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

Good afternoon, ladies and gentlemen, and welcome to Sarepta Therapeutics' SRP-9001 New Clinical Data and Integrated Analysis Conference Call. [Operator Instructions] As a reminder, today's program is being recorded. At this time, I'll turn the call over to Doug Ingram, President and Chief Executive Officer. Please go ahead.

Thank you, Michelle, and thank you all for joining us today for a review of our most recent results regarding the functional benefits and safety of SRP-9001, our gene therapy. After literally 1.5 decades of rational design, testing, optimization, brilliant preclinical evidence and then repeatedly confirming across multiple studies, SRP-9001 safety and efficacy profile, we, at Sarepta, have developed, as you can imagine, a significant conviction in the potential of SRP-9001 to change the trajectory of this disease, Duchenne. So you can also imagine how difficult it would be to exceed our expectations in the transformative potential of this therapy, and yet the results that we will be discussing today have done just that. They have met or exceeded our expectations on every metric considered. We see impressive statistically significant and clinically meaningful results. We see results using our commercially representative material at target dose that meet or even exceed our prior clinical material experience. We see results that bolster even further the powering of our ongoing placebo-controlled trial EMBARK, a trial that was already robustly powered. We see results that diverge from natural course of disease so that the impressive results we see at 1 year grow greater as the trajectory of the treated children diverge from the natural course of this ferociously degenerative disease, and we see a stable and differentiated safety profile. The potential meaning of these results to the lives of Duchenne patients should be patent to all of us. And with that, let me turn the call over to our Head of R&D and Chief Scientific Officer, Dr. Louise Rodino-Klapac, who will walk us through the results. Dr. Rodino-Klapac?

Thank you, Doug. Good morning, and thank you for joining us today. It's a great pleasure that we share with you new clinical data for SRP-9001, our gene therapy in development to treat individuals with Duchenne muscular dystrophy, or Duchenne. The data from our 103 study is particularly important because it's from our commercially representative material. It can be extremely challenging to achieve the same level of safety and efficacy when scaling up from small scale to commercial scale. So we are thrilled to see that these results are not only consistent but may be improved over previously reported results. This is a shining moment for the program. So I want to thank all of our CMC colleagues for all of their contributions. Here are our forward-looking statements. I or one of my colleagues may make statements or predictions about future events today. Please review our public filings for the risks that may come with these sorts of forward-looking statements. Now in terms of the agenda, today's presentation includes the following new clinical results. For SRP-9001-103, our ENDEAVOR study, we'll be sharing results from Cohort 1's 20 patients 1-year functional data from baseline and compare it to external control. For SRP-9001-101, it's a 4-year follow-up data from our first 4 patients, which includes NSAA and timed function tests. For study, SRP-9001-102, Part 1, 20-patient 2-year functional data. And finally, our integrated efficacy analysis, which includes Study 101, 102 and 103, a target dose in 52 patients. This includes 1-year functional data compared to propensity-weighted control. We will also touch upon the expression results from all of these studies and review our safety data, which continues to be a key differentiator for our program. Duchenne muscular dystrophy is a debilitating disease. SRP-9001 is intended to not merely treat but to change the course of this progressively degenerative disease. Duchenne results from a mutation in the gene that codes for protein called dystrophin. Dystrophin is a sort of shock absorber in our muscles. It attaches to the sarcolemma or muscle membrane and distributes force as we move, protecting our muscles from damage. Boys with Duchenne lack this dystrophin shock absorber to protect their muscles, and the consequences are as certain as they are brutal. The course of the disease is predictable. At a young age, a Duchenne boy will begin to struggle. Sometimes typically between 10 years of age and very early teens, he will become wheelchair dependent, followed by respiratory and cardiac complications, and ultimately, death will follow typically in the third decade of life. Our goal is to change all aspects of this cruel disease with SRP-9001. We believe today's clinical results advance us one step closer to changing the course of this disease. SRP-9001 is a full-body or systemic infusion of gene therapy. It uses a specific gene construct, which is designed, empirically tested and optimized after decades of research. It delivers a gene cassette that codes for a truncated but functional form of dystrophin to skeletal, diaphragm and cardiac muscle. And it's the only therapy in development that uses rh74 as its delivery vector and MHCK7 promoter to drive expression. It's these unique and rationally designed elements of SRP-9001 that explain why this therapy results in properly localized expression of the dystrophin protein across muscle, why it has a differentiated and unique safety profile and why it results in the functional improvements that we have seen in multiple studies to date. Sarepta has dosed more individuals with Duchenne than any other gene therapy currently in development. Today, we will discuss new clinical results from Studies 101, 102 and 103 as well as an integrated analysis of all of these studies. As a reminder, Studies 101 and 102 use clinical material as shown in the blue circles and Study 103, commercially represented material as shown in the purple circle. Our EMBARK study, or Study 301, as seen in the last column on the right, is advancing well and on pace. For these analysis, we'll be using prespecified external control analogies. Due to the lack of a placebo arm in Studies 101 and 103, we used a rigorous propensity score matching analysis to compare the treatment effect of SRP-9001 against an external control for Studies 101, 102 and 103. The propensity score method is a validated and commonly employed method to reduce bias and precisely predict treatment effect. Propensity weighting has served as the basis for approval of 2 drugs: defibrotide for veno-occlusive disease, or VOD; and burosumab for the treatment of X-linked hypophosphatemia, or XLH. The external control group includes data from the CINRG Duchenne Natural History Study, the FOR-DMD study and the placebo arm of an Eli Lilly study evaluating a treatment for Duchenne. All patients in the external control group received steroids. The inclusion and exclusion criteria rigorously matched baseline characteristics. The propensity score weighting included age, NSAA, rise from floor and 10-meter walk/run, which are all prognostic factors in Duchenne that are known to impact the progression of the disease. As a second step, a linear aggression model is used to correct for any remaining imbalances in baseline covariate. And so now turning Study 103, also called ENDEAVOR. Study 103 is an open-label clinical trial of SRP-9001. The study has enrolled 40 participants with Duchenne, including 20 ages 4 to 7 and expanded cohorts of older ambulant and nonambulant individuals as well as the younger cohort. Today, we will be showing functional data results from the 20 patients in cohort 1 between ages 4 to 7 who are ambulatory. As a reminder, we presented the first functional results from Study 103 in the first 11 participants in cohort 1, ages 4 to 7. We demonstrated a 3-point improvement from baseline on NSAA just 6 months after treatment. In terms of our baseline covariate, as you can see, the propensity score weighted balanced the baseline variable as well for the SRP-9001 arm and the external control arm. The baseline functional characteristics for both arms are nearly identical for all measures. And now for the data. We are thrilled to report that patients in cohort 1 of the ENDEAVOR study improved 4 points from baseline on NSAA. Pretreatment, they had a mean baseline NSAA of 22 at week 52, improved to a mean of 26, which is approaching the top end of the NSAA scale. To put these data in context, for example, these patients can now perform 2 activities unassisted that they were not able to perform prior to therapy for 4 activities that they needed assistance with but they can now do on their own. Equally impressive, SRP-9001-treated patients improved 3.8 points using unadjusted means and 3.2 points using least squares means at 52 weeks on NSAA compared to the propensity-matched external control, which is highly significant with a p-value of 0.0001. Simply put, the unadjusted mean difference is the main change from baseline of the SRP-9001-treated group minus the mean change from baseline in the external control group. The least square mean difference is commonly used for multivariate analysis to account for any imbalance. These results demonstrate that commercially representative SRP-9001 improved motor function and further confirms our confidence in the treatment effect of our therapy, increasing the probability of success for EMBARK. We've also see mean improvements across key secondary functional end points, including the time test. Shown here, Study 103 demonstrated mean improvements across key end points, which include time to rise, 10-meter walk/run, time to ascend 4 stairs, and 100-meter walk/run. In summary, patients receiving SRP-9001 improved significantly on every functional measure. The external control set only collected time to rise and 10-meter walk/run data. So these are the 2 -- only 2 analyses available using external control. For time to rise, we observed that SRP-9001-treated participants improved 0.9 seconds on unadjusted means and 1.2 seconds on least squares mean at 52 weeks compared to the propensity-matched external control group, with a p-value of 0.0001, which is highly significant. Now for the 10-meter walk/run, we observed that SRP-9001-treated participants improved 1 second using unadjusted means and also 1 second using least squares means at 52 weeks compared to the propensity-matched external control group, with a p-value of 0.0018, which is also significant. In summary, the results of Study 103 are truly remarkable, and our conviction is further strengthened by the consistency across all measures. And now I'll turn to our long-term data results. These data are particularly important because they answered 2 of our most common questions. Question number one, are these data clinically meaningful? And the second question, is the effect durable? So first, we will look at our original 4 patients after 4 years of treatment on SRP-9001 from Study 101. To remind you, SRP-9001-101 is a single-center, open-label clinical trial to evaluate the safety, tolerability and proof of concept of a single dose of clinical process material. The trial enrolled 4 ambulatory participants between the ages of 4 to 7. Participants are being followed for 5 years after treatment, while safety and efficacy continues to be evaluated. I will be showing you 2 analyses for Study 101. NSAA changed from baseline over 4 years in the 4 treated boys and then in comparison to external control youth using propensity score rating. First, once again, you can see that the propensity score weighting has balanced the baseline functional characteristics well for both groups. Now turning to the data. As noted on this slide, the data show that patients and SRP-9001-101 demonstrated a mean increase of 7 points in total NSAA from baseline to year 4. So I'd like to pause here because these data are particularly important. As these are older patients, around 9 years of age at year 4 and because Duchenne is a disease that gets progressively worse, these patients would now be in the steep decline phase of their disease. However, instead of declining, they have increased their function and importantly, maintained that increase, thereby demonstrating distinct treatment effects that increases over time, supporting the durability of SRP-9001. We are showing individual patient-level data here. So it's clear that all patients have remained stable and well above their baseline for this time period. No single patient is driving the mean of the group. We are pleased that all participants are continuing to benefit 4 years later. Now when we compare the treated patients to the propensity-matched external control, we observed a nearly 10-point difference using unadjusted means and a 9.4-point least-squares means difference, with a p-value of 0.01 at 4 years. As an example, SRP-9001-treated patients can now do 5 activities that those on the external control group are not able to accomplish. This durability question needs no further explanation. I will spend the bulk of the time on the clinically meaningful question. We are pleased to see that the treatment effect has continued to increase over time. We feel confident that we can detect a statistically significant difference at 1 year. However, this time period does not define the total clinical effect of SRP-9001. As expected and as shown on the graph on the left, the treatment effect increases over time. Importantly, please note the precipitous decline in year 3 and 4 of the external control group, as these boys are now in the steep part of the decline phase of their disease, while the treated patients remain stable. This graph represents the treatment effect of a disease-modifying agent. To further support this point, next, we'll share the functional results from 20 patients who have received SRP-9001 in Part 1 of Study 102 and now followed for 2 years. As a reminder, Study 102 is a double-blind 1-to-1 randomized placebo-controlled clinical trial of SRP-9001 in 41 participants to Duchenne between the ages of 4 to 7. We again analyzed the propensity score weighting method to compare NSAA data for Study 102, Part 1 participants to an external control group. Once again, the groups are very well matched. And at 1 year, as noted on this graph, in Study 102, Part 1, there's a 3-point median difference between the SRP-9001 group and the external control group. At week 96, this grew to a 5-point median NSAA difference, with a p-value of 0.0001. It's also important to note that an influential outlier with a 17-point NSAA decrease was observed in these data, which skewed the mean estimate and therefore, the comparison of the medians are more appropriate and quantile regressions applied using the same for propensity score weight to test the equality of the mean. It's important to note this patient would have been excluded by 301 inclusion-exclusion criteria. The fact that only approximately half of the patients in the treated group received the target dose, make these results even more impressive. Now speaking of target dose, we will now move to the integrated efficacy analyses for all patients in Studies 101, 102 and 103, who received the target dose of 1.33E14 vector genomes per kilogram compared to an external control. When you pull all of the patients to receive the target dose from these studies, from 101, 102 and 103, we have a robust pool of 52 patients in the corresponding analysis. In terms of the baseline covariance, unsurprisingly, once again, the groups are very well matched at baseline. The integrated analysis of 1-year functional data from patients who received the target dose of SRP-9001, Studies 101, 102 and 103, the treated patients improved 2.4 points in NSAA total scores from baseline. When compared to the propensity weighted external control group, NSAA changed from baseline 1 year after treatment for the treated patients who's 3.1 points higher on unadjusted means and 2.4 points higher using least square means, with a highly significant p-value of 0.0001. These data firmly reinforced the consistency of NSAA improvement now across 3 independent trials. Again, we also looked at key secondary functional end points for the time function test. Importantly, these data show mean improvements such as time to rise and 10-meter walk/run. Now to conclude this section, we are showing expression data across all of our studies. We are very pleased with the consistency of the results generated from both our clinical and commercial manufacturing process with Study 103. And finally, I will now share an update on safety. Our collective safety data shown here has been generated from all patients in SRP-9001-101, 102 and 103, for a total of 84 patients. The most common treatment-emergent adverse events was vomiting. We had one new treatment-related SAE myocarditis in an 11-year-old boy admitted to treat nausea and vomiting. While the patient had no symptoms or signs of systolic dysfunction, raised troponin was noted incidentally during his hospitalization. Function was preserved on echo and cardiac MRI, but MRI findings were consistent with myocarditis superimposed on Duchenne cardiomyopathy. The patient received 3 days of IV methyl-pred. Both events had additional chronic cardiac medications added, and cardiac MRI at 1 month showed normal function and partial resolution of myocarditis changes. An echo at 4 months showed normal systolic function. We have now dosed the largest number of Duchenne patients and have dosed patients over 80 kilograms. We continue to see no clinically relevant complement activation, and we remain pleased with the strength and consistency of SRP-9001 safety profile. We will end with a summary of today's data and next steps. These new data and our integrated efficacy analysis has demonstrated that SRP-9001 performs well above what natural history will project, giving us increased confidence in the probability of success for EMBARK, our global Phase III pivotal trial. We remain on track to complete enrollment in the EMBARK study in the middle of this year. And finally, we remain in discussions with FDA regarding the potential accelerated approval submission. We will provide an update when the discussions conclude. Thank you very much for your attention today. I will now ask the operator to open the call for Q&A.

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

Across the totality of the data across the various studies on NSAA, I know you put some benefits into context. Are there any trends on which domains are improving within the end points? And any differences there, age-wise or otherwise?

Louise?

Thank you for that question. We certainly look across the domains. As you know, given the totality of this evidence and the 5 point changes, you're seeing a positive effect on multiple domains within the analysis. As we mentioned, you're showing improvements in multiple and then also gaining these effects. There's no particular domains that we see improvement and that we don't see in others. And so across the board, we're seeing improvement.

Our next question comes from Ritu Baral with Cowen.

I wanted to just ask about controlling for steroid use in the propensity-matched controls versus the dosing that's used around 9001 treatment. Did you look at what steroid dose the propensity-matched controls we're using and if the difference might impact the delta on NSAA? And just a very quick follow-up question from a client. The timing of the FDA discussions that you've had so far, about when were they? Were they very recent or were they earlier this year?

Yes. So I'll answer the last question, then I'll turn the steroid-related question over to Louise. We've had some discussions for -- we haven't given the exact timing, but we've had some productive discussions with the FDA. We've had an opportunity to share with the FDA all of the information that you have seen today. We wanted to make sure that the FDA had an opportunity to review that and digest it before we publicly disclose it, and that's why we chose this meeting to do that. As I've said before, we'll say it here, we've had some productive discussions with the FDA. We're continuing to be in dialogue with the FDA, can make no prediction on the outcome of that, but I am confident that we'll get to a conclusion in a relatively not-too-distant future. And then of course, we'll provide an update once we do. The exciting thing about this data, frankly, is across all of the various studies and then looking at the prior data that we've disclosed as well but really focusing on this data here, is that it certainly is helpful in discussions with the FDA regarding the potential for filing for accelerated approval, but also beyond that issue, just with respect to EMBARK. EMBARK was a very well-powered study. But these results only increase our conviction and, frankly, increase the powering of that study. So we're very excited about what we're seeing here. And we'll provide an update on those discussions with the FDA once we have a conclusion on them. And with that, Louise, perhaps you can touch on the steroid-related issue?

Sure. Regarding the steroid uses and those external controls, they were using standard of care dosing, which is consistent with what our patients are on. Now our patients in the first 60 days do get increased steroids. But at the time when we're looking at 1 year, they're all on standard of care well past that period of using those increased steroids, so they won't compound the treatment effect.

Our next question comes from Colin Bristow with UBS.

Congrats on the data. First, somewhat of a housekeeping question on patient numbers. I think there's 65 patients across cohort 1 for 103, and then Studies 102 and 101, in the pooled analysis, there were 52. I didn't quite see the footnotes on the slide. Could you just help me with that delta? And then just a quick one on the functional data for 103, can you talk about any differences or any meaningful differences in ages and baseline characteristics for the additional 9 patients we got today versus the original 11 that comprise the 24-week data we saw at the micro-dystrophin data.

Yes, Louise, can you take both of this?

Yes. So the first question around the numbers of patients. In the integrated analysis, we included all patients that were on target dose, and that's how we arrived at the 52 patients for the integrated analysis. And then the baseline characteristics for the 11 versus the 20 are similar. As you know that we had added more important 5-year-old [indiscernible] went to 20 patients. And so the mean ages slightly lower than the first 11, but relatively consistent with the first 11.

On the -- sorry, just on the target dose issue, what was the primary reason for patients not receiving the target dose?

Yes. So in Study 102, if you'll recall, we -- when we retrospectively titered those patients using our now validated methods, it was noted that there was some variability in the lot using the clinical material in Part 1 of that study, and that's why there was some patients that receive less than the target dose.

And just so I'm sure everyone knows this, but to the extent, the doubt -- so we started Study 102, Part 1, using clinical supply from Nationwide Children's Hospital, and we used the titering method that they had, which, at the time, was a supercoiled PCR method. Between Part 1 and Part 2, we were able to develop a much more precise titering method. So all of the kids that crossover received the linear titering method that we developed, and so they all had target dose. But the kids in Part 1, as Louise mentioned, 60% of those kids had less than a target dose. All of the kids in 103, of course, are not only at the target dose, but that's our commercial process material. That is the process we'll be using to launch this therapy, and that is the dose we would intend to launch this therapy at. So it's very telling that we're getting these results in that cohort of patients, and of course, a very robust cohort as there are 20 patients in Cohort 1 of 103.

Our next question comes from Kristen Kluska with Cantor.

Congrats on these data that you presented. Could you talk more about the importance of the earlier intervention here, especially that you had followed up with patients for over 4 years, and you've also looked at different patients across different age groups? And I understand that PPMD very recently submitted a nomination package to add DMD to the newborn screening panel?

Yes. Let me comment very briefly on it, and then Louise can talk a little bit more about the age-related issues and the lag. First, know this, we are working with our -- with PPMD and others to try to get newborn screening in place, and there's an enormous amount of value in newborn screening for Duchenne. As it relates to this particular data, the real point of this data, when you look at 1 year versus 2 years versus 3 years now versus 4 years, it's not -- there's a separate issue from when you intervene. The real issue is that we are seeing a strong benefit at only 48 weeks. But what that miss is that, that benefit is going to compound as kids have been on the therapy for a longer period of time because against natural history they're going to start diverging considerably. That was one of the things we had previously been saying that you reduce treatment effect. You see a 3-point delta in NSAA, which is brilliant at 1 year. It would be brilliant. But that doesn't describe the benefit. Because, of course, these kids now, if they're on a disease-modifying therapy and if SRP-9001 is anything and is effective at disease modifying, these kids are now diverging from natural history over time, and that's exactly what we see. You see on median at 2 years, you see a 5-point delta versus a propensity-matched natural history cohort. And then when you look at these 101 kids, you're out here at 4 years, and it's really as much about the time on therapy than the age they were dosed. These kids are 4 years. They're now over 9 years old on me. Anybody who knows Duchenne knows what that means for kids that have Duchenne muscular dystrophy. They are in that steep decline phase of Duchenne. And yet these kids are not right now. These kids have been rock solid stable for the last couple of years. There's 7 points above their baseline from 4 years ago, and they are almost 10 points above natural history using a very conservative propensity match approach. So the biggest point is that the benefits that we are seeing here confirmed, at least our hypothesis, that the benefits are going to continue to compound as these kids diverge from the natural course of this disease over the long history of Duchenne. But Louise, maybe you have some comments on sort of intervention times?

Yes. I think you catched a lot. I'll just add that. Regardless of what age patients have been treated with 9001, we've seen a similar treatment effect in comparison to well-matched natural history control. And so the age of treatment is not indicative or not mandatory in terms of seeing that benefit. And as newborn screening comes on board in the future, I think the opportunity to treat patients earlier is certainly something that we're supportive of and recommending that it's great for patients to have that opportunity to be treated earlier as the future comes on board.

And the full experience -- to really fully understand the experience in both expression and safety and ultimately, a function across age groups, people should know that we have dosed much older children and young men, I mean those young men that are nearly 20 years old. We've dosed very heavy individuals over 80 kilograms. And we've also dosed very young. We've actually dosed down now to 3 years already. So we're really building out a very comprehensive set of experience on both expression and safety and ultimately on function with 9001.

Our next question comes from Tazeen Ahmad with Bank of America.

Just to follow up on your comments about not clinically relevant incidents of complement activation. Can you just give us a little bit more color on what the threshold is for you to determine what's clinically relevant? And then can you just give us some color on the new incidence of myocarditis that was observed in the study?

Sure. Louise?

Yes. In terms -- we haven't seen complement activation. So we haven't seen any clinical manifestations or things like TMA that have been seen in other studies. So we just haven't seen any clinical signs of complement activation. In terms of the patient with myocarditis, this was an 11-year-old boy who was treated in the second cohort of 103. This is the older ambulant cohort. He was, in the first week, which is common, had nausea and vomiting and was admitted for IV. And during that stay in the hospital, he -- it was noted that he had an elevated troponin, but he had no symptoms or signs and his cardiac function was normal. So out of caution, he received 3 days of methylprednisolone. And now at 4 months has returned or normal function -- continues to have normal on echo. To this point, it was observed incidentally, but clearly something that we do already monitor for troponin, and we'll continue to monitor for in the future.

And just one thing to add, Tazeen. The way that we're describing clinical complement activation is the exact same way that we've done previous disclosures. There's absolutely no change to how we're describing it now.

Yes. This should avoid any doubt. We see no clinical complement activation. We never had -- that is not an issue thus far with our therapy, even though we have dosed now well over 100 patients. We dosed the largest patients, I think, in Duchenne. We've dosed the youngest patients in Duchenne. We've dosed the oldest patients and the lightest patients. We just have not seen in this yet clinical complement activation or TMA or any of the kidney issues that others have observed with other vectors.

Our next question comes from Brian Abrahams with RBC.

Congrats on the updated data. Maybe just a follow-up on the patient experience of myocarditis. I'm curious, what do we know about their mutation? Is this one of the 0 dystrophin mutations that had been associated with myocarditis from other programs and if not immune related, any mechanistic explanation there? And then I'm also wondering the degree of FDA and EMA awareness of that case, and whether or not any additional monitoring, be it more troponin draws or imaging might be -- need to be implemented for the ongoing Phase III.

Louise?

Yes. Troponin elevations can commonly be seen in patient. We have seen them at pre-dosing in some patients. So this patient does not have a mutation that would be subject to immune-mediated myositis. And so that's not something that concerned about in this case. We already monitor troponin very frequently, and so that's something that we'll continue to do. And certainly, that this information on these cases have been shared with regulatory agencies.

Our next question comes from Gena Wang with Barclays.

Also congrats on the data. So one is a clarification question. For Study 102, you have 28 patient analysis with correct dose. The placebo arm is 21 patient crossover. So the remaining 7 patients from active drug actually received the correct dose. And my second question is regarding your start of the control on external control arm. I wanted to understand when you have a multiple-year follow-up, do you match with the beginning patient baseline as a control and a follow as the external control for full year data? Or do you match when, say, 1-year follow-up, when you have a 2-year follow-up, you match with the 2-year age to the external natural history as your control?

Louise?

Yes. So for the second question, we matched from baseline for the propensity-matched controls. Yes. So they're not rebaselines year-over-year. It occurs at the true baseline. And then the question on Study 102, the -- Part 1, as we said, 60% did not receive the target dose in Part 1. And then for Part 2, all patients received the target dose.

Our next question comes from Tim Lugo with William Blair.

Congratulations on the data, which continues to improve. For, I guess, a question on the nausea and vomiting patients. It looks to be about 40% to 50% of these patients could experience some of it. Can you give us some color on how many of those patients eventually require IV fluids and hospitalization? And then for the myocarditis patient, would that patient ever have been picked up if it wasn't for the hospitalization?

Yes. So the nausea and vomiting is common. We have not seen that was the only SAE in this case that has had required hospitalization for the nausea and vomiting. So it's not a common occurrence. Usually, it can be controlled with antiemetics and patients do fine with that. In this case, troponin is measured. This was an off-cycle time point, but it would have been potentially caught by the next time line in terms of the troponin elevation, with the current existing monitoring within the protocol. So it's first identified during this time, but would potentially be caught by the next time point. Again, the patient, this was incidental and he does not have any [indiscernible] for troponin increase.

Our next question comes from Yun Zhong with BTIG.

So I think on your discussion with the FDA on the accelerated approval pathway, outside the functional data that you presented this morning, do you think you would need or do you plan to submit anything specifically to support a correlation between micro-dystrophin expression? I assume that, that will be the surrogate end point and clinical benefit. And also on using the external control, there was a previous question on the steroid use, but I wonder, can you remind us when the external control data were collected? And anything that can potentially may have improved over the years in terms of standard of care outside that would use might affect the validity of the comparison, please?

Yes. So I'm going to touch briefly on the second part, but then I'll let Louise really dig into it. But just so we're clear, those are the kinds of things that one has to be careful about and make sure we're balancing for us. So there is a cutoff date on the types of information we can use. So that we're not going back so far in time that literally standards of care could have changed in some way that confounds the data. I'll let Louise touch on that in more depth than that. On the first part of the question about discussions, the short answer is we've had very robust discussions with the agency, and we've had a significant amount of analytics on the data that we have, expression, safety, associations between various measures and functions. And we're in active dialogue with the agency. And once we have an answer one way or the other on the filing of the BLA, short of EMBARK, then we will certainly discuss that and we'll let everyone know. Until such time as that could occur, and we should all remind ourselves that our current pathway to approval would be EMBARK, and we're very excited about EMBARK. We're obviously dosing EMBARK, even as we speak. We're on the readout in the middle of next year on EMBARK. And of course, while we were very confident in the powering of EMBARK, before this data, you can only imagine how confident we are today on EMBARK. So we're excited about the -- any of the pathways, but obviously, the pathway that could bring this therapy to patients the quickest is our [indiscernible]. With that said, Louise, perhaps you want to touch a bit on the external control issues that we -- what was that?

I'd have much to add in the fact that when we derived to the external control data sets, obviously, we looked for rigorously controlled trials that used standard of care dosing. We certainly wouldn't review something that wasn't method or inconsistent with what it's been now. Obviously, we will evaluate that looking forward. As I mentioned, the early surrogate in the trial does not, at the time point, does not affect the time point at what year where we're looking at these functional changes. And so we don't have any concern about that compounding the data in any respect.

The other thing we should say about the use of external controls and the way we're doing it is this is a very rigorous, but also a very conservative approach. The alternative to this approach would have been essentially a post hoc analysis where you grab an external control and you try to match them on 1 or 2 covariates. We've gone much more sophisticated in that as -- number one, it was prespecified. So this is all prespecified analytics. It wasn't post hoc. Number two, on the way we're doing this, that we use a regression analysis, which, in some regards, arguably dampen to the overt effect but also has the benefit of being very rigorous and, in fact, at least 2 therapies in the not too distant past, have been approved on the basis of this prespecified propensity match analysis and regression analysis that we're using here. So I think we feel very good about the approach that we've taken. And we think that it has an enormous amount of rigor with it.

[Operator Instructions] Our next question comes from Judah Frommer with Credit Suisse.

I was hoping you could just circle back and maybe provide a little more context on the 2 drugs you mentioned that were approved on propensity match controls. Anything you'd call out in terms of differences or similarities maybe in level of unmet need patient population? Anything that you could use as kind of a guidepost for potential accelerated approval here?

I'll let Louise touch on that specifically since she knows the names better than I do of those 2 drugs. I will say on the concept of unmet need, it's very difficult to get a greater unmet need than in Duchenne muscular dystrophy, ferociously degenerative disease, which is inevitable in its fatality. But with that said, Louise, is there anything else you want to say about the particular therapies?

Now, just in both of these are for rare diseases and patients in pediatric and adult populations in terms of the rarity, not more rare than Duchenne and no more unmet need than ours. So these are both reviewed, and those approaches were accepted by EMA and FDA. So it's certainly a guidepost. There are some differences. But in terms of the unmet need, they don't surpass those of Duchenne certainly.

And probably more than anything else, the point of mentioning the other therapies that have been approved versus this propensity-matched approach, it isn't that our therapies are exactly the same or our populations are identical. It's to say that this is a very rigorous and sophisticated approach that creates, in essence, essentially a synthetic placebo and provides a lot more rigor than one might use, for instance, with a simple post hoc analysis where you grab some patients and have a natural history cohort. So I mean the point here is that the FDA and I think EMA as well has embraced the concept of propensity-matched external control at times.

Our next question comes from Joe Schwartz with SVB Securities.

This is Beth on for Joe. I was just wondering if you've been able to glean any insights on how micro-dystrophin expression may correlate to NSAA benefit based on the 12-week biopsy samples you've taken? Also wondering if you've taken any longer-term biopsy samples to understand how the micro-dystrophin expression may play out at longer time points?

Louise?

Yes, so great question. So certainly, we look at the population level at micro-dystrophin expression, 9001 dystrophin expression versus NSAA. And it's clear from the state of today that the expression is driving those functional results. We will be -- we've looked at longer time points in the study and other studies, and then that will be shared in the future as we publish data sets moving forward, specifically on the 102 data. What we know about durability from those preclinical studies and clinical studies is that AAV in particular is very durable and the results we've seen so far are very consistent with those results. So we're happy with that.

I would now like to turn it to Ian for an e-mail question.

Sure. So unfortunately, Gavin from ISI couldn't log on to the call. So I'm just going to ask a question that he asked, Louise. So he asked, in Study 103, do we see a difference in NSAA improvement from baseline in the 4- to 5-year-old group versus the 6- to 7-year-old subset?

Yes. Thank you for the question. We actually didn't do that analysis as you've seen, we see an effect, regardless of age, in 103. So we're seeing an effect size in comparison to the external control that's robust regardless of the age. We did that with some previous analysis in 102 that was compounded by some baseline imbalances. But as you can see from this data, we're seeing robust results regardless of the baseline age in this group.

Our next question comes from Ritu Baral with Cowen.

Louise, I wanted to just round back to the myocarditis patients. Can you tell us how long it took for that patient to biochemically resolve and for his troponins to come back to, I guess, normal levels? And then on follow-up, did function return to baseline levels in this patient?

So the troponin elevation resolved quickly and the function remains normal. So as mentioned at 1 month and 1 year, there was normal systolic function. So the function actually never was adverse in terms of this case. There was some myocarditic changes on MRI, but the function had already remained normal throughout the event.

Our next question comes from Danielle Brill with Raymond James.

I just had a follow-up as well on the myocarditis patient, sorry if I missed this before. But did that patient have a mutation in exons 1 to 17? And then a quick follow-up on EMBARK. What is the status of enrollment? When do we expect that to be completed?

Yes. I'll answer the first one, Louise, and correct me if I'm wrong. The answer on the first one is the child wasn't within the 1 through 17, and this wasn't any innate immune response or anything like that. And then on EMBARK, we've said we're going to complete dosing around the middle of the year. We're on track for that. And so things are going quite well there. We'll provide an update to folks once dosing is complete.

Our next question comes from Eka Gigauri with OpCo.

I'm Eka dialing for Hartaj today. Congratulations on the data. Our question is about manufacturing. So with these data in hand and interactions with the FDA that are ongoing as well, could you comment on your commercial readiness and manufacturing scale-up to potentially meet the demand if you were able to file for an accelerated approval?

Yes. Thank you for that. We will be ready. So we are -- we scenario plan all the time. And in the event that we could get -- as everyone knows, our base case is that EMBARK will be the pathway for our approval, and that remains the case, unless and until we're able to say something different than that. But we scenario plan so that we would be prepared to serve the community upon launch, either after EMBARK or earlier if that was possible. So the one thing I will say is, obviously, we have a lot of inventory to build, but we'll do that, and we'll be in good shape.

We'll now have Ian read an e-mail question.

Ian, you're on mute.

Apologies for that. Salveen had trouble getting on. Salveen from Goldman Sachs. So Louise, the question was, for the integrated data, how do we reconcile the commercial versus the clinical product in the analysis? And then her second question is, has the FDA provided guidance on how to determine propensity matching?

Good question for the commercial versus clinical material. What's the beauty of this is that the 9001 dystrophin is the same in our clinical and commercial material. And as you can see, we saw consistent results in terms of expression. In addition to that, as you also saw, we performed the integrated analysis, but we also performed individual analysis with each study. And so we saw a consistent effect across those studies, and then when we integrate the data together. In terms of the guidance on propensity matching, there's multiple guidelines available that we complied with. Some of these include the FDA E10 guidance will be adaptive design for clinical trial guidance as well as DMD and rare disease guidance. So that's just to name a few. So we certainly comply with those in terms of designing the propensity score matching.

Yes. And just a follow-up from Louise. So Salveen, just so you know the clinical commercial process, the clinical material was used in Studies 101 and 102, and then the commercial process material was used in 103. So like what we said, we can easily break it down that way.

Our next question comes from Gena Wang with Barclays.

[indiscernible]

Yes. Apologize, Gena. I don't -- unless I'm alone, I can't hear the question. Apologies for that.

Can you hear me now?

Yes. Yes.

Can you hear me? Okay. Okay. Perfect. The question is that we accelerated [indiscernible] to submit another approval or another alternative is the rolling [indiscernible] the time FDA will need to make a decision is very close to your Phase III trial EMBARK readout? Do you think regardless will pass if it was due, would be [indiscernible] back to [indiscernible]?

I apologize, Gena. I didn't fully get the question. I'm going to try to deduce what the question was, which is I'm not -- what I think you were asking, and apologies for not getting it fully, was that assuming that we got an answer from the FDA that allowed us to submit a BLA, then the review time for that BLA might be -- might run close to the readout for EMBARK. I think that might be what you're asking about? And the short answer is that if we could file a BLA for accelerated approval, it would provide an enormous opportunity to get this therapy to patients as quickly as possible. And we'll see what the FDA's perspective on that is. And then once we know, we'll let you all know. But -- the exciting thing about the data today is while we were very pleased to be able to provide it to the agency and to have discussions -- productive discussions with them, we're also very excited about what this means for the powering of EMBARK and the probability of EMBARK, which is our base case. And again, I apologize, Gena, if I completely misunderstood your question. I'm sorry about that.

We have another e-mail question. Ian?

I'm sorry, I don't see another e-mail question.

This question is from Brian Skorney from Baird.

Okay. So the question is [indiscernible] presented 3-year data from Study 101 last year and included a percent meter, 100-meter walk. I think patients were generally in the range of about 50% to 70% predicted, and it was noted that natural history would be about 44% in 3 years. I think natural history comparisons are really helpful, but I also find comparison to non-DMD subjects are helpful, too. Can you provide any color on where these patients are sitting relative percent predicted at 4 years? Just trying to get a handle on not just how to differentiate these patients are comparing to DMD, but if they're also getting closer to non-DMD controls?

Yes. Thank you for that question. So essentially, the percent predicted is the correlate of what we showed in terms of -- we showed a decline in the amount of time that it takes to do that activity. So the percent predicted to sort of the inverse of that. And so you're right, we've seen a consistent improvement in the percent predicted. We don't have the exact number on comparison to natural history or certainly something that we could do. But to your point, these patients are definitely more approaching a normal control than the natural history of DMD, but I don't have the exact numbers for you today.

I mean, these kids are -- the 34 points is the maximum of the scale. I think the normal range in these kids sort of varies from, what, 31 to 34, Louise? And we've got kids over 30. And so obviously, they're doing well.

Our next question comes from Brian Abrahams with RBC.

Do you expect a clear answer from the FDA on whether the data you have in hand would support accelerated approval? Or are your discussions centered more around exactly what types of data and analysis you might need to provide in a filing and that you plan to file regardless and the decision would really come down to being a review issue as to whether or not this could be approved on an accelerated basis?

In the end, I mean, I think everything is a review issue once you file a BLA, we're not going to get an answer in advance. It's definitive, but we will get a clear view from the agency and I suspect on whether it would be wiser for us to file for a BLA for accelerated approval now or whether we should await EMBARK. And obviously, we'll be guided by that. We are not going to file a BLA unless we have an enormous amount of conviction that we're going to have a very positive review of that BLA. So I think we will get -- I think we get a relatively definitive view on whether it makes sense to file a BLA or whether we should wait for EMBARK.

Our next question comes from Chang Liu with Needham.

This is Chang for Gil. So will there be another integrated analysis 2-year mark? And if there is, what should we expect from the natural history? It seems intuitive that the trend should go down in the second year, but the external control in 101 is actually increasing?

Louise, I don't know if we have a planned integrated analysis for year 2. The integrated analysis that we did that you saw today was really in support of our discussions with the FDA about the full data set. Obviously, the thing we're most excited about next from a data perspective is the readout EMBARK, which will happen around the middle of next year. Louise, do I have any -- incorrect about what I just said?

No. All I would add is that our primary end point for the study was at 1 year and our target is to the integrated analysis at 1 year. We may choose to do additional analysis. We look forward that but the primary end point was 1 year.

There are no further questions. I'd like to turn the call back over to Doug Ingram for any closing remarks.

Thank you, Michelle. Thank you all for joining us today. If you don't mind me extemporaneously congratulating my colleague, Dr. Louise Rodino-Klapac. I would remind everyone that Louise is not simply our Head of R&D and Chief Scientific Officer, but she, along with her colleague, Dr. Jerry Mendell, designed this therapy and painstakingly tested it and optimized it over many years. So this has been an enormously important day for Duchenne, and I think it's an important day for Louise as well and Sarepta. Congratulations on that, Louise. We look forward to providing additional updates across the year as we achieve milestones. As it relates specifically to SRP-9001, this includes the outcome of our discussions with the FDA. So we'll certainly provide you an update on that and also on the completion of dosing for EMBARK, which we will update on that as well. And until then, I would ask everyone to have a good day and be well.

This concludes the conference call. You may now disconnect.