Sarepta Therapeutics, Inc. (SRPT) Earnings Call Transcript & Summary

Good day, and welcome to the Sarepta Therapeutics conference call and webcast for the SRP-9001 Micro-dystrophin R&D day. [Operator Instructions] As a reminder, today's program is being recorded. Now I'd like to turn the call over to Doug Ingram, President and CEO, for opening remarks.

Thank you, Catherine, and good morning, everyone, and thank you for joining us today for what we are calling throughout as Micro-dystrophin Day. First, remember, we will make a number of forward-looking statements today. Please refer to our various public filings for a list and discussion of the risks and uncertainties that come with making predictions about what may occur in the future. Now in a moment, Dr. Louise Rodino-Klapac will walk us through recent updates from our 3 ongoing studies, the Study 101, Study 102 and Study 103. And then we'll provide information on the protocol for our pivotal trial, Study 301, a study that we call EMBARK. SRP-9001 represents the most advanced and significant opportunity in gene therapy today. The goal of this meeting is to share additional data from what is indisputably the most comprehensive evidence set for any Duchenne gene therapy to convey to you the basis for our deep conviction in the transformative potential of SRP-9001 for those living with Duchenne muscular dystrophy and to detail the thoughtful design of EMBARK, our Phase III trial, which we recently announced has already initiated. We are indeed proud of how far we have come with SRP-9001. With 77 patients dosed thus far, we have the largest database to inform us on the performance of SRP-9001 and to inform our program and enhance its probability of success. We have become leaders in the manufacture and release of gene therapy. We are the first Duchenne gene therapy sponsor to have been greenlighted in the U.S. to commence a genuinely global pivotal trial with sites eventually in the United States, Europe and Asia. And with the wealth of patient data so far, some of it stretching beyond 3 years, it is beyond genuine debate that, as of now, we have a highly differentiated gene therapy construct in terms of safety, in terms of tropism, in terms of expression and in terms of functional signals. And yet we advance forward with the humility of a patient-driven team that never forgets that the patients we serve and for whom we exist are in a daily fight against a brutally degenerative disease, where time is of the essence. So while SRP-9001 may stand alone today, nevertheless, we know that we are indeed in a race. We are in a race against the disease that daily robs children of their muscle and their function and, invariably, their lives. Hopefully, by the end of our discussion, you will agree that we have kept that race in mind and have built a program with the greatest probability of rapidly advancing to the community a potentially life-altering therapy. And with that, let me turn the presentation over to Dr. Louise Rodino-Klapac. Louise?

Thank you, Doug, and good morning, everyone. It's a pleasure to be here today to really describe the breadth and depth of our data with SRP-9001. And just to quickly review the agenda for today, as Doug mentioned, we'll be reviewing data from our 3 trials, 101, 102 and 103. In terms of new analysis for the 101 study, we'll be showing you 3-year functional data in comparison to natural history. And then we'll also be sharing additional analysis for Part 2, where we've also compared to natural history as well. And finally, in our 103 study, in our commercial process material, we'll be sharing functional results from the Cohort 1 first 11 patients. And this is all leading up to the study design for our 301 EMBARK study. So if we go to the next slide. This slide really needs no introduction. As Doug mentioned, DMD is a devastating disease and children need urgent treatment. SRP-9001 was designed specifically to target all aspects of DMD, which includes skeletal muscle weakness, cardiac manifestations and respiratory manifestations. And it is truly a race to be able to get to these patients. And on the next slide, just to review the 3 components of gene therapy that are essential, really, and that just includes the vector, the promoter and the transgene. The vector, which really derives how you get to your target, in this case, skeletal muscle and cardiac tissue, and then really defines the safety or immunogenicity profile of your construct. Next, the promoter, which turns it on and off in the tissues that you like or a desire and then also the transgene. So the next slide just to talk specifically about the components of SRP-9001. As you know, we have selected the AAVrh74 vector. We selected it specifically for several reasons: One, in comparison to all other vectors, we saw the transaction profile, specifically in muscle, which had a broad distribution across all the muscle cells, also in cardiac tissue. It has also has a low level of pre-existing immunity. We see approximately 15% of patients with antibodies to rh74. Next is the promoter. We are using the MHCK7 promoter, which expresses nicely in skeletal muscle and cardiac tissue, which is critically important to any therapy for Duchenne. And then lastly, our transgene, which I'll talk about in a few slides, which was designed specifically to maintain all of the critical components of the disturbed gene and protein. In the next slide, just to highlight rh74. As I mentioned, it has broad uniform distribution. And we currently have it in gene transfer trials for both Duchenne and limb-girdle muscular dystrophy and remains to have a differentiated safety profile. We've now dosed over 80 patients in both DMD and LGMD trials combined. Also specifically, just the MHCK7 promoter to specifically target cardiac muscle as well as high levels of skeletal muscle transaction. Next slide. Now micro-dystrophin is a short and functional form of dystrophin. If you look at the left and normal muscle, dystrophin is a shock absorber that has critical link inside the cell to outside the cell. So when the muscle contracts, dystrophin absorbs that shock. It's too large, as you know, to fit into adeno-associated virus, or AAV. And about 30 years ago, a patient was identified by Dr. Kay Davies, who is a 61-year-old ambulatory Becker muscular dystrophy patient. Becker muscular dystrophy is a form of dystrophinopathy in which patients have shortened version of dystrophin. So it's a much milder course of disease than Duchenne. So this particular patient retained critical components of dystrophin and was still ambulating until an older age of 61. And so when we designed our micro-dystrophin construct, we took lessons from that patient and made sure that we were maintaining these critical regions. Specifically, if you'll note, if I can repeat, 1, 2 and 3 directly bind to the muscle membrane and really are important to maintaining force and resistance to contraction. Also, we've influenced the critical components to restore the dystrophin-associated protein complex. So all of this design and iteration over many years has led to the development of SRP-9001. And it shouldn't be forgotten in terms of the differentiation of our construct. So now turning to NSAA. Apologies, just wanted to highlight the key considerations for evaluating gene transfer therapy. First and foremost, we always highlight safety. It's critically important for each of the trials. We'll go through the safety profile. Next, in terms of biopsies, it's important to know, is the transgene getting into the cell? We do this by measuring vector genome copies per nucleus. Is it making protein? And this was measured by the Western blot measuring the total amount of protein. And then is it getting to the right place in order to do its function? So is it getting to the muscle membrane? We measure this by percent positive fibers and intensity. And then finally, is all of this leading to functional improvement? And our primary functional outcome across these studies is NSAA, which I'll talk more about, and the various time function tests as secondary outcome. So now, specifically, we'd like to highlight the understanding of the NSAA. This is such a critical outcome. It's a 34-point scale of 17 different activities and these represent daily life. And each patient could have 0, 1 or 2 for each of these activities: 0, you're unable to do the function; 1, you can do it with assistance; and 2, you can perform it without any difficulty. And each of these measures is , as you can see on the left, it goes from most difficult. And these are left early on in the course of disease to least difficult or last week. So some of the more difficult activities are things like hopping on one leg and rising from the floor. It's important to note that these tests are administered by professionals. And this whole scale was developed by investigators to really mimic clinical meaningfulness. So now thinking about one of the more difficult activities rising from the floor, what we'd like to do on the next slide is really show you what this means because I think seeing it in action really links to the point in how different one point can mean on this scale. Now we're going to see the first video. To direct your attention to the video screen, on the first video on the left is the boy rising from the floor. And this is a 2 point on the scale. Are you able to play it? Is it playing? Sorry, I'm unable to -- he's able to get up from the floor without any difficulty at all. Next is 1 point on the scale. And you can see that this child has significant difficulty getting off the floor. He needs to use his arms to be able to push him off -- push himself off the floor and has significant difficulty. And now 0 on the right is a hard video to watch because he is unable to get off the floor. And this really speaks to the impact of the disease and how it impacts not only a child having difficulty getting off the floor, needing assistance to get to a sitting position, then also the family because he becomes completely dependent. And so this just speaks to the significant impact of a 1-point change on the scale. And so as we go through the NSAA data for each of the components, please keep this mind on the impact of these changes. Okay. Now turning to Study 101. To remind you on the next slide of the study design. This was a single-center study of 4 boys, open label with baseline biopsy, biopsy at 12 weeks. And then the boys will be followed for a total of 5 years. Patients have confirmed DMD and patients between 18 and 58 can be negative for AAV antibodies. So we'll review the results that we've previously shown. On the next slide, I'd just like to point out the baseline characteristics for the first 4 patients. As you can see, at treatment, there was a mean age of 4.8 and a baseline [indiscernible] mean of 20.5. Now first, just to review the biopsy data. This is their expression. For weeks, you can see all 4 boys post gene therapy had robust expression with a mean intensity of 96% and percentage of dystrophin-positive fibers of 81.2%. On the next slide, it was also important to show that this micro-dystrophin expression led to the restoration of the dystrophin-associated protein complex. As you can see on the right, this is beta-sarcoglycan, one of the components of the DAPC, was up-regulated, correctly localized to the sarcolemma post treatment. The next, just to summarize the expression results with the -- by Western blot. We had 74.3%. When it was adjusted for fat and fibrosis, it was 95.8%. Intensity, 96%; 81.2% of positive fibers. And we looked at vector genome copies per nucleus. There were 3.3 copies. So next, now turning to the functional time test results. What we want to point out is that NSAA will show you that data has consistently improved over time, but the plan function tests have also been supportive and improved over the course of 3 years. We could see an all-time function test of each of the patients has improved. So for time function test, you're looking for a reduction in the amount of time that it takes to do the activities of regular -- you see actually a positive response with the exception of the 100-meter percent predicted when you are looking for a positive response. So this is where you want to see a rise in the amount of prediction that it takes to do the activity. And so next, turning to the NSAA. And so here, this is each of those 4 boys individually showing their NSAA scores over time from 1, 2, 3 years. And on the right, you're seeing -- we saw a mean of 7.5 point improvement from baseline in the 4 boys. And what's important to note, as I showed you previously, is that the boys were 4 to 6 at treatment. But now after 3 years, they're at 7 to 9 in terms of age. And so at this point, we would be expecting a downward trajectory for these patients, but yet, they're stable or improving. And so what we did in the next slide is did a comparison to natural history. And so we used a model to predict where these boys would be over the course of 3 years. And we developed the trajectory model from approximately 300 patients across 3 data sets and then modeled the trajectory over the course of 3 years using an [ MRM ] model to fit. As you can see, we see a highly significant difference between the treated patients and natural history comparator arm with a highly significant p-value of less than 0.0001. And so what's really important about this, and as you can see from the graph, is that the treated boys continued to improve over the course of 3 years, whereas the natural history comparison are declining over that. We really see an increased separation over time. And this has -- lead to about an 8.6 difference when treated compared to natural history. So really strong, and this just speaks to the point that the effect will grow over time. And then we're really looking for a difference from natural history over the course of time. And it will become apparent in our next trials as well. So if we now turn to Study 102, so for safety summary. In terms of safety on 101, we now have 3 years of various follow-ups and no additional safety signals. We had treatment-related SAEs that were mild and moderate and all resolved. And these all occurred within the first 90 days of treatment. We had no additional safety within the second or third years of follow-up. The 3 patients had -- 3 patients that we did see safety signals were elevated, damaged [ GPs ] in the first 3 months. And they all resolved with various treatments. And then the most common treatment-related AE was vomiting. We had no AEs that were associated with clinical complement activation. And so to summarize we've seen no additional safety signals since our last report. It was all within the first 90 days. Now turning to Study 102. To remind you of the study design, the study was designed with one stratification factor, and that was age. And so this is a study of 41 boys, stratified by age where half were 4 to 5, generally half, and the others were 6 to 7. This was a randomization factor. The Part 2 of the study, half of the children received SRP-9001. The other half received placebo. The primary end point was NSAA at 1 year. And then the boys crossed over to the other treatment. We did have a weight-based dosing at 1.33x10^14 vector genomes per kilogram. That's using our linear qPCR standard. As you recall, we started off Study 102 with a different titering method, using supercoiled qPCR. And then it was later identified when we developed our validated qPCR method with linear standard that 60% of those patients had a -- did not receive the target dose. And so we'll talk about that later on. We did convert to 1.33x10^14 of linear qPCR. And that's all children received that dose moving forward. Next slide. These are the baseline demographics. As you can see, half with the placebo, half with the 9001. Approximately 8 children per group were in the 4- to 5-year-old group. The rest, majority, were 6- to 7-year-olds. Next slide. So these are the baseline demographics in the intensive treatment group. And what we'd like to highlight here is that in all cases, this is going in all patients. The placebo group is better in all parameters relative to the SRP-9001 group. So all of the time function tests, in addition to the NSAA, the placebo group was better at baseline. This is all patients. When you look at just the 6- to 7-year-olds, this is even more profound. The next slide. These are the micro-dystrophin biopsy end points. This is just a summary slide. So at 12 weeks, we saw a mean of 28.1% by Western blot. As far as immunofluorescence, with 63.7% intensity, 32.9% dystrophin-positive fibers and a mean of 1.6 copies per nucleus. And this is Part 2 of the study. And so now translating into function. As we mentioned, the primary end point was NSAA. And this is looking at all patients in the study. So as you can see, we saw an improvement at every single time point in the treated group. That was a baseline -- the change from baseline of 1.7 points compared to 0.9 points in the placebo group. This is not statistically significant. And the reason for this becomes crystal clear on the next few slides. If we go to the next slide. As you recall, we had a prespecified analysis with the 4- to 5-year-olds. We did see a statistically significant improvement in NSAA versus the placebo group at week 48. As you can see, the baselines were very well balanced with 20.1 and 20.4. And in just 8 children per group, we did see a significant difference at 48 weeks. Now if you look in the next slide of the 6- to 7-year-olds, we see a much different picture, where the baseline NSAA for the 2 groups, placebo versus treated, were significantly different from one another, 19.6 versus 24. It was highly significant actually with a p of 0.0046. Now what's important to note is that when we look at the placebo group here as a validation for our natural history model, we use that prediction model to model out where the placebo patients would be in the 6- to 7-year-old group. And in fact, it was highly concorded with only 0.3 point difference between what was observed at 48 weeks versus what our model would have predicted where p was equal to 0.0013. So this really validates our natural history model that we've used for the additional study. Now on the previous slide we have noted was that the treated patients -- let me go back, sorry, for 1 second. Yes. What I'd like to highlight here is on the bottom, the purple graph, the treated patients. Although they were significantly affected, and in fact, many of these patients had a rise time of greater than 5 seconds, they remain stable over the course of the 48 weeks. And so when we compare those on the next slide to natural history arm, where we've matched for both baseline age and NSAA, we do see a significant difference from what natural history was predicted with a 0.3 change in SRP-9001 group versus a 3.2 change in the natural history comparison. So an approximately 3-point difference in treated versus natural history. So now turning to safety. In terms of safety for 102, the -- we've seen no additional safety signals, but let me just summarize. So approximately 85% of the patients had treatment-related SAEs. But these were primarily mild, with the most common being vomiting or there was no clinically relevant complement activation observed. There was a total of 4 patients met with 5 degenerative SAEs. 4 of those were in the treated group and 1 in the placebo group. 3 were random analysis, 2 in treatment group, 1 in placebo. These resolved. And we had increased transaminases in another patient that resolved with increased steroids. No other clinically significant laboratory finding. No complement activation -- clinically relevant complement activation. So in terms of safety, no new safety signals since the last report. And so now turning to Study 103. Again, this is our commercially representative process material. And we'll be looking at 6-month functional data. We previously showed expression data for the first 11 patients. So we go to the next slide. We'll talk through the study design. And this is a multicenter, open-label study looking to evaluate the commercial process material. In total, 32 boys have been treated. 20 in Cohort 1, this is between ages of 4 to 7. 6 in cohort 2, this is older ambulatory between ages of 8 to 17. And in cohort 3, this is 6 patients who are nonambulatory. In this study, we did expand the patient population to study all mutations. As I mentioned in the previous studies, we studied 18 to 58. This is, again, a linear qPCR dose of 1.33x10^14. This -- we will be showing results of the first 11 of 20 ambulatory boys in Cohort 1. On the next slide, this is the first 11 patients of Cohort 1. I just wanted to point out that the majority of these patients, 9 of the 11, are age 6 to 7. The other are age 4 to 5. You can see various demographics here. So the next slide, just to remind you of the expression profile that we saw. We saw robust expression correctly localized to the membrane, again, with our commercial process material at 12 weeks. We saw a mean expression of 55.4% by Western blot, 70.5% by -- dystrophin-positive fibers and 117% in terms of mean intensity. And so now on the next slide, we'll show the 6-month functional data for the NSAA. And what you can see is that we saw a 3-point improvement on NSAA at 6 months in the first 11 patients. And if you look at just the 6- to 7-year-olds or 9 of those patients, we saw an improvement of approximately 2.9 points from baseline. So it's virtually identical to the total cohort. So very pleased to see this improvement across 6 months. It's -- when we look at natural history, we will predict approximately 0.5 point decline in natural history if we were to look at this age group. So that would be a delta of approximately 3.5 over the course of 6 months. And now turning to safety. In general, safety was similar to what we've seen in our previous studies with the most common AEs being vomiting. We also saw increases in liver enzymes in response to corticosteroids as we've seen previously. And no clinically relevant complement activation was observed. We did have one new SAE in cohort 2, which is immune-mediated myositis. This patient experienced muscle weakness post infusion. We did have a cellular immune response that was detected to his mutation, and I'll talk more about that. There was no associated myocarditis. He did receive plasmapheresis and discharged [indiscernible]. And he has now returned to his pre-event function and is doing well. As you recall, an immune response to the transgene was the theoretical risk that Dr. Mendell and I had previously identified. And as you remember, our first 2 trials, we excluded certain patients for the 3 events. So then in Study 103, based on the safety profile, we had expanded mutations in 103. And in total, we had dosed a total of 9 of the 32 patients with mutations in the N-terminus. Only one of those patients had difficulty, and that was the SAE I just described. This patient had a particular mutation in the N-terminus that was large and different from the others. We've taken the conservative approach to exclude these new patients in the 301 study. But we'll be studying this group separately. So in addition to the 8 additional patients that we've already treated in this area of the gene, we'll be studying to understand exactly what patients -- or what mutations are at risk. Ultimately, based on the data in hand, we believe that the number of mutations at risk is quite small. And we'll do this in collaboration with external consultants to determine the final number. But it would be well below 3% in terms of the final number that would potentially be affected. But we'll confirm that over time with data and in consultation with experts. So now turning to the -- just to summarize so far what we have described across SRP-9001. What I hope you can appreciate is now we have a wealth of data across 77 patients with SRP-9001 in 3 ongoing studies. 45 has been dosed with the clinical process material and 32 have been dosed with commercially represented material. 59 were ambulatory boys and 12 -- reaching ages of 4 to 8; 12 ambulatory boys, greater than 8; and 6 are nonambulatory boys. We have a wide age range of 4 to 19 and then also a very wide weight range between 13.5 kilograms and 80 kilograms. So we have successfully dosed older and nonambulatory boys up to 80.1 kilograms, which speaks to the breadth of our data. We see durable clinical activity with sustained functional improvements compared to natural history, now having data up to 3 years in Study 101 with consistent positive expression results. And all patients are at least 6 weeks of follow-up. And as I mentioned, the first 101 boys are now here for follow-up. We also have a consistent safety profile. Now turning to the 301 design. As you know, we will -- we've taken all the learnings from our 101 and 102 study and then applied them to the 301 study. So I've highlighted the key features here of 301. We have an N of 120 patients. And we've now included 2 stratification factors, which include age and baseline NSAA. In addition, to ensure homogeneous trajectory, we will also have an inclusion criteria for time to rise with patients having a rise time of less than 5 seconds. We'll have a balance between the age cohorts of at least 50% of the patients, 4 to 5. This is a 1:1 randomization between the 2 groups. In terms of the study design, it's similar to the 102 study, where you have patients -- the patients receiving treatment while the others receive placebo. The primary end point is at 52 weeks for NSAA. Patients are then crossed over to the other treatment and then followed in a long-term extension study. This is a global study with 43 sites in approximately 10 countries. And in terms of the dose, just again, the 1.33x10^14 is the linear standard. As I mentioned, the inclusion criteria includes the rise time less than 5 seconds; stable dose of corticosteroids, which is the same regimen we used previously in 102; and antibody titers less than 1 to 400. We are excluding mutations between the exon 1 to 17, as I mentioned previously, and also mutations with an exon 45 because of the [indiscernible] disease. The primary end points are NSAA at 52 weeks but with various secondary end points of expression and time function tests. So this study is very well powered to seek the benefit of -- in the NSAA at 1 year. On the next slide. So just to summarize, 9001 is the only candidate in development with results from a randomized placebo-controlled trial. This contract has been very carefully designed and studied over time. We've developed a robust and reproducible manufacturing process and improved assay matrix acceptable to FDA and other regulatory agencies for release. We've now demonstrated robust micro-dystrophin in Duchenne patients with this commercially processed -- commercially representative process material. And the totality of the clinical data that was demonstrated and showed you today shows you that the micro-dystrophin construct can serve clinically meaningful benefit to Duchenne patients. This benefit is sustained and we've now shown up to 3 years of follow-up in terms of function. Our broad 80-plus patient experience, including those of limb-girdle, including patients over 70 kilograms has, to date, observed a safety profile that's consistent and manageable. We continue to use only a single-drug steroid regimen. But now where are we going next? We have several upcoming milestones, which include announcing expression of functional data for SRP-9001 102 Part 2, and that will be in early 2022. And then as far as our 103 study, our commercially representative material, we'll have a Cohort 1 full expression and function coming up as well as cohort 2 expression function as well as cohort 3. And so this will tell you the breadth of data in both our younger patients 4 to 7 older ambulatory patients and older nonambulatory patients, really addressing the full spectrum of the disease in Duchenne. And so with that, I'm going to turn it back to Doug for question and answers.

Thank you, Louise. Catherine, let's open the line for questions.

[Operator Instructions] Our first question comes from Tazeen Ahmad with Bank of America.

Can you hear me?

Yes.

Okay. Great. Just wanted to get a sense about Study 103. So based on the data that you presented today and based on your understanding with FDA about what needed to be shown and by when, Doug, can you give us an update on where you stand today?

With just what? The path forward?

Whether or not you had to agree to show a certain level of efficacy or equivalents between the commercial grade supply and your previous data and how today's data would potentially address that.

Okay. Yes. So just so we're clear, that is not an issue at all today. So just to remind us, there was a period of time when we were considering some bridging issues between the clinical supply and the commercial supply. As we stand here today, just so we're absolutely clear, we've had very productive discussions with OTAT and FDA, OTAT, as everyone hopefully knows, is the division within CBER and FDA that is responsible for cell and gene therapy. They have blessed the commencement of our next trial, which is Study 301 and all of the assays associated with that trial. So we are in good shape. We don't have any additional work to do or bridging to do as well. And so as we announced very recently, within the last 2 weeks, we have already initiated Study 301 with the blessing of the FDA and, obviously, with regulatory authorities around the world. And we're getting -- going as fast as we can with Study 301.

Our next question comes from Gena Wang with Barclays.

So Doug, maybe wanted to ask you now with the new evolving data, the Study 103 data you showed today, and then beginning of next year, you will also show Study 102 2-year data. So what kind of data package will make you reconsider the potential -- approval path? And then very quickly just on Study 301, is the study powered to detect significant differences in Group 4 to 5 and also 6 to 7?

Yes. So let me answer the second question first, yes. You'll notice we have at -- an N of 120, this is a robustly powered study. And I think kind of going to the point that I made at the inception of the meeting, we have prioritized probability of success. It is, from our perspective, the wisest way to ensure that children around the world will get this therapy. And that's why we didn't cut any corners. We've been very specific with respect to the protocol. We've got floors, we've got ceilings. We've got -- we're eliminating -- or limiting kids to being under 5 seconds in rise time, not because, obviously, the therapy is going to limit that but just so that we reduce any heterogeneity when we see it. And of course, the N of 120 answered that same question. And we did that not only for the proven analysis but for the separate prespecified analysis as well. And just to remind everybody, so there is no misunderstanding, we are stratifying on the basis of not only age but baseline characteristics that we will not have the issue that we had in part study 1 -- Part 1 of Study 102 nor will we have any of the titering issues that we had in Part 1 of 102 because we use our linear approach. Now let's go to the first part of your question, the question about the path forward. I want to be very clear about the discussions that we've had with the agency today. So our goal this summer was to get our pivotal trial Study 301 going, get them to bless all of the assays, get them to bless the protocol. And we've done all of that to the satisfaction of the FDA. And obviously, we're doing the same thing with regulatory bodies around the world. And I will repeat what I've said many times, which is the base assumption that everyone should have is that Study 301, EMBARK, we should probably start using the name of the study now, EMBARK, that is the pathway to our approval in the United States and around the world. It is also true, Gena, that the data that has evolved is becoming significant in the breadth of the data and in the signals. I mean in almost every single way you look at this, the potential for profound effect comes off, not just the expression levels, the safety profile, the functional signals, the restoration of the dystrophin-associated protein complex. We're not showing some of this stuff today. But we have the independent study of muscle MRI, it's clear. So we will see what happens early next year. As everyone knows, we're going to have very informative information early next year. Part 2 of 102, we'll have that completed right at around the end of this year. And we'll see that and have that data available in the first quarter of 2022. And there is going to be a wealth of information. That's 41 patients. Half of the -- approximately half of those kids who have been on that therapy for 2 years against a prespecified, not post hoc because it's a number that is still blinded, a prespecified natural history cohort. And then the other group will have been tracked for a year off therapy and tracked for a year on therapy, again, against the natural history cohort if you look at their trajectory. If that data is additionally compelling with everything else we're seeing, then we will, of course, consider whether -- given the wealth of the information that we have and the data set that we have, whether the totality of evidence justifies us beginning to have conversations with divisions, agencies around the world about a more accelerated approach to bringing this therapy to patients who do not have the benefit of time. But I want to be very clear, from my perspective, the base case assumption, and I would argue for investors, the investigation assumption ought to be that Study 301, as it stands today, will be the pathway for approval for SRP-9001 in the United States and around the world.

And our next question comes from Brian Abrahams with RBC Capital Markets.

I really appreciate the comprehensive overview and contextualization of the data. I was wondering if you could maybe expand a little bit more on the immune SAE that you observed in terms of which mutation, whether there was any overlap with what any other companies have observed with their transgene or might you expect some differences there and any kind of monitoring that you'll incorporate into the ongoing studies. And then I'm also just curious on that front. The natural history data that you showed as a comparison for Studies 102 -- 101, 102 and 103, to what degree does that represent all the different mutations matched by age and NSAA versus just the ones that are being included in these studies?

Yes. I'll turn those 2 questions over to Dr. Rodino-Klapac.

Yes. In terms of the SAE, this was the mutation in the N-terminus and it was large -- significantly larger than the others. We do note that other sponsors have seen cases of immune-mediated myositis. And we do think it's related to this class effect of certain mutations, certainly not related to anything that we tested. So in terms of monitoring this, this is something we're already monitoring for and that we will continue to be monitoring for closely. We're being conservative by excluding all these mutations, as I mentioned, in the next trial. But we'll be sure to identify the exact mutation at risk because, ultimately, we believe that it's a very small number that would be at risk specifically. The size of the mutation does appear to matter. But ultimately, over time, between the data and the consultation with experts, we'll determine the exact amount. And so it is specific to each program in terms of constructs. So you make different exclusions in other trials, not just based on the -- what's included in the individual construct. So the second question was around the natural history. And so this is a comprehensive data set that was derived from 300 patients. So it does represent our population. You develop a model. And then in certain cases, where we can -- these are -- been matched for both age and NSAA. So it does take into account the considerations or the characteristics of our patients in comparison.

And one final thing, just to bring us back a bit on this mutation-related issue. And I know -- we know, for instance, that Pfizer has had 3 of these. And then 2 of them actually were associated with the myocarditis, ours was not. But I want to be clear that we feel pretty confident that this is a mutation-related issue that just may very well be the case that Pfizer just got unlucky with their 3. And if you look back, actually, we're -- the data that we've developed has been very informative but for our ability to bring this therapy to the greatest number of patients possible in Study 101 and 102 because of the theoretical risk that if you have a mutation in these ranges, you might have an immune response to the gene itself. We were very conservative. Then in 103, the ranges were expanded. In fact, they were -- those issues were eliminated. And while we're very -- we wish we had, had none. It was actually quite informative in that 1 through 17 range of exons, we dosed 9 children with mutations in that range that would have -- historically, we would have worried about theoretically. In fact, 8 of the 9 had no issue. And the ninth that had an issue, it isn't just random. That mutation is qualitatively different than the others in a very significant way. There's -- there are 2 issues there. There's this enormously large mutation -- deletion and it's in 1 through 17. Now we're taking a conservative approach for Study 301, of course. And while we have limited the exclusions from 101 and 102, we are going to maintain this idea of excluding exons 1 through 17 during Study 301. But we are separately -- and frankly, we're going to do it as rapidly as possible. We're going to start other development and studies so that we can really frame out what is this combination of where the mutation exists and the enormity of the deletion that would be a genuine risk. So we're pretty confident, given all the data we have right now and given what we've done in 103, that by the time the SRP-9001 is approved, it will be a fleetingly small group of children who would have to be excluded. And that's, of course, extremely important work because if we remain with this conservative approach that we're taking for 301, 8 of the 9 kids that are -- today have been treated and are doing very well won't -- wouldn't have been treated, and that's not right for the community. So we are going to do the work so that when we launch this therapy, we're going to be in a place where a fleetingly small number, we suspect, of children would be at any risk of this issue.

Our next question comes from Salveen Richter with Goldman Sachs.

Could you just comment on your confidence level here that when you've gone back and looked at the previous studies that these 2 stratification methodologies and the titering aspect are really the only trial design changes that need to be made here?

We've tweaked a number of things in the protocols. But there is no doubt that in Study 102, the therapy performed brilliantly. The study did not for the top line because of the issues that we've mentioned. The first is, of course, as we went back and used our linear titering method and retested the lots that were released through the supercoiled method from Nationwide Children's Hospital concluded that 60% of the kids were dosed at substantially less than the 1.33 e^14, which was our target dose. That is 100% solved. So we're very clear. It was solved at the crossover for 102. And of course, it's all 103 and then will be solved in EMBARK, of course, because we're using this linear method going forward. And it is also clear, you see it jump off the page for us, that the problem with the 6- and 7-year-olds is that the baseline characteristics were so wildly different that it was comparing the long group. There was just no way to intelligently compare -- to actually use this very sophisticated natural history modeling that we use. We would have seen a strong statistical significance if we had, had the right group against the 6 and 7. That is also fully solved in Study 301. We are stratifying on the basis both of baseline characteristics as well as age. So we will not have a repeat of that issue. And then of course, if you look at Study 102 Part 2 and you look at the 4s and 5s, even with the titering methods, when you get the baselines right, in a short 48-week period you see a strong statistically significant benefit of this therapy in the 4 to 5. And that's in 16 patients. And yet, our Study 301 is 120 patients across those age ranges with those tight entrance criteria. So we have an enormous amount of conviction in the design of the study and, frankly, in the probability of success of the study.

Our next question comes from Alethia Young with Cantor.

Maybe just 2. One, can you talk about kind of how you're thinking about enrollment time lines for the next commercial study in light of everything going on in the space? And the next one is just kind of a big picture question. Are you -- I mean there's a lot of obviously some concerns, not exactly on gene therapy but around safety. So I wanted to get your perspective on how you're thinking about kind of the safety and the FDA is thinking about safety overall with cell and gene therapy.

Okay. So on the first one with the enrollment, so we're going to take -- obviously, if we had started much earlier in the year, I think I had talked before about given certain time lines, we're now obviously initiating this trial. And we'll have kids enrolled and dosed this month. So it will -- first things first, it will enroll robustly. And some -- sites come online. And we'll have eventually over 40 sites up and running. We'll be starting in the U.S., which I'm very excited about. And then we'll be in Europe and Asia and the like. The study will robustly enroll. Just to give you an anecdote, one of the reasons we're being a little circumspect about exactly what sites exist because we have heard stories anecdotally of patients trying to move countries just to get the chance of being in a study. And that -- we would not recommend that. And so we're being careful about that. The study will robustly enroll. I am going to say that, conservatively, we'll have this enrollment complete and dosed by the first -- by the end of the first half or in the first half of next year. So -- but it will enroll very robustly. There is an extraordinary demand for potentially transformative therapies such as this one. On the safety side of things, the FDA has reviewed -- let me focus specifically on our therapy. I don't want to be overly competitive. So we'll talk to others we've seen. But we have, and you saw it in the Advisory Board that occurred in September, there have been issues that have been discussed with respect to other AAVs, things like effects on dorsal root ganglion, neurological effects, integration-related issues and then, of course, this concern that people have regarding complement-mediated aHUS in kidney failures and the need to use SOLIRIS and the need to use dialysis on kidney failures. We see none of that. We have had none of that in our program. The FDA has looked across all of our safety database, including, obviously, all the information that we provided today and on that basis has blessed our ability to move to our pivotal trial. That's why we've been able to initiate it. So we feel very good about where we are as an organization from a safety perspective, an expression perspective, functional signal perspective and a performance perspective.

[Operator Instructions] Our next question comes from Anupam Rama with JPMorgan.

On Study 301, just thinking about your inclusion/exclusion criteria based on time to rise. What portion of 4- and 5-year-olds and 6- and 7-year-olds will be excluded from this? And then, logistically, how many attempts do the kids get to meet the criteria?

Louise?

In terms of the 4- to 5-year-olds, it would be a low percentage that would screen out for that, very small. 6 to 7, I can't give a specific number. But obviously, the goal of doing this is to create a study that has a high probability of success. And we have successfully treated patients with rise time above 5 seconds. And so our -- and we've proven that within Study 102. And so I think that's an important thing to keep in mind. The -- I'm sorry, was there a second part of the question?

How many times did -- I think the question that Anupam asked is how many attempts does the child get to rise before you decide that they are definitively over 5 seconds.

The physiotherapists will do the test once officially.

One final thing we should say. Look the exclusion criteria -- just so we're very clear and I'm sure everyone understands this. The goal of the majority of these exclusions for purpose of the trial is to simply reduce heterogeneity so we see an effect. Obviously, I think I'm telling people what they don't need to know. But one should not imagine that, commercially, this therapy wouldn't be available to children who had rise times over 5 seconds. That's not the goal. We just -- in a controlled experiment, you have to reduce this. Now there is one significant exclusion criteria that is meaningful commercially. And so ensuring that, that exclusion percentage is low is unbelievably important. And that is exclusion for preexisting binding or neutralizing antibodies because we're not excluding that for heterogeneity issues. We're excluding that because as it currently stands in gene therapy, if one has a previous experience with something that looks like the capsid and then therefore has neutralizing or binding antibodies related to that capsid, it isn't currently safe to dose those children. The good news about our construct right now, and it's stayed very stable over a long period of time, is that we're in the mid-teens, around 15% or so excluded. We don't like 15%. We'd like it to be 0. And Louise and team are doing an enormous amount of work to eventually someday have a pathway through the use of other therapies or other treatment paradigms to knock down pre-existing neutralizing antibodies to bring these children back into frame. But the good news is the vast percentage of children will not have a preexisting neutralizing antibody. And the second good thing at least so far, remember, is that we're a little different than other companies. We're not simply working with rh74 for Duchenne. We also use rh74, which is our capsid, for limb-girdles. And limb-girdles affect adults as well as children. So we've had an opportunity to look at screen-out rates over ages. And interestingly enough, it does not get significantly different as one gets older. So even looking up at full adults with limb-girdles, the screen-out rates we're seeing from an rh74 perspective at the seroprevalence looks about 15% stably. So that is a really important issue to ensure that we can reach the maximum number of patients possible.

Our next question comes from Colin Bristow with UBS.

And thanks for the helpful presentation. Just a little bit more on the controlling for baseline characteristics and specifically trajectory. I think if you look at the [indiscernible] data, certainly at the younger age, is like around 4, 4, 5. The different classes can have very similar baseline NSAA scores. And so how well does the time to rise control for this? And did you, at any point, consider just doing serial NSAA to just try and at least plot trajectory on a shorter-term basis? And then just second quick question. What's the longest duration data you have for expression? And any commentary you could give on just what you're seeing in terms of trends over time would be helpful.

I'm going to turn the first part of the question over to Louise. On the latter part of the question, obviously, we have 3-year data from the children from a functional perspective. From an animal model perspective, one of the things we've often said is that, in one sense, muscle is a very difficult place to go for gene therapy because getting the tropism and getting the gene to the right place can be very challenging. Obviously, we've solved for that. We're now seeing with respect to our expression and our commercially representative material some 3.8 genome copies per nucleus. The good thing about muscle is that once you get there, you don't get significant turnover. So you see very good long-term durability. So we've been -- we've seen solid, consistent durability for as long as we've been able to look across models for micro-dystrophin generally, mice models, nonhuman primate models; with our partner, golden retriever models. So we will see we predict very good long-term durability with this therapy. And we're certainly seeing it functionally, of course, in the first 3 kids -- first 4 kids. But with that, I would turn over the first part of the question to Louise.

Sure. So characterizing the baseline characteristics like age and baseline NSAA is highly predictive. And in terms of the rise time, it's been shown that a rise time less than 5 seconds is predictive loss of ambulation within 1 year. And that's been demonstrated by experts in the field. I would just add that as far as the predictability of that, you can just look to the model work that we did. So using the model that we generated, we then applied that to the 102 placebo patients. And using those baseline characteristics, we were able to see very highly concordant data between what was predicted and what we actually observed. So there was less than -- there was about 0.3 points difference between the 2. In terms of the predictability of applying the NSAA and age criteria using our model, it's highly predictive of what we would expect. And so that kind of leads to our overall confidence in the 301 study. But it's very well adequately powered based on those criteria.

And I would just add, finally, just to keep reminding us, I mean this is an N of 120. So we have robustly powered this study so that we should be in good shape.

Our next question comes from Brian Skorney with Baird.

I was hoping you can maybe help us sort of understand the key differences between Part 1 and Part 2 of Study 102. It seems like the Part 2 analysis that we'll see early next year is becoming increasingly important. So when we sort of look at that Part 2 data, what should we really be looking for as the proper comp here to determine the effect size? Are we looking at something like the 2.9 point benefit, which I think is what you're saying you saw in the 6- to 7-year-olds in the 103 cohort at 6 months? Just help us think through what would really be the expected difference here between sort of the better baseline controlled Part 2 102 and the Part 1.

Sure. The first thing to know, Brian, is that this will all be prespecified, I mean I think post talk about it. But we'll have a prespecified natural history cohort, both against the kids that have been on therapy for 2 years as well as the kids that have been on therapy for 1 year plus a trajectory analysis plus a pooled analysis, I believe, as well. And so the broad strokes, our goal is to see a significant difference between the natural history, what a kid would do with these trajectories versus what they're going on therapy. But beyond that, I will turn it over to Louise to comment.

Yes. I will just point out some of these are several pieces of data from 102 in comparison. You'll have the 2-year data of the treated patients that were treated in the first -- Part 1. We'll have the comparison of the patients treated in Part 2 compared to natural history as well as the initial placebo. And so there'll be several pieces of data coming from that, that will be critically important, and also, the trajectory of data from the patients that were placebo and then crossed over. So taken together, those 3 pieces of data will be highly supported and segment is prespecified.

And then remember, the 2 confounding issues in Part 1 in Study 102 was the titering method that was previously used as well as a mismatch in the baseline characteristics of the 6- and 7-year-olds. I think the use of a natural history set -- because we won't be able to use the placebos. No kid will be on placebo by then. But the use of the prespecified, rigorous matched natural history set can correct for that issue, of course, in Part 2. And then as you know, in the Part 2 crossover kids, half of the kids that were crossed over, that titering issue was resolved. We used the linear titering method. So we should be in a good shape to get very informative information out of Part 2 of Study 102 early next year. We're very excited to share with everybody in early 2022.

Our next question comes from Joseph Schwartz with SVB Leerink.

I was wondering how many sites will you use to enroll the 120 patients in EMBARK? And how will you ensure that you don't sacrifice any rigor when you go from -- I think there were just 1 or 2 sites for Study 102 and 3. And so you don't try to power up the trial but potentially dilute it by having more variability across many different clinicians evaluating patients with NSAA.

Yes. So we'll ultimately have about 41 sites or so around the world. And the second point is a really good one. It's something we obsess over. I would say the way we're going to ensure that we don't increase standard deviation as we move to sites around the world is through sort of our quintessential obsessive detail-oriented world view. We are going to stay very, very close. We've got great monitoring. We've got a great clinical operations team. We've got the insight of people like Dr. Mendell, who knows how to do this, who's helping inform all of the other sites in precisely how to do this. And then we are going to obsessively focus on the rigor of the exclusion and inclusion criteria, the performance of the functional tests, the way the infusions are performed to ensure that we don't increase standard deviation across sites as we move across the world. So that second one is a real straightforward issue. It's just obsessive operational excellence and execution. And I would hope people would see, particularly with the work that we've done this year, that we're an organization across tech ops, commercial, medical affairs, development, clinical operations. We're obsessively a detail-oriented and executing organization. That's the best way. And then finally, of course, I keep saying it over again, we've got an N of 120. So we're taking a conservative approach even with respect to that. We are not cutting corners with respect to Study 301.

Our next question comes from Yun Zhong with BTIG.

So just a follow-up on the enrollment criteria. I wonder, are there any external natural history data besides what you have seen from the subgroup analysis from the Study 102 that suggest that in addition to making sure that the baseline condition is similar, you have to look at the 4 to 5 years old? That's a pretty narrow range to make sure -- I understand that increases data homogeneity, but just wondering if any additional evidence.

Dr. Rodino-Klapac?

Yes. I mean the short answer is that we certainly took into account all natural history data sets in designing this trial and informing our model. There are external data sets which support our model in terms of the prediction of our natural history. It's hard to say. Certainly, we took into account all the available data. And that was -- informed our model and consistent with our model in terms of the probability of success and powering our study.

I should also say -- I won't name them because it's premature. But there is an independent modeling approach that is unrelated to us that have applied to our data. It is almost identical. This is the 3-year data, so really the long term, almost identical to the information that we have. We're not at liberty to share it yet. That will be published early next year. And then we'll see the additional confirmation of the validity of our model. So we feel very good about where we are. And I will just finally say just -- and I don't think this comes as a surprise to anyone. No one in the -- at least in the sponsor world, has more access and more patient level data to inform their programs than Sarepta does given the amount of work we've done in Duchenne muscular dystrophy.

Our next question comes from Tim Lugo with William Blair.

This is Lachlan on for Tim. I was just -- wanted to clarify, Doug, when you were talking about the regulatory path forward in your base case. Are you assuming that you could file after the first cohort of EMBARK? Or would you be expecting to wait the second cohort and have a cohort with 1 year and a cohort of 2 years' worth of data before you went to the agency?

Yes. Thanks for asking that question so I can make sure that people know that. It's the first cohort, unquestionably the first cohort. That is our goal. That is our primary. Thank you for that.

And our next question comes from Ritu Baral with Cowen.

This is Anvita on for Ritu. Just a quick questions from us, please. First, in Study 101, do you think there's a slight roll-off of effect in Patient 1 at Year 3? Any thoughts on why that might be? And then Study 103, how many 12- to 19-year-old have you dosed? And how many would you say were heavier weight across the cohorts?

I think I missed the first part of the question. If I'm not mistaken, you were asking about whether there was a dip in the first patient and whether that was meaningful. If I'm right about that, the answer is no. It's not meaningful just within the variability. That child has done -- if you look across the study, look across the 3 years, he has done just brilliantly and far better than natural history would have assumed. So of course, we're going to get moment-to-moment variability. These are function tests. Sometimes kids can be tired, et cetera, and it creates some variability. But those kids are all doing brilliantly. And then I'm trying to -- what was the second question? Louise, did you capture it?

I'm not sure.

Apologies for that -- the second question.

Our next question comes from Zhiqiang Shu with Berenberg.

I'm wondering if you can comment on how many patients in your Phase III EMBARK trial would be eligible for your 3 approved drugs, I guess if you think you are running into a situation where patients might receive those drugs in the trial and can potentially confound your results?

Yes. Louise, correct me if I'm wrong, but they would not be permitted to take another Duchenne-related therapy while they're in the trial. Am I correct about that, Louise?

Yes, correct.

There are no other questions in the queue. I'd like to turn the call back to Doug Ingram for closing remarks.

Well, thank you, everybody, for spending time with us this morning. As you can well imagine, we're very excited about where we're going with 9001. We've done -- I think the organization has done a brilliant job moving this therapy forward as fast as possible from a bank -- everybody involved. Our technical operations folks, our process development analytical folks have just done a brilliant job with respect to our commercial process material so that we're in a place where we are starting our Phase III trial in the United States and in Europe, in Asia. I want to thank our development team for the great work they did in developing what is a very sophisticated protocol. And of course, our research and translational group as well, the entire regulatory group did just a wonderful job of representing us with regulatory bodies around the world. Everyone has just done a fantastic job getting us to where we are. And I also want to thank our investigators, of course, who have done -- just been enormously valuable. And Dr. Mendell, who deserves just more credit than one could give a single human being for what he's done for Duchenne and SMA and gene therapy generally. Obviously, I want to thank Louise, who was the -- one of the co-designers of SRP-9001. We stand in a wonderful place right now. We're very excited about where we've come. We have an enormous amount of conviction to the transformative potential of SRP-9001. But I will end with what I said at the beginning, which is it is always dangerous with success to get complacent or arrogant. We stand alone. SRP-9001 is a unique construct with unique qualities. And it is undoubtedly to anyone whose objective, quantitative, qualitatively safety tropism function a therapy that stands alone. But we have an enormous mission in front of us. Every single day, patients and their families are not simply suffering from Duchenne muscular dystrophy but are degenerating and having their muscle stolen from them. So we are going to execute aggressively to enroll Study 301, to get 301 moving as fast as possible and to be fully enrolled. We are then -- as we look forward to the rest of the year and into early next year, we're very excited to come back and talk to you early next year about the results of Study 102 Part 2, which we'll have. And along with our just wonderful partner, Roche, who has just been a fabulous partner with us as we develop this therapy, we're really excited about moving this program forward as fast as possible, exploring the art of the possible once Study 102 Part 2 comes out and getting this therapy, once confirmed, to patients living with Duchenne muscular dystrophy around the world. And I look forward to updating everyone on this call and the community as we do that. And with that, please have a great rest of the week.

This concludes today's conference call. Thank you for participating. You may now disconnect.