Good afternoon, and thank you for joining Guggenheim's Third Healthcare Innovation conference. I am Debjit, one of the Therapeutic Analysts, and my privilege to host Michael Severino, the CEO of Sarepta Therapeutics. Thank you so much for making time for us, Michael.
It's a pleasure. Happy to be here.
Let's start off on a high level. I think you probably accepted one of the most challenging positions in biotech. What made you decide to come to Sarepta at this juncture?
Well, I see the setup at Sarepta as unique in a number of ways. One, it's very rare, almost unheard of for biotech at our scale, to be in the position that we're in, which is that we have a commercial business with a leadership position in an important condition like Duchenne that generates substantial revenue, but where we can use that revenue to drive a pipeline with the value that I believe exists in our pipeline, and do it on the basis of our own cash flows, do it without the need to raise equity. When I look at the commercial business, I think it has very real potential. Obviously, 2025 was a challenging year for us, but I believe we've done a lot to address the concerns that arose in 2025.
We've focused the indicated population to the ambulatory patients for ELEVIDYS, where benefit-risk, I think, is quite clear. We've introduced new data, long-term data, functional data, imaging data that I think support the presence of substantial benefit there. We're managing risk effectively, and I think that will be a durable business. Our PMO business is very robust. I think we can use that experience, use that cash flow, not only to deliver a lot of good to Duchenne patients, but to advance this really promising siRNA pipeline as well. It was that combination that led me to join over the course of the summer.
Got it. You just reported your third quarter. The exon skipping business looks really strong and sticky. Could you walk us through the assumptions? We have got competitive launches, at least for one of your exons, hopefully next year. Your business seems to be very sticky, with over 90% of the patients being home dosed. How do you see that evolve once you have competitive launches?
I'll first start with the attributes of our product. All of our exon skipping therapies, but particularly exon 51, given that is what is likely to face competition in the coming year with the Dyne program under FDA review. That program, our product has been on the market for more than a decade now. It has a very well-understood safety profile and a very favorable overall safety profile. We also have extensive long-term evidence, so real-world evidence supporting its use, prolonged time to ambulation, reduced decline over time in things like pulmonary function and cardiac function and overall mortality, things that really matter to patients, to their families, to physicians. That whole body of evidence, in our mind, is what creates a lot of the stickiness of these therapies. Above and beyond that, we have tremendous experience navigating this commercial marketplace. We understand the reimbursement pathways.
Our access groups are very effective at ensuring coverage for patients who are qualified. We have robust patient support programs in place, at-home dosing, which is critically important for these patients. On the order of about 90% of patients participate in the in-home dosing program, which is a tremendous feature in terms of ensuring the high levels of adherence that we've been able to demonstrate. All of those things are what lead to the stickiness of these programs, and all of those things I think will support us well when competitive entrants come into the marketplace.
Given that 90% of these patients are home dosed with a caregiver, how often do they see a physician before they can contemplate switching to a different therapy?
Well, that's going to depend on the individual physician practice in their offices in terms of how frequently they see these patients, but it's typically going to be on a frequency of once a quarter or longer. But again, we view the visit to the physician not as the primary determinant of whether a patient stays on therapy or switches to a new therapy. It'll be what is their experience with that therapy? Have they been stable over a long period of time? Are they comfortable with the safety profile that they've observed and the benefit that they've derived? All these things are going to be part of the discussion between the patients, their families, and the physicians when they talk about what potentials exists in the marketplace.
Got it. Now, if you think about the exon 51 market, you have the new incident patients, you have patients who are probably already on ELEVIDYS, sorry, on EXONDYS, and the older non-ambulant patients who may or may not be on EXONDYS. Where do you think you're likely to gain or lose market share if Dyne is approved next year?
Well, again, I think that the stable patient base is going to be very sticky. I think that our product will do very well in the face of competition with those stable patients who have been on the therapy for, many of them, for an extended period of time, have experienced the benefit, and are very comfortable with the safety profile that they've observed. I think that's going to be a core part of our base. We'll compete for new patients based on the attributes of the products. For our product, the important features are going to be the ones I already mentioned. The time in the market, the well-understood safety profile, the data to support the benefit-risk. That'll be our value proposition.
We'll have to see what the label looks like for Dyne, when their PDUFA date comes, in terms of what they'll be able to say. But we feel good about what we have to offer to this patient population.
Got it. Moving to gene therapy and ELEVIDYS. Third quarter would, as the way, and if I understood correctly, this is kind of the new baseline in terms of patient starts. Do you expect to see this move higher as physicians now have a better understanding on how to manage these patients, especially post-therapy?
As we described on both the most recent quarterly call and the second quarter call before that, we feel like what we're observing now is a clean baseline. Given the six-month journey from start form to dosing, the demand that we're seeing now is the demand that was generated much earlier in the year, before our commercial sales force expansion was in place, before that sales force had the new data at their disposal, the three-year data from EMBARK, some of the imaging data in 8- 12-year-old patients, which are critical in understanding benefit risk as patients progress in their journey with Duchenne. Now that's all in place, the data are in the hands of the field force, they're out having these discussions with physicians. Those conversations are going well, and from this point forward, we'll be able to assess the effectiveness of those conversations.
We believe in the long-term potential for growth for ELEVIDYS, but we want to see some longer-term trends before we predict the exact timing or the nature of that. Of course, there are data readouts that are coming, data readouts from Cohort 8, that could also impact that trajectory, so we want to see how that plays out before we give longer-term color.
Got it. What do you think is the regulatory hurdle for any other gene therapy entrants for DMD? I mean, ELEVIDYS had to go through two separate studies before it got approved. Do you think FDA is going to kind of ask the same from your potential competitors?
Well, certainly the standard up to this point has been for randomized controlled data showing a clinical benefit, and that's the standard that we've been held to. Once you have that in a particular population, at that point, it is possible to look at the correlation between microdystrophin expression and benefit and extrapolate. But prior to that point, that's not something that the agency has been willing to do, at least up to this point, and for good reason. We've seen other programs that have shown microdystrophin expression that didn't translate into clinical benefit long term, so we know that all these products are not equal with respect to their ability to deliver that benefit to patients. So we believe that randomized controlled data is an important part of the data package that should come prior to approval.
Obviously, there are some of our competitors out there who are planning recent or in the near- future regulatory interactions. We'll see how those go. But if the standard is consistent, it should take randomized controlled data, and that's what we believe an appropriate standard is, given the uncertainty between microdystrophin expression for different programs and the ability to translate that into clinical benefit. Also the need to have a robust safety database before understanding what benefit-risk really looks like in these populations. We've dosed more than 1,200 patients in the ambulatory population, more than 1,500 overall. We have a very well-understood safety exposure database. It's very hard to compare that safety exposure database to something with a markedly smaller exposure history from single-arm studies.
Got it. Let's switch over to the pipeline because that's where all the attention is. You're at the cusp of providing updates on your DM1 program and the FSHD program. The DM1 data comes first. Could you talk to the way you're trying to deliver the drug product to DM1 patients? Because there are various approaches there, and Sarepta has a slightly different take on the approach on the delivery side.
The delivery vehicle—
Yes.
Specifically, yeah. We believe delivery is obviously an essential part of the equation that goes into what benefit you ultimately can provide to patients. And we have a delivery that we think is unique in the space and provides real advantages that we've seen not only in preclinical studies, but in our single ascending dose studies. We use an αvβ6 peptide ligand to deliver. We know that αvβ6 has very good surface expression on muscle cells. We know that it rapidly internalizes and it recycles, so you don't get depletion of surface availability of αvβ6. We know that from preclinical studies.
We've shown that that translates into very high muscle concentrations, preclinically robust knockdown, preclinically, and that has translated through to our single ascending dose studies, where we also see very robust muscle concentrations in both our DM1 and FSHD programs without evidence of dose-limiting toxicity. So we know we can get the agent there. We also know when we deliver agent, it's very, very potent, and we can show that through a number of different assays. And so we think that, that combination sets us up well to be in a position to bring forward best-in-category offerings in both our DM1 and FSHD programs.
Specifically with DM1, with the upcoming data, where should we focus on? Is it going to be splicing? Is it going to be some of the functional measures like vHOT or any other specific attributes that you think the street should focus on when you look at the upcoming Sarepta data versus what's already known in the field?
I would focus on DMPK knockdown and splicing correction, because that's what the phase I-B study is designed around. We know that DMPK mutations are causal in this condition. So DMPK mutations in the nucleus sequester MBNL, and that leads to abnormal splicing patterns that lead directly to the disease. That's the causal chain between the fundamental underlying mutation and the disease that we observe clinically. That's what the phase I-B study is designed to interrogate. We'll assess functional measures. But functional measures take more substantial studies and longer periods of time to show meaningful impact. We'll look at vHOT, we'll look at other functional measures. We'll show the data that we think are relevant, and we'll show the vHOT data.
But I would caution folks not to over-index on changes in small populations in vHOT, particularly when one is trying to compare amongst different programs. The studies just aren't designed and aren't powered to do that. But they are very well-designed and very well-powered to show that causal biology, that we're knocking down the toxic species, that that corrects splicing, and that is what we believe will predict, in larger and longer-term studies, the best ability to deliver a clinical benefit.
Got it. With the DMPK and the splicing correction, what level of DMPK knockdown and splicing correction would make this compelling and make you commit to a registration study?
We know what others have shown. We know that, for example, the Avidity Novartis program has shown between about 10% and 15%. We also know that at least at a top-line level in their phase III study, they weren't able to show a difference in vHOT, which was the primary endpoint in that study. They have commented publicly that they've seen promising signals in secondary endpoints. That sort of tells me that perhaps that level is on the cusp. We'd want to be substantially north of that. Others may have delivered into that 20%, sort of 25% range. I think that is an area that starts to become interesting. Our goal would be to come in even above that. But we'll see what we're able to achieve when the data come in just a little bit later this year.
Got it. Is there a sense from your thought leaders, et cetera, what level of these knockdowns would translate into a clinically meaningful benefit which can be captured in, say, a six-month or a 12-month study?
Yeah. I don't think that there is yet a firmly established minimum threshold. As I said, I think in that 10%-15% to 20% range, you're getting into the realm where it's quite possible in longer-term studies. I think as you get further north of that, I think one can have greater confidence in the ability to show a benefit in longer-term clinical studies. We're going to aim to go as high as we can go with respect to splicing correction. We have a program that has shown the ability to increase muscle concentrations as we increase systemic dose without any dose-limiting toxicity, so we can push on this pathway quite hard. We're going up to very meaningful doses in the program. We're going up to 8 mg per kilogram in the DMPK program expressed as siRNA.
Our goal will be to get the most robust splicing correction that we can.
Got it. If you get to the thresholds that you think you can, do you think there is a biomarker-based approval feasible for DM1, or it has to be a functional endpoint, and we yet don't quite know what a functional endpoint should be for DM1?
I think accelerated approval, based on everything we understand about the space, should absolutely be on the table. I think what that accelerated approval is based on still has some open questions. I think it's potentially still vHOT. I don't think we know whether the Novartis program has fully explored the potential of vHOT as an endpoint, so it could be vHOT. If it's not vHOT, I think a biomarker-based approach is also very much on the table. Splicing correction, as we said, is on the causal path. The spliceopathy is what leads to the clinical syndrome. Showing a robust impact on splicing with a validated assay, I think, provides an alternative approach to vHOT. So I think either is quite viable.
Do you think the Novartis approach was too limited? Because the original plan was to go higher. Obviously, they had a few SAEs, which kind of stopped them at about two to four. Whereas, as of yet, it looks like you guys have a much cleaner profile.
Well, in my opinion, it's limited by the ability of knockdown and correction they were able to achieve, whether that's truly based on systemic dose, because they also seem to have a ceiling that they could achieve in terms of actual muscle delivery based on systemic dose. That may be due to the delivery mechanism that they have, so a bispecific transferrin receptor antibody. They were limited by anemia in terms of how high they could go. There's also the possibility that that bispecific antibody leads to clustering and receptor downregulation and can limit the amount delivered despite escalating doses. Those are theoretical concerns. I don't know fully what explains the biology that they've observed, but they did seem to be limited in terms of the extent of knockdown and splicing correction that they should be able to achieve.
We know that those same limitations, based on all the data we have to date, don't apply to us. We can dose escalate very effectively. As we dose escalate, we get more into muscle. Our reagent is very potent at knocking down DMPK, not only in the cytoplasm, but as we showed at the Oligonucleotide Therapeutics Society meeting just about a month ago, we get nuclear knockdown as well. We believe we're in a very good position to test this pathway and to deliver benefit to patients.
Got it. Before we talk about FSHD, I think I want to spend some time on Huntington's because I do not think people talk about it as much, and this is a very differentiated program, especially given it is a subcu dosing option. There was a slight pushout with the data. Now you expect somewhere middle of next year, unless I am getting my timelines right.
Yeah. Somewhere around the middle of next year is when we would expect proof of biologic activity. There is some uncertainty as to how far up the dose curve we need to get, just because we are not there yet, before we believe we have enough data to show a convincing proof of biologic activity. That is what drives our statement around the middle of the year. The program is progressing well. We are confident in the ability to get to that proof of biologic activity. It is just a question of which cohort will it be that we decide is most illustrative, most informative to show more broadly. So that is the change in the timeline, or that is the driver for the timeline.
Given the recent data from Alnylam, where do you think this program has to be with the advantages that you have with dosing?
Well, we think that we can show very robust knockdown. An important feature when one thinks about our data compared to the data from other programs in this space is that we are delivered systemically, and we are acting in the CNS. That is not just a matter of convenience. Yes, subcu delivery is more convenient than intrathecal delivery. What we are looking at here is Huntington protein knockdown in the CSF. When you deliver a therapeutic into the CSF and then measure the CSF, you do not know whether you are actually impacting the entire parenchyma, including the deep brain nuclei where you need to go, or you are just operating at that interface between the CSF and the parenchyma. We are delivering systemically.
For us to get into the parenchyma and for that effect to show up in the CSF, it essentially proves the principle that we're crossing the blood-brain barrier. We know from preclinical models, and from first principles, when you do that, you should be getting into those deep brain nuclei. We think our knockdown data can be interpreted very differently, or should be interpreted very differently, than an intrathecally delivered program. Showing robust knockdown in the CSF for us, we believe, is very strong proof of principle, and has a high likelihood of translating into downstream clinical benefit.
Given the advantages, and let's assume that come May 1, uniQure does have an accelerated approval, do you think the same path is open for something which is that as differentiated as your program, or do you have to conduct a randomized study?
I think for a condition like Huntington's with a mechanism that's novel like ours and the features that I just described, the accelerated approval pathway should certainly be on the table if we can demonstrate the kind of data that we would expect to. Patients with Huntington's have very limited options today, and I don't think there's any one therapy that's going to fully address that such that the accelerated approval pathway wouldn't be open.
How many patients do you think you can get on the higher end of the dose curve so you can sort of convincingly make the case that this is doing exactly what you outlined?
Well, we'll go up the dose curve, and we have the ability to expand cohorts. I think it's a little bit early to predict exact numbers until the data starts to come in. But I think within a reasonably sized I-B expansion, one can make a compelling argument that we're impacting the biology in a way that's likely to translate into clinical benefit.
Got it. Let's go to the FSHD program because that's an imminent readout. The bar is sort of out there from competitor data. Where do you need to go or where do you need to be to aim for an accelerated approval using the DUX4?
Well, others have achieved roughly 50% knockdown of DUX4-regulated genes in biopsy material. Looking at circulating biomarkers, the data had been a little bit less clear, probably lower than that 50%, maybe more in the 20%- 30% range. Again, we would aim to be substantially better than that. In our single ascending dose study, in pooled cohorts, but all of the doses lower than the doses that will come in our update later this year, we showed roughly 90% inhibition of DUX4-regulated genes.
We're really setting a very significant bar there. In the multiple- dose study, we're looking for something that is consistent with that number broadly. It doesn't have to be exactly 90%. I think anything greater than 70% would be an exceptional result and likely to predict long-term clinical benefit. We'll also have the ability in that study to look at circulating biomarkers.
We'll look at the same circulating biomarker that others have looked at, which has been referred to as cDUX, but we'll have additional, at least one additional circulating biomarker that we can look at. I think we'll have a good body of evidence moving forward to show that we're impacting that pathway in a way that should translate into real clinical benefit.
Unlike, say, DM1 or HD, this one seems like an indication where FDA would like to see a randomized study and not just a biomarker-based approval. Is that your expectation?
I think the potential for accelerated approval is still on the table. Obviously, one would need to do a confirmatory study that would be randomized and would look at longer-term clinical endpoints. If we generate the kind of data that we think we can generate, we believe accelerated approval is on the table. Of course, we'll have to take the data that we generate when we have them and discuss that with the agency, but our view would be that accelerated approval should be possible.
Awesome. We've got like 40 seconds left. When you think about 12 months from now, where do you think Sarepta is?
Well, I think we'll be a very different sort of company in a very favorable way to where we are today. If you think about where we could be 12 months from now, I would expect that we'd have both DM1 and FSHD in registrational studies, with the potential for accelerated approval, so approval not too far out on the horizon. We would expect to have proof of biology of our Huntington's program. If you think about that's three high-value programs that I think aren't reflected in our valuation today. That can be really transformative for the company. That just sort of fuels everything that's after that. There are many other programs pre-clinically that we could pursue, and many other venues that we could use to continue to expand our presence in neuromuscular and neurologic conditions.
I think 2027 is going to be a very transformational year for the company.
Wishing you the very best, Michael.
All right. Thank you.
Thank you.
It's been a pleasure.
Thank you.