Great. Good afternoon, everybody. It's my pleasure to be moderating this panel with Doug and Oxana from Dyne. We'll get the CMO and the CSO perspective on where Dyne's at. I think in this panel, we're going to talk about each of the key programs and delve into some of the common regulatory questions that investors have been asking in the DM1 program. Maybe we can start, Doug or Oxana, whoever would like to take it away. Maybe you can just give us a couple-minute snapshot of your R&D efforts, how things are going in DM1, DMD, and then we can do Q&A. Thank you very much.
Yeah, I can start. Thanks, Paul. You know, Dyne's mission is really all about neuromuscular diseases, starting with Duchenne muscular dystrophy and myotonic dystrophy, and behind that, FSHD and Pompe disease. We are excited about all of those programs, all of which have different cargoes. The preclinical data for FSHD and Pompe disease is great. We have been able to present that at World Muscle and here at MDA in Dallas right now. We have two late-stage programs, obviously, DYNE-251 for DMD. We read out some data here at MDA, which is really encouraging for that.
That registration expansion cohort is fully enrolled and moving towards a readout in the fall. DYNE-101 for DM1 is also in late stage. We are really enthusiastic about it. We feel like these are positioned to deliver best-in-class therapies in the near future. We hope to be filing from these next year and launching in 2027 for both of those programs.
Awesome. All right. Doug, I hope you won't blame me for asking all the regulatory questions that I feel like I've been getting from investors, and I know that you have too. Maybe just to kind of start, right? Dyne laid out a path leveraging splicing as a surrogate endpoint for DM1 and a registrational cohort that, right or wrong, was, I think, smaller than some investors had expected because I think many of us just kind of look and index what different companies are doing comparatively to one another. You came in as CMO, I think, last summer. What gave you confidence in outlining this path without another FDA discussion since last May?
Oh, yeah, that's a great question. I did. I came into Dyne in July. I'm a neuromuscular neurologist, have taken care of DMD and DM1 patients for most of my career, and had followed Dyne since its inception back in 2017, but was really looking for something that was able to deliver payload.
Oxana, is that just Doug's video?
I believe so. We had this problem earlier. Just to maybe continue on what Doug started, DM1 program, as we know, molecular pathobiology is really a derangement of splicing occurring within the nucleus. Therefore, going after splicing correction and thinking of it as our biomarker for accelerated approval is something we absolutely want to do. It's, in fact, our obligation, we feel, to pursue accelerated approval that is available in the U.S. As Doug started to mention last year, we did have a discussion with FDA regarding our plan for accelerated approval, and particularly discussing CASI-22 measure of splicing correction. We had agreed on the plan that we would need to come up with a statistically significant change in CASI, improvement in CASI for our patients, and correlate this change with also functional outcomes, trajectory trends.
Yeah. Can I clarify that, Oxana? I think there is so much—I think there's a lot of focus on correlation, and in my perception, maybe some misunderstanding of what correlation means. When you say correlation, do you mean that the actual change in splicing has a high R-squared on a patient level with change in functional endpoints, or are you just saying that on an average basis, you improve splicing, and on an average basis, you also clearly improve functional endpoints? Do you know what I mean?
I know what you mean. The answer is, I would like to simplify this answer. If you think about our power assumptions for this accelerated approval, our expansion registration cohort, right, we would be looking into statistically significant change in splicing at month three. That is with power over 90%. That's how we powered this study. We will see it. As far as functional observation, there's a directional change. There is no statistically significant change in function that is required, but we need to observe effect on function, not at that time, but later, right? This could include either measures of strength, timed functional tests, or PRO.
Yeah. Again, so it's not about—because you guys say correlate, but I feel like, am I interpreting correlate as really more like there's an association, like function's improving too, not like you don't mean like a quantitative correlation on a patient?
No, because what we are saying is the ability of biomarker to reasonably likely predict a change in function, but later in time.
Yeah. OK. I guess the other question—and we can just keep going here until Doug joins—why is this the best path? I guess it's funny. If I look at your guys' phase one two data, I mean, look, the drug clearly improves splicing. You could make an argument that the VHOT data is more consistent, right? I mean, if anything, VHOT separates for all doses at all time points. Splicing mostly does, but you have the outlier issue. There's a high degree of variability. Sometimes placebo gets worse. Maybe in the next batch of patients, placebo might get a little bit better. It almost feels like splicing has a greater signal-to-noise ratio than some of the other measures you've shown efficacy on. Again, why did you guys double down on this way of doing it?
OK. Let's start with DM1 disease molecular pathobiology. It is a disease of splicing derangement at the molecular level. We are thinking about effective therapeutic. What it needs to do, it needs to deliver to the nucleus, correct the defective splicing. Therefore, you start with selecting ASO to ACHIEVE delivery to the nucleus and correct splicing. If you think about DMD as a devastating disease with a multitude of manifestations, we know it affects skeletal, cardiac, smooth muscle, and CNS. All of those manifestations are caused by missplicing of multiple genes and can be measured by this CASI-22 gene panel.
How do we expect to correct all these functional defects where myotonia is only one of them? In fact, with underlying a single gene responsible for myotonia change. Myotonia effect is not going to predict on other functional tests or PROs. It's just one aspect. What can predict what happens to all these functional measures is correction of splicing. That's the only thing because that's what causes it. What I started with is we have selected ASO. ASO gets there. We have shown as we increase the dose in our cohort in MAD portion of ACHIEVE, we saw correction of splicing from the initial dose, 1.8, just showing some patients responding, some not, as expected.
It's a very low dose. As we go to 3.4, we began to see effect through the patient cohorts continue to 5.4 and 6.8. You mentioned it's not a good measure. It is, in fact, a good measure. We only had a single time point that fell off the chart. In the small patient.
I didn't mean to say it wasn't a good measure. I just meant to say that there is noise, right?
Oh, yes.
Yeah.
There is noise in every measure. There is noise in measuring numerous serum biomarkers. Consider CK in Duchenne is extremely noisy and means a multitude of things. In this case, it's a fundamental pathobiology. That is why it was a productive discussion with FDA circled around this as a right surrogate marker that has the potential to predict functional outcomes by the biology, by the pathobiology of disease.
Yeah. Yeah. OK. Go ahead.
I just wanted to say, since we are talking about accelerated approval pathway, because of our platform, because we have this measure available to us, this pathway is open. That's why we are pursuing it. Let's say we would go straight to phase three. What is the loss? There is no loss. Phase three is available. It's in the plans. We just have this unique ability to look at our registration cohort and submit based on for accelerated approval. That's all.
Yeah. Yeah. I think, well, I think the one thing that someone might say is, well, the loss might be that is there an in-between? Could you have set up a registrational cohort that's 100 patients and powered for all the functional endpoints? Maybe that delays you six to nine months, but now you've really diluted the risk to that not playing out the right way for you. What would you say to that?
At this point, I'd pass it to Doug.
You can hear me, right?
Yes. Thank you.
Oh, good. OK. Yeah. I mean, there are lots of ways you could do this. For us, really highly powered on CASI, on myotonia, we got this roadmap laid out for us for accelerated approval by the FDA. We are executing on that roadmap. We know what they want. They told us.
Right.
You do not have to have statistical significance on the functional endpoints. We will be close. Forty-eight is a pretty big number. The rest of it is going to come from the phase three. Obviously, there is cardiac, there is smooth muscle, there is CNS, there is a lot of that. This is kind of the balance that we tried to strike, get in as quickly as possible, be responsive to the FDA, but just do not delay the phase three. Let's get the whole thing done as well.
Yeah. Can you talk about the powering scenarios for splicing, Doug? Assuming you end up enrolling up to 48, which it sounds like you may do, how resistant are you to some of the noise issues that can happen with this biomarker? Let's say placebo doesn't get worse, but placebo has no change or a small improvement. How can we kind of stress test those assumptions?
Yeah. I mean, first of all, I'm happy to go through kind of the numbers. I get my statistics group, and I've done this with a bunch of people, right? They kind of walk away saying, we get it. I'm happy to do that. Let me just say, the way we've statistically powered this registration expansion cohort is based on very, very conservative numbers. The change in CASI that we modeled is less than that which we observed. Although the placebo worsened in ACHIEVE, we have modeled that it does not worsen.
OK.
One final aspect is you talked about the CASI. What if we have a QC failure or things like that? We went back and looked. Out of the 150 samples that we've looked at, 96% of them are exactly what you want. They do not fail QC, and two adjacent biopsies tell you exactly the same thing. We've built in kind of new statistical and analytical methodology. Even so, we've powered the registration cohort to account for 10% missing data. You put all of that together, and you walk away saying, my God, those were very conservative assumptions, less than that which has been observed. We're fully powered 90%-95% for CASI, for myotonia, and for very strong directionality on muscle. We feel like that's an option that we've got to take.
Yeah. Yeah. OK. You guys are going to be engaging again with the FDA. You guys had talked about this as relating to your confirmatory study, but it feels like it's inevitable that you would also have further discussion on this cohort. I guess maybe I'll leave it to you. What could this dialogue look like? What questions might you propose to the FDA? Do you think you'll be able to come away from this meeting with reaffirmation for us that this plan is intact and more formally blessed?
I mean, obviously, there's a lot of urgency on our side, getting all the clarity we need. Really, we feel like we've got the clarity on the registration expansion cohort because they laid it out. Sure, we will talk to them. The main purpose of that meeting is to get feedback on the phase three design, statistical methodology, size, et cetera. There is lots that we will get in terms of clarity from them on the phase three. In terms of kind of when we would externalize all of that, we're going to incorporate this feedback from regulators.
At some point, we're going to provide this fulsome update on the accelerated approval pathway on the phase three trial and on that path towards full approval. We're not saying when that's going to be, but we anticipate very good regulatory engagements. We will talk about it when we have an update.
Yeah. OK. If you're an investor and you're looking out for this FDA meeting, what's the worst-case scenario? Is it that you just have to enroll more patients in the registrational cohort? It sounds like you're very confident in splicing. There's no reason to think that that would have changed. What's the risk?
Hypotheticals. But could they say, well, we want a few more patients? OK.
Yeah.
What if you want some more of a safety database? Remember, you can add more patients, but you've also got the phase three that will be started this year. You can easily cut data and bring it over to enhance the robustness of the safety database. There are some things that obviously they could say that would impact that. I don't know what else they could say. We'll have a dialogue. Again, it's mainly about the phase three.
Right. It's mainly about the phase three. I guess in that, again, I'm not trying to split hairs here, but ultimately, from your perspective, in that meeting, do you feel like you would know leaving that meeting if the FDA was uncomfortable with your filing plan for accelerated approval?
It's possible. Regulators speak. It rarely is as definitive as we would like. To be fair to them, it is going to be a matter of reviewing depending on the data and how good is the efficacy and what about the safety? Do they feel like that has kind of been able to establish a full benefit risk?
Right.
We have said this before. The roadmap was laid out. The exact specifics of that registration expansion cohort, we did not go back to them. We did not have time. We do not know the exact specifics. We are thinking about how we could be responsive to the FDA and still have an accelerated approval pathway. The backstop here is phase three, up and running, full approval. If it is 12-18 months or 12 months behind, it is a big mark. It is a good drug. It is going to be commercially and kind of very transformative for patients. Even in that case, worst-case scenario, still pretty exciting for us.
No, I get it. Worst-case scenario, it totally feels like you have a drug here. I'm just trying to kind of get a better understanding if right or wrong, there's some skepticism in this in the investor community. Are we going to get the feedback that you feel like people want to get at some point this year? Because it sounds like on the one hand, you might talk about this accelerated path. On the other hand, it's like, why re-ask the question if you already got the feedback on splicing? Some of these things are review issues. I guess maybe ask another way. I'll try one more time. Then we can move on to another topic.
Come the end of this year, do you think we're still going to be asking the same questions on the regulatory side and still going to be perceived as a little bit gray? Or do you feel like we should be able to have materially more information from this FDA meeting?
My sense is you'll have more information. How definitive, I don't know. It won't be where we are now. I recognize that there's kind of a need to really clarify that regulatory overhang. We're doing as much as we can. I think you'll get some. I don't know.
Yeah. OK. All right. Thank you, Doug, for entertaining 10 questions.
Sure.
I appreciate it. Can I just clarify one thing? Did you say you were 90% powered for VHOT at 48 patients? Or was that a misspeak?
That was a good speech.
That was a good speech . You're 90%-95% powered on splicing and powered on VHOT and powered on quality of life too? Is that right?
Not fully powered on the functional measures or strength measures or PROs. It's been tough to model the PROs because especially the MD HI seemed to only move in the last cohort. It's a little bit hard to model that. Our view is we do not have full power. When I say full power, I mean north of 80% or 90% for those measures. We do not.
Right. Right. OK. Fair enough. All right. Let's talk about DMD a bit. You guys recently had some super interesting data over the weekend where you showed evidence of long-term benefit. It's a small sample, though. I guess as you take a step back with the DMD program, and we know a number of exon skippers haven't worked in confirmatory trials, what gives you the confidence that this time is different? That this time we're above that dystrophin threshold and the clinical data we're seeing in these small ends is real and reproducible?
Yeah. I'll be interested to have Oxana jump in here. I mean, as we've seen the data play out over time, initially, it was fairly short-term data in the clinic, 10 mg per kg. Then we saw 20. Now we've got data that is out to 18 months on the 10 mg per kg, 12 months on the 20 mg per kg. What you're seeing is really robust improvement from baseline, not just a slowing of the decline. People haven't seen that before. It is maintained for as long as we've looked, including in the 10 mg per kg out to 18 months. It's on SV95C. It's on North Star. It's on Time to Rise. It's on 10-meter walk/run. Multiple endpoints, multiple time points, multiple cohorts.
What's fascinating is that when you look at the two cohorts, the 10 mg per kg and the 20 mg per kg, they're very different. The 10 mg per kg, younger, higher functioning, you still may be getting improvement at least at the very beginning of that just by virtue of developmental motor milestones before DMD starts to pull these boys down. That is not true in the 20 mg per kg. They are older, less functional, much more fat tissue, less muscle fiber. Yet, they respond from baseline on SV95C and all the other measures as well. It felt to us like this is finally telling us something. These patients broadly in DMD, not just the young patients, but even the kind of older patients, they can respond. They can improve from baseline multiple measures. We're excited about it.
Yeah. I can add, thank you, Doug. Really, don't go far. Just start with the platform. If you look at correlation and prediction of our preclinical package, how we designed platform, what we saw in preclinical species, we are super happy to see aspects of it clearly showing up in our trials, one of which is, for example, deep tissue penetration with our platform and delivery to every muscle fiber and uniformity of that delivery. If you recall, in our section in the patient sample of even at 5 mg per kg, we saw a very uniform widespread expression of dystrophin at sarcolemma.
If you're looking at the functional change even at six months, that is very exciting for us, seeing continuity of it to 12, 18 months, even more foundation for its distribution, localization, and this correction of the whole muscle type because of the way platform works. If you think about patchy distribution in case of naked PMO, for example, you cannot expect to see this muscle type acting in concert. In our case, we can. Even if you think of how profound is 3%, over 3%, and 8% adjusted dystrophin, is that the right number? If you think about it, the way it distributes, it appears to be the right number. It appears to be sufficient to allow us to see even functional benefit.
Yeah. Yeah. How do you think about demonstrating functional benefit in the registration cohort?
We've got, obviously, the primary endpoint, dystrophin, accelerated approval pathway in the U.S. fully based on dystrophin. We confirmed that back in December. We don't need function. Obviously, we're excited about function. SV95C, really, really good objective accelerometer-based endpoint. EMA has done beautiful work on this. They have qualified it as a primary endpoint and the MCID of 0.1 meters per second. Both of the 10 and 20 mg per kg cohorts are well and above that MCID, double it actually. That's a very important endpoint.
Obviously, this is a bit of a newer endpoint. It's got many advantages because it's done in the kid's home. It's not kind of effort-dependent. He wears the accelerometer for a couple of weeks prior to the clinic visit. Obviously, we're still going to get the standard 10-meter walk/run, Time to Rise, and North Star to contextualize that. We're looking to look at all of those in the registration expansion cohort.
Yeah. Yeah. OK. Great. Anything else you'd like to add, Doug or Oxana, before we wrap things up?
I just wanted to say that since we are presenting at DMD, our DMD programs, very exciting. I think what we can project is how applicable our platform is for expansion of the pipeline. Also, we did not touch today on CNS delivery of our payloads. That really sets apart our platform. We have, speaking of DM1, seen in preclinical models delivery to the CNS. We see a reduction of toxic foci formation not only in the heart, for example, but also in CNS in our mouse model. If you look at NHPs, we've seen widespread throughout the brain distribution, including deep brain, very uniform, very much like in muscle.
I also wanted to say that for the first time at MDA, we are sharing our CNS effects in DMD mouse model, actually a functional effect on anxiety. Single dose is capable of relieving anxiety in these mice. It just tells you that on top of it, Pompe program and also its effect in CNS and clearance of accumulation of substrate from lysosome tells you and sets you so well for our expectation of benefit, CNS benefit to our neuromuscular diseases. Just initial data in our MDHI on fatigue and sleepiness also is and support from all the other type of payloads that also deliver. Really sets up us well for fixing cardiac, skeletal, smooth as well as the CNS symptoms.
One more thing from me just to say that back to DM1 and ACHIEVE, we're looking at a variety of endpoints. Additional details on that trial can be found on clinicaltrials.gov. So anybody who wants more detail on kind of what those endpoints are, that's now available.
OK. OK. Great. Thank you both very much. Appreciate it. Always a good discussion. Best of luck.
Thank you.
Thank you.
Thanks for your time. Appreciate it.
All right. Thank you.