Dyne Therapeutics, Inc. (DYN) Earnings Call Transcript & Summary

Welcome, everyone, to the 43rd Annual JPMorgan Healthcare Conference. My name is Tessa Romero, and I'm one of the senior biotech analysts here at JPMorgan. We're pleased to welcome to the stage Dyne Therapeutics. And presenting on behalf of the company, we have President and CEO, John Cox. John?

Thank you Tess. My name is John Cox. I'm the CEO of Dyne. It's a pleasure to be here. We are trying to build a leading neuromuscular disease company, and we think we're on the verge of doing it. We had an event just last Friday, where we went into great detail on our clinical data around DM1. And we're going to show some of those slides and some basic slides on the company, but I'm going to move through those fairly quickly because some of the questions people are asking is about our registrational trial and around the novel surrogate endpoint. And I really want to spend a few minutes on those slides and add some additional detail to make sure people have clarity and the kind of confidence that we have in what we're doing going forward. We will have some forward-looking statements. Just briefly on the company, kind of the value proposition of the company is this: This is a company that has 2 programs late stage, both in areas where there is significant unmet need, DM1 and DMD, both with tremendous opportunity and with a path to approval and launch as early as 2027 and 2 programs. That transforms the company when we do that, and it transform the lives of patients, and I think it's very important for shareholders. It's all based upon a very differentiated platform, which I'm going to touch on in a moment. And I'll just say we fully own our assets, and we're well positioned financially. Now let me touch for those newer to the story, just a bit on the platform. The platform is novel. It is a platform that uses an antibody fragment to target a transferrin receptor, one that is found throughout all kinds of muscle tissue and also found in the CNS. It was specifically designed to hit a particular epitope on the transferrin receptor, which is designed to transfer iron into cells throughout the body. But to hit that particular epitope for a very short period of time and transfer a payload into tissue. And that's what we've been doing. And we've shown in animal models, and now we have shown it in humans and a very important point around this is that the payload is something that we can interchange. We are using a payload for DMD that's an oligonucleotide called a PMO. We use the same exact platform now for DM1, except we are using an antisense oligonucleotide. And we're using that for a reason because ASOs get to the nucleus, and that is where the core pathobiology of DM1 exists, and we're trying to address that. The final point I'll make is we not only get to skeletal muscle and cardiac muscle, we get to the CNS in animal models. And one thing we have been wondering is, would we be able to not only get broad pharmacology across muscle tissue but would we see signals in the CNS? And I'm going to touch on that in a moment because we're excited about it. Here we go. Just quickly on some of the preclinical data, and I'll move through this pretty rapidly. But people have known for a very long time in genetic diseases, neuromuscular diseases. They've known the genetics. They've known the oligonucleotide sequence that could direct to it and deal with it. The problem has been delivery. When you put these oligonucleotides naked into the bloodstream through an IV, you do not get the distribution. Distribution has been the challenge. The core of this company has been distribution of genetic medicines broadly into tissue, could we overcome that issue in humans? Here you see in animals, 25 to 50 fold the distribution and penetration of oligonucleotide in the diaphragm, heart, in other words, in the skeletal muscle versus a naked oligonucleotide. And then below, we're showing in the CNS, similar type of penetration. That penetration is if you compare it to an intrathecal delivery of a genetic medicine, this is deeper, broadly distributed throughout the brain. So a remarkable distribution profile for our platform. A couple of points now on DM1, and these are kind of the summaries. First, we have completed our multi-ascending dose study, and we are now moving into our registrational expansion cohort. We have identified a primary surrogate endpoint of splicing correction at just 3 months because we believe it predicts broad functional benefit, including patient-reported outcomes. And we anticipate being fully enrolled by the middle of 2025. The dose that was selected, we have gone through a very disciplined effort to select the dose guided by splicing correction in the tissue of these patients to select a dose that would -- we thought would best predict functional benefit. We've made that decision. Our CMO here, Doug Kerr, has done that, and he'll be up here on the stage with me in a few minutes for Q&A. And then finally, we have a significantly favorable safety profile. There have been 0 serious related TEAEs at any dose, at any time point in this particular study. Now let me just touch for a moment on DM1. This is a disease that is really caused by a toxic or a mutant DMPK that is found in the nucleus. It is genetically caused. Its manifestations are varied, and you can see that. High picture on this or our top point on this is the life of these patients is shortened by 20-plus years. And that life is affected in many different ways, including myotonia, which is just an ability to relax the muscle. But muscle weakness, cardiac arrhythmia, pulmonary issues and very importantly, CNS manifestations, devastate their lives, and it's a very significantly -- the number of patients affected by this disease is over 100,000 patients across the U.S. and Europe. Now here is the central pathobiology that DYNE-101 was designed to address. First, you have to be able to deliver the oligonucleotide broadly to these tissues. We've just described that. You can see this hairpin loop that is caused by a CUG repeater expansion of DMPK. That hairpin loop in and of itself isn't the problem. The problem is that it binds splicing proteins known as MBNL. And if those are bound, you do not get splicing across a broad set of genes and the result is the heterogeneity that you see with this particular disease. Our view has been if we could design an oligonucleotide to bind to that hairpin loop and unleash these proteins, you would first start to see myotonia effect. Modest splicing correction, you see myotonia effect. From there, the view was if you address that central biology, you should see broad functional benefit around strength, functional outcomes and patient-reported outcomes. And that's what we're going to talk about and show because we think we've proven that and addressed it. This just points to the heterogeneity again of the disease. The pathology is complex, so are the manifestations. I'm not going to go through all of these. Our goal is to address broadly the tissue by the heterogeneity of the disease by getting to the tissue and ultimately getting to the nucleus. And I think we've seen that in our clinical trial. Now let me make a couple of points about the ACHIEVE trial and particularly the multi-ascending dose portion. That portion is now complete. ACHIEVE trial was designed, and we need to get clarity on this. It was designed with really 2 major components: one, multi-ascending dose, to select a dose in a rational way that would affect broadly the disease. We've done that. Using splicing to guide us at that selection. The second piece of that purpose of the multi-ascending dose was to determine whether you could use splicing which has been shown through a measure called CASI in natural history to project the deterioration of the disease, could you use that splicing as a surrogate endpoint for an accelerated approval path. We think we've demonstrated both of those points by completing the multi-ascending dose. The second part of this entire program was that within the same protocol, we could then switch to a registrational expansion cohort using all of that patient data from the MAD as they move forward as 1 cohort at the new dose and then have an expanded cohort of x number of patients at the dose which we selected at 6.8 and moved that as a robust package forward for accelerated approval. That has always been the plan. I'm not going to go through the safety slide in any detail, except as I've said before, no serious events or TEA events related to the drug. 855 doses, we have a building safety database, 72 patient years of follow-up and Doug can talk about that in any more detail. So let me come back to the central pathobiology and this point again. In drug development, getting after the mechanism of action of this disease; first, the question is, are you getting the drug to the tissue? What you see on the left side is the ASO is getting to the tissue in humans. And this is all at the 6.8 mg per kg dose that we have selected for registration. The second question, is it having a pharmacological effect? Is it engaging with the target? Target engagement is seen by knocking down the DMPK, profound DMPK knockdown at 3 months, 6.8 mg per kg dose. And then you ask, what pharmacological insights can you gain? Do you correct splicing? Splicing is corrected here, 25% improvement. I'm going to add a point here for folks that wanted to get in the stats. This slide has slightly changed from what you saw on the previous event tech because we wanted to put absolute numbers. We still have the percent improvement, which we think is the important point, 25% improvement on CASI. But for those that want to do stats and understand error bars, you now have them, they are right here. The question from that is, do you get broad functional change if you get that kind of splicing improvement? We've got that on the right-hand side, and we can show that we correlate that at 6 months along strength, along time function measures and you'll even see on a patient-reported outcome, a significant change at 6 months. Now some of the questions were, okay, if you're going to be counting on CASI as a surrogate endpoint, nobody has ever done it before. What's the risk? How confident are you in it? We're confident in it. And there's a few reasons. On the left-hand side, we've just taken from some of the natural history papers, just examples that show that CASI correlates very well across multiple measures for showing deterioration of the disease over time. We have, in the middle, just information about CASI as a surrogate endpoint in our hands. The assay is reliable. The variability in the assay itself, very low, and we've shown it to you here. The assay has been part of our package with the FDA and something we've even shown and discussed with the FDA. The muscle biopsies that we take, some concerns can be that you have muscle biopsies, you could have variability or you could have failures. And we had 1 cohort that we showed as an outlier that was contributed to by some of these failures. So the question was, if you have those kinds of issues, how do you know in a registrational cohort it's not going to be popping up and being a problem for that? We've accounted for it. But if you go through even looking at the rate of QC failures of the biopsies - very low. And then if you look for discordance, which we do across the same biopsy and taking multiple slices of that biopsy - very rare that you have discordance. So the assay is something we feel very robust about. The other question was, does it correlate with a function? It doesn't just correlate with a function. Myotonia easy. Does it correlate with quantitative muscle testing strength? Yes, it does. Does it correlate with 5x sit to stand? Does it predict in a reasonable way, functional benefit, which is the bar for accelerated approval? And the answer on all 4 of those is yes. And for those of you who wanted to see some stats, our values, and you want to see some pretty good P values here, they're right there on the page. So we feel good about CASI. Now let me get to some of the data. This is some data that Doug had presented just last Friday, and I've added even -- we've added a little more detail to it again. Again, back to the central biology, the surrogate endpoint is the -- is now in our view, now that we have demonstrated for the first time that you could use a surrogate endpoint in a reliable way and predict multiple functional benefits, we're going to use it. And you can see on the left-hand side, CASI at 3 months; again, gives you all the data you need. And if you look to the right, what you see across multiple functional aspects, including PROs, significant improvement to the right with the treated patients and from the no-effect line, movement to the left on placebo. And we've done one other piece to add some clarity for people. And that is -- in this field because there hasn't been a drug for patients, you can't just go out and say, what's the minimal clinically meaningful change. There's no standard established. So the question is, are we having effect that is meaningful to patients? We've used standard statistical models. You'll see in the footnote how we did it just to put in some projected MCIDs. So you can see that we've exceeded that. It's statistically based, but that's what we have to do. And you can see on every 1 of these, our means are moving to the right of what is statistically meaningful. All of that comes back to what we believe we have to demonstrate for the FDA and gives us the confidence to move forward with selecting our registrational dose. Now the MDHI, which I showed on the previous slide, I probably should spend a little more time on it. Doug can talk about it. It is a patient-reported outcome that we think is this really the gold standard in our view, for patient-reported outcomes because it includes everything like from daily living to all kinds of functional activities, but also includes multiple measures on CNS and this is important. It gets back to the CNS point that we were making before. When we looked at the 6.8 mg per kg dose, we got excited about it. The functional measures look good. We have seen good functional measures and good CASI improvement as well at the 5.4 mg per kg but what really stood out and differentiated is this. And when we show this to clinicians and Doug is one of those clinicians that's treated DM1 patients, that change means something. Now it's 6 patients. But we're not going to leave that on the table. We're going to go for that as well as the functional benefit you saw because we know how important this is to the effect on people's lives with DM1, and we're going to pursue it. And we believe that it is possible because that's what our animal data told us that we could get to the CNS. Now I'm going to spend a minute on this slide, too, because I think there's been questions about the regulatory aspects and robustness of how we're approaching it. In our event, we only showed you 1 portion of the regulatory strategy, which was what we were doing for accelerated approval. We had said the number of patients was approximately 32. When we selected 32 not as a final number, we actually have a range. We can go higher in the number, but 32 was a number that gave us statistical significance powering for our surrogate endpoint. We can increase that. We can increase that to increase the probability of other functions having whether it's statistical significance or a meaningful correlation that would be something that the FDA would view in an accelerated approval and the type of thing that we might get on the label, we've listed all these on the right. But that number is 32 to 48. We have put on the bottom our powering information. So for those who just being transparent for those who want to do the stats, and see what our Biostat folks have done and how we are powering it, how we're thinking about this trial and why we have confidence in it, you have it. But I'll take it even further. What we also didn't talk about is that the multi-ascending dose portion of that trial. All 56 patients get moved up to 6.8 dose Q8W. All 56. Those 56 are moving forward, concurrent with, we're continuing with those with the registrational cohort that could be from 32 to 48. You put that all together, that's roughly 100 or more patients in an accelerated approval package that we would have at the time of the 6-month endpoint that is listed here following the full enrollment that you see on that slide, okay? So it is extremely in our view, powerful, robust package for the FDA. The other piece that we have not talked about is that with this data that we just put together for you in December -- well, we were putting it together in December, the completion of the multi-ascending dose study, we unveiled it just this last week. With that package of data, we want to sit down with the FDA for an end of Phase II type of study so that we can start what you see on the bottom, which is a Phase III study. The end of 200 is a rough estimate. We are going to -- but the point is it will be a fulsome study that takes everything that we have been saying so far in the accelerated approval and the functions, allows us to build on that and gives Doug and the team a chance to say, You know what, let's really study CNS in depth. Let's see if we can deal with the sleep issues. Let's see if we can really prove the cognitive issues out. And let's also get into the cardiac issues that are often the reason for death of these patients. So that is the entire kind of package. It's a little more complex, but it gives us a chance to move very rapidly with a robust accelerated approval package and follow it up with a concurrent Phase III that we would like to be starting this year once we have good alignment with the FDA. I hope that clarifies. Now let me move just for a few moments on to DMD. DMD, we had presented data on our 20 mg cohort back in September. We're not presenting new data. But we do want to update on the progress towards registration and the progress towards being able to launch. And we are moving very rapidly with that. We saw absolutely unprecedented dystrophin data with the 20 mg. I'll highlight it just briefly. You've seen it before. Nobody has had that level of dystrophin for Exon 51 on minimal treatment. But we also saw a functional benefit, small patient numbers, and nobody has seen that either. And that's what gives us confidence. And at the 20 mg per kg cohort, we have not seen any serious related treatment-emergent adverse events either. So the safety profile is building across the platform. So let me just touch on a couple of points. I think people are very familiar with this devastating disease. I know there are treatment options out there. I have not met anybody satisfied. The disease is not cured by gene therapy. I wish it was, but it is not. There is a massive unmet need for something that doesn't just talk about dystrophin, but starts talking about functional benefit improvement. And that's what we're trying to do for this patient population. And we want to get after it and we will make it happen soon. We've been through this type of trial. It's a similar type of trial design that we have to achieve, but we're a little further along. We are in the registrational expansion cohort, 32 patients. We have the same kind of patient numbers, similar patient numbers in -- from the multi-ascending dose cohort moving forward as well at the 20 mg per kg dose. Here's the safety profile. I commented on it, 837 doses administered to date, representing over 65 patient years of follow-up. I'm happy to answer questions on that. This is a refresher. This dystrophin data, whether you calculate it unadjusted or you calculate it adjusted is unprecedented dystrophin data for this patient population, full stop. Small patient numbers, but that's why we're continuing with the study at 20 mg per kg. You know what's more important, it's not a debate about who's got dystrophin and how much. It is, does the dystrophin in delivered by our platform -- does the distribution profile of that dystrophin broadly to the tissue because of our delivery platform enable that muscle to function. And what we saw here at just 6 months were functional measures that the FDA uses or the EMA users, where we had moved and seen clear functional improvement in just 6 months. You don't see that in DMD. Most are trying to get there at 12 months and it's clinically meaningful on stride velocity at 0.1, and we have exceeded that from baseline. We're not talking about slowing the progression of the disease. We're talking about improving from baseline. That's a big statement. It's only 6 patients, but that's what we're trying to show, and we're moving rapidly. So you can see the trial design. I'll give you a couple of dates on here, full enrollment this quarter. We will have that trial fully enrolled this quarter. We're very close to full enrollment. From there, it's a 6-month endpoint, primary endpoint, and you can do the math that early 2026, we could file, and we could have this drug on the market in 2027. And I think there's real opportunity. Finally, I'll just point to the pipeline. I'm not going to spend much time on the pipeline, but the platform to us has been proven now in humans with 2 different payloads in 2 different diseases that gives us the confidence to start preparing additional Exons candidates for DMD and build out a franchise. It also gives us the confidence to move forward with FSHD. Love to talk more about that, not today. And it also, because in Pompe disease, you need to get not just to the heart but to broad skeletal muscle and to the CNS. And if you've seen our preclinical data, we have clearance in the preclinical models of glycogen even in the brain. We want to move forward with a broad pipeline. So I'll just wrap up and then I'm going to be asking Doug to come up. You can see our time lines. We're not talking about building a great company in 10 years. We intend to build a great company soon. In 2027, we have a chance to launch 2 very meaningful programs that will be transformative for the field. And I think it will be transforming for our shareholders as well. Thank you very much. So I'm welcoming Doug Kerr. He is our Chief Medical Officer. You can see his background, neuroscience, neurologist, treated patients. I could go on.

Okay. Great. Thank you, John. Welcome, Doug. So I thought let's just get right into some specific questions on the data that you unveiled last week. What is the biological basis for why you would see a plateau in CASI with increasing dose?

You want me to take it?

You bet, I have talked enough.

I mean we've always said we've had preclinical models. We've had natural history. We always said that at some point, that we'll be diminishing returns to the skeletal muscle. And we don't think we've hit that. But what we look at when we go between 5.4 and 6.8 mg per kg is we do get more drug there. And there is more of an impact. When you actually look at the functional measures in the skeletal muscle, it is better at 6.8 mg per kg than it is at 5.4 mg per kg. So it's hugely different. But what really differentiates 5.4 and 6.8 mg per kg is what we've always hypothesized and now may be true, that we're actually accessing the harder tissues in myotonic dystrophy. And so we've always thought that the skeletal muscle largest organ in the body would soak up a lot of drug. That's on target sync. But above a certain level, you're going to access the CNS, the heart, other tissues that are relevant to myotonic dystrophy patients. That's what we think we're seeing. So we're seeing a much more broad deep benefit at 6.8 mg per kg relative to 5.4 mg per kg. If you look at the safety profile across those 2 doses, there's really no difference, no issues limiting our dose, that's why we decided on going forward with 6.8% mg per kg.

Why are you not dosing higher?

What we're finding is that we've got a transformational drug that affects virtually every aspect of myotonic dystrophy right here. And what we did is we did a very rigorous study. We've got all the pharmacology. We've got clinical benefit across strength, time function, CNS. It's time to go. We could dose fine forever. It's time to get this to patients.

I might just add one point. If our Chief Scientific Officer, Oxana was here, she would be saying, I had predicted that all you have to do is get to 20%. Not. It's hard to get to 20%. Nobody's gotten there before. But if you get to 20% CASI correction, as I was explaining in the -- with the biologic, you will unlock a broad functional benefit. We hit it at 5.4 mg per kg, and we hit it again at the 6.8 mg per kg. We don't need to keep going up to get the broad functional benefit.

Can you just talk a little bit about what data you're analyzing to get to that projected MCID that you had on that slide there? I can't remember which slide number it was.

Yes. So what we've done is we've shown several things in this MAD. John talked about it. First of all, analytic consistency of CASI, is it a good measure, intrapatient, interpatient loss of data? He showed that data. It is very good as a biomarker. That's really important point one. We don't lose samples. We don't have discordant data. It's like 4%. There really aren't many biomarkers that are that good. The second point, doesn't predict function? Absolutely. And if you think about, for example, neurofilament as a prediction of function in neurodegenerative diseases, it's pretty weak. The R value there is like 0.15 to 0.2. Well, you saw R values up there from John, the predictive capability of CASI for function is 0.4, 0.5, 0.6. It is a profoundly good biomarker. But the question test that you asked is, does it now predict meaningful data? So the way that we calculated MCID, remember there are lots of ways to do MCID. What most people do is a distribution-based statistical methodology to define what is meaningful to patients. That's what was done here. You can see in the footnotes, the distribution-based statistical methodology that did that. There's another way as well. And we started to do this, and that's an anchor-based method. They're largely the same. And when you look at that slide, it shows that the functional change that we've observed and achieved exceeds the MCID on all of those measures. Now there's more work to be done, lots more patients, more time. But if you wanted to know if the drug is having an impact that you think is meaningful to patients, that's the first evidence that it indeed is doing that.

As, just to -- we can talk to stats all day long on that. And then you look at MDHI and it's a 40-some percent improvement and you put that in front of the doctors or you show any of these data to doctors. There is no -- nobody says, "I'm not sure it's clinically meaningful."

Have you seen any of the safety blinded here since these last cuts that you reported last week? I mean, I think they're a little bit on the older side. Have you seen anything change materially?

No, we haven't seen anything. I mean that data table there, I think is December 6, it's updated. No related serious TEAEs. We've seen nothing that is dose limiting, no new events on safety at any dose within ACHIEVE or DELIVER for that matter.

And can you just please clarify what you mean when you say no persistent related anemia?

Only that after phlebotomy, you get a transient drop of hemoglobin that may drop slightly below 10, and then it pops right back up and is related to phlebotomy drawing a blood, but not related to the drug. We see nothing over time. No trends downward, no persistent anemia of any kind.