LV mass.
Yeah.
I would add, I mean, if you needed 5% of normal to correct the disease or show an improvement, then we should not be seeing an improvement in our Cohort One patients.
Right.
Yet we are. That thesis of 5% is based on preclinical evidence, and maybe that the clinical picture looks different. The only way to explain the effect we're seeing in Cohort One is by looking at the distribution or the immunohistochemistry assay.
Yep.
We're getting to about 50% of normal. The preclinical lines of evidence would suggest you need 40% or higher. That would actually explain why we're seeing this benefit within the Cohort One, Two patients, despite them not being at 5% of normal via LCMS. So maybe that, you know, historically, we have oriented a little bit too much on the wrong assay, that actually IHC may be the assay that, you know, determines the picture of correction in the disease more so than LCMS.
Okay, great. So yeah, I mean, maybe that's a good segue into actually talking through the data and, you know, can you maybe to start, just talk a little bit about in the patients that have seen the biopsy in SUNRISE-FA by dose, what you're seeing in terms of frataxin expression levels on Western, what you're seeing on IHC, and how you think-
And maybe just even taking a step back, I think just highlighting safety, right? 'Cause that's the first thing, and so we saw no safety signals. So I think that's really important anytime you're in gene therapy. Then going into the biopsies, we looked at these two modes of expression, whether that was Mass Spec or IHC, and in both cases, we saw improvement. So, Mass Spec in our Cohort Two, we're getting close to that, if not above that 5% range in improvement.
Mm-hmm.
Significant percentile improvements, but also getting to that [BOE] of 5% or so. Then in terms of IHC, again, if we're saying around 40% is what we think we need, we don't know, but this is preclinical models, we're getting closer to 50% in this second cohort. Keep in mind, we're going up to a, a, to an additional cohort, so we expect to see those biopsy parameters to improve even more as we go up, if they correlate with the preclinical work we have thus far. That's the biopsy data. Then if you allow me, I'll go into the clinical data.
Yeah.
In the clinical data, we had the measurements, the biomarkers of interest, which I mentioned, which is LV mass, wall thickness, and troponin. LV mass, we saw a reduction of around, 10%- 11%, and that gets better over time. So as we go from six months to 12 months to 18 months, we see this continued improvement. That's supportive of this idea that we're causing remodeling. We know that takes time. So as patients get further out, we're seeing improvements in their cardiac weight. That's important because we know across cardiomyopathies, increases in LV mass are associated with worse outcomes. So we think it's a very compelling endpoint with regulatory guidance that we can lean into when we have discussions with the FDA. We looked at wall thickness.
Wall thickness is now, as opposed to measuring the mass of the whole heart, looking at one specific area. The reason for that is sometimes wall thickness increases before the whole LV mass changes. So it allows us to look at patients that we enrolled that were maybe at an earlier stage of disease.
Mm-hmm.
You can think of it like an earlier stage cancer when we're looking at wall thickness. In this case, wall thickness improved by nearly 14%. So again, a measure of a biological effect of our drug. We wouldn't expect that to happen spontaneously. Wall thickness by MRI, the measurement we have, doesn't have that degree of inter-user variability, unlikely to be by chance. And then finally, which I think is compelling data as well, is Troponin. So Troponin is down by 50% in our patients. That's a robust amount. Again, unlikely to be play of chance in my mind, and also one where there's guidance of using that as a secondary endpoint, 'cause we know that biomarker across any kind of disease associated with worse outcomes.
Yep. Yeah.
So those are the, I think, the primary endpoints that we showed.
Just in terms of effect size, I think it's important to highlight. So, you know, an LVMI, a 10% improvement in LVMI is a clinically relevant measure as evidenced by another genetic cardiomyopathy, where the FDA has agreed on that as a co-primary endpoint, and 10% is the threshold. We're reaching 11.4% at 12 months already, within the patients in our study. Troponin, there's another program, secondary endpoint is Troponin. The 30% bar for that, we're reaching 50% already in our study. So I think, you know, we're clearing a clinically meaningful bar that's validated by, you know, regulatory precedents.
Mm-hmm.
Across the two key endpoints, two of the three key endpoints that Eric described. So we're excited by the results and think that this, you know, gives us good scope for a great discussion with the FDA.
Great. And then, I mean, I guess that's a kind of segue or maybe answers the question in terms of what the regulatory framework is and having a discussion with the FDA. But obviously, there's been a flexibility applied to both [omaveloxolone] and the approval of Skyclarys, as well as
Yeah
... flexibility in and of itself with gene therapy. You know, how do you kind of see that, those regulatory frameworks coming into play here? I mean, it seems like you would pursue almost a full approval with LVMI as a potential marker, but-
I-
Is Frataxin in play as-
Yeah, I think it's working towards an accelerated approval.
Okay.
Which is the pathway that, you know, I think a lot of the gene therapies are utilizing today. And as part of that accelerated approval pathway, we would seek to reach alignment on with the FDA on a, imaging endpoint, such as LVMI, as a, as a co-primary combined with, protein expression. And that's a. You know, there's a precedent that exists around that. Sarepta is one. There's another genetic cardiomyopathy with that precedent behind it.
So that's the framework we'd seek to reach first on the surrogate endpoints. Once we have the biopsy data from our Cohort three patients, I think that's the time point in which we would seek alignment on the, you know, full design of the final pivotal study and therefore, put us in a position to move forward. So I think of it as a two-step process. Step one is the endpoints themselves, and then step two is the final dose and the ultimate size of the study, in order to take it forward into the pivotal.
Got it. And then, obviously, part of the focus of rh10 is its cardiotropism.
Mm.
But it does get expression in other tissues as well. At a certain dose level, do you start considering looking at some of the neurological components here? Is that a possibility?
I think striated muscle is pretty interesting, and-
Yeah
... In the past, I think people just attribute everything to the ataxic components of disease, but we know this is a muscle disease as well. So we're not looking at it, but, you know, it wouldn't surprise me if you saw improvements in quality of life based on improving not just the cardiac tropism or cardiac transduction, but striated muscle as well.
Yeah. I mean, we're collecting the neurologic scales, you know, mFARS and so on. The therapy is not designed to benefit the neurologic disease. It's our thesis that there may be other modalities better suited to treat the neurologic disease than gene therapy, for example, because you need to transduce a certain number of cells, for example, in the deep cerebellum. It's unlikely that a systemically administered gene therapy would be able to achieve that. So we were looking at the neurologic scales, but you know, I wouldn't set an expectation that there would be any benefit on the kind of neurologic disease progression.
Okay. And then in terms of the decision to move up to a third dose cohort, other than me pestering you repeatedly to do so, can you kind of walk through the rationale behind that?
That was it.
That was it.
I'll say, you know, I spoke to LVMI. We're clearing the clinically meaningful threshold that's validated by-
Yep
... other, you know, other studies. I spoke to Troponin, clearing the threshold. You know, what I didn't say is that we're clearing this "5% bar" on protein.
Right.
We're getting just to about 5%. So it would be our goal with the third dose to clear the 5% bar with that dose, and that was the primary reason-
Yep
... you know, for dose escalating, was just to ensure we're, you know, achieving a level of protein expression that could clear that, that threshold for a pivotal study. Albeit with the caveat of what I said earlier, that it may be that LCMS is not the right way to look at this, and IHC is the appropriate way to look at it. If that's the case, the 50% to which we're getting to today would be, you know, sufficient-
Yep
... and actually corresponds to the effect that we're seeing in some of the cardiac clinical biomarkers.
Okay. And then, just in terms of the upcoming data points, how do we expect sort of a cadence of updates from, you know, the data that the patients you've already treated through to the patients that you're dosing in cohort?
You know, I think what we can guide to this year is, we'll be presenting the third biopsy in Cohort two at an academic conference. We will also be giving an update on the sort of regulatory dialogue that's ongoing. So those are the updates that could be expected within 2024. I think, you know, longer time points of Cohort two, additional patients, Cohort three data, that would be a 2025 update. I think, you know, closer to the time of that update will give guidance on the specific, you know, quarter in which we'll give them.
Okay, great. Moving on to your next cardiomyopathy program, [LX2020]. Maybe you could just walk us through this disease, how twenty-twenty is sort of designed to hopefully correct for some of the phenotypic issues here, and maybe even how the FA program could help translate the risk of this program.
Yeah. So arrhythmogenic cardiomyopathy is a significant cause of morbidity and mortality in patients with heart disease. It's usually the most common genetic cause of arrhythmogenic cardiomyopathy is are mutations in a gene called plakophilin 2. This affects between 60,000 toF 80,000 patients in the US alone, and that doesn't include, as I said, Europe, which has a large, well-defined population. It's a very morbid disease, so patients have sudden death often when exercising. They develop right-sided heart failure and often require heart transplantation. They have severe exercise restrictions. In summary, we think it's probably the largest opportunity in cardiac gene therapy, if not all of gene therapy currently being evaluated, so a great target for us. The biology in this case is also or the thesis is, again, reductionist.
It's that these patients don't have enough of this protein plakophilin. Plakophilin is a component of the cardiac desmosome, which allows cells to stick together. So if you're missing plakophilin, cells, cardiomyocytes can't stick to each other, and they start drifting apart, and you have infiltration of fat into the heart, and that's an nidus for arrhythmia. So the thesis is we can restore plakophilin, we restore the Desmosome, cells start to stick together, you see regression of fat, and that's been now confirmed in multiple, very robust preclinical models, multiple investigators, US, Europe, all showing the same thing. In plakophilin-related deficiency in murine models, if you improve it, you improve cardiac function. So it's a really exciting program with a really strong preclinical data set, a really high unmet need.
And you asked about the relationship between-
Okay.
FA and this program. We are using AAVrh10 for this program as well. So I think the clinical validation that we're seeing from FA certainly will translate to PKP2. Albeit we are using a higher dose because it's a different protein function. This is a structural protein versus frataxin's enzymatic, you know, function. We're also using a different promoter that's cardiac specific, but nevertheless, I think some of the learnings on cardiac tropism that we have from FA clearly give us a read-through to this program as well.
Walk through the clinical plan here, the design of the phase I study, and what can we learn in this indication from the initial move?
Sure. So the phase I study, obviously, the most important thing for us is safety. We're using a higher dose, and this is the highest dose I believe used for rh10, so I think that's the primary goal. But obviously, we are doing, or not obviously, but we're doing cardiac biopsies in these patients, so we should get some sense of tissue transduction using a cardiac-specific promoter. So we'll be able to evaluate transduction in the context of a cardiac promoter, as opposed to a ubiquitous promoter like we used in FA. We'll be able to evaluate not just the expression of this protein, but other proteins associated with the desmosome and see if those are all restored, 'cause we know from human biopsy data that they're all reduced.
So every component of the desmosome goes down when you're missing PKP, so we can look at those. And then ultimately, you know, obviously, readouts that we're thinking of clinical importance, the most important one is probably premature ventricular contraction.
Mm-hmm.
That's extra heartbeats. We all have extra heartbeats, skip beats. Patients with arrhythmogenic cardiomyopathy have, at a minimum, around 500 of these a day, and they're noticeable. They cause an impact on their quality of life. The more of these you have, the more likely you are to die or get sudden death. So we're able to quantify them, so we can precisely measure the amount they have over a week-long period, and we hope to show improvement in that as a... we could see that as a endpoint. The reason we think it's such a robust endpoint is there's already some regulatory guidance around it, so there's investigator-initiated studies where the FDA has agreed that reduction in PVCs is a suitable clinical endpoint as a, as a, a surrogate for survival and quality of life, I should say.
Great. Are you just focusing on one dose? Is there any escalation potential in this?
Yeah, sorry. So there's two cohorts with two E 13 and six E 13, but we saw, you know, evidence of improvement in the preclinical models at this two E 13 dose, so we think it's an exciting dose for us to start with.
Yeah. So, the low-dose data, you know, I think will be coming over the forthcoming months. I mean, it is, you know, it's three-month data. It is the lowest dose. I think safety is the focus. We'll also, you know, speak to biodistribution as part of that readout. I think, you know, we will dose, likely dose escalate to the forthcoming dose, and I think the collection of that data across both doses is probably the data set that shows the most compelling picture of efficacy. So I think at the early stage, we're really focused on biodistribution as the, you know, as the focal point of the early readout.
Right. And maybe it came from earlier in just what the difference is between Lexeo and some of the other companies, but thoughts around differentiation from other PKP2 gene therapies, and you know what?
Yeah, the biggest difference is each company using a different capsid, a different serotype, and I think each capsid has its own advantages and disadvantages. We have experience now with AAVrh10 that gives us a lot of confidence in the capsid. You know, the preclinical lines of evidence to support the cardiac tropism profile were very compelling. We're now seeing the clinical picture playing out in line with that. So we would expect that to be an advantage in this program as well. In our view, you know, if you're going to convince 60,000 people to consider being treated with gene therapy, you need a therapy that is safe. In our view, the therapy that's the safest is ultimately the one that will be the best-in-class treatment.
And safety in this context probably means the ability to achieve disease correction, but do so at the lowest dose. And that would also correspond to potentially the least immune suppression, 'cause obviously, the immune suppression also comes with its own safety risks as well. So that's what our goal is here, to achieve a compelling safety profile, while also, you know, correcting the disease in the way that we saw in the preclinical studies.
Got it. And then in the last couple of minutes, you're also investigating an APOE gene therapy for patients with Alzheimer's disease. I guess, technically, it's the lead program. You have the most patients dosed, right? Can you walk us through the rationale here and the basics of sort of this APOE biology-
Yeah, yeah.
-and how you're potentially addressing a subgroup of-
Yeah, it's very interesting. You know, so basically, the APOE gene is a major determinant of risk or protection against Alzheimer's disease. Most of the population are APOE3s. APOE4s, especially homozygotes, have a 15x higher likelihood of developing Alzheimer's disease. On the other end, there's APOE2s, which are about 8% of the population. They have the lowest risk of developing the disease if they develop it at all in their lifetime, but interestingly, if you're an E2/E4 heterozygote, the existence of E2 removes the E4 risk, and you go back to normal, so this is the sort of genetic thesis behind the program, where we're using APOE2 as a therapeutic, that which we believe can stop or slow many of the pathogenic processes that are believed to be associated with Alzheimer's disease.
This study, we read out the low-dose data associated with this at CTAD in 2022. By adding the APOE2 gene, and this is at our lowest dose, we showed a reduction in amyloid tau and phospho-tau. The degree of the reductions, you know, were meaningful. For example, in tau, we saw a 15% reduction. You know, I think lecanemab's reduction was 15%. We already saw this at our lowest dose, albeit this was a small N, it was only three patients, so all of the caveats, but I think the early data was interesting. Now we've enrolled the next three cohorts, including our highest dose. We've completed enrollment of the study.
This is data that will be read out this year, you know, likely at the CTAD conference again in Q4. And what we're looking for in this study is the thesis here is that correcting the genetics has a downstream impact on multiple different pathogenic mechanisms of the disease simultaneously: tau, amyloid, you know, phospho-tau, and so on. So we're gonna be looking at those biomarkers. Are we observing an improvement in amyloid, tau, phospho-tau, and so on? I think those are the key biomarkers that we'll be evaluating. In addition, we'll have PET scans of both amyloid and tau. And to note, obviously, lecanemab and aducanumab were accelerated approved based on amyloid PET scans.
So to the extent there is an effect seen in PET scans, it could open the door to an accelerated approval pathway for this program as well, albeit it's not a rare disease. It's obviously, you know, one that has substantial unmet need, in particular, in this APOE4 population.
Yep.
Because the amyloid antibodies have not shown compelling efficacy within that population, but they also have a substantially higher risk of ARIA, which is the brain swelling disease, the most serious side effect of the amyloid antibodies. So with donanemab , there's a 40% risk of ARIA and no statistically significant improvement in efficacy. So from a risk-benefit perspective, if you're an APOE4-
Right
... you probably don't opt for, you know, such a treatment. Well, this is one of the most advanced, you know, clinical stage precision medicines treating APOE4 homozygotes, so this may be the best, you know, modality to treat patients of that genetic profile. So this is the data that we're reading out at CTAD this year. I think, you know, for the Alzheimer's field, for people that are interested in kind of the whole kind of neurodegeneration area, this is a very exciting readout because it'll link the genetics to, you know, what is a very complex disease in a way that I don't think has been done, you know, previously, so.
Great. And then I guess in the last couple of minutes, can we just revisit sort of the financials?
Yeah
... what your current cash position is, and you know, what sort of cadence of data points that is able to walk us through?
Yeah, so we have current cash of $175 million on our balance sheet. This gives us a runway into 2024, sorry, 2027. Apologies.
Yeah.
It should be longer than that. 2027. So this puts us on the other side of the completion of our, you know, phase I/II study in FA. This gives us the capital to start initiation efforts of the pivotal study for FA. This puts us on the other side of our phase I/II study for PKP2 and the data readouts there. It also puts us on the other side, obviously, the Alzheimer's readout, which will occur this year. So a number of different, you know, clinical data catalysts that would occur within this runway, many of them in the next, you know, six to 12 months, for the company. So I think it's a very exciting, you know, picture ahead of us across both CNS and the cardiac programs.
Great! Well, with that, I guess we could leave it there. Thanks so much for the time today.