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

Good afternoon, ladies and gentlemen, and welcome to the Sarepta Therapeutics' Micro-dystrophin SRP-9001-102 Top Line Clinical Data (Part One) Conference Call and Webcast. [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 CEO, for opening remarks. Please go ahead.

Thank you very much. Before I begin, let's move to the next slide, please. I would just remind everyone that we will be, of course, very likely making forward-looking statements today. And I would ask folks to look to our public filings and SEC filings for the various risks and uncertainties that are intended when one makes forward-looking statements about the potential issues that'll occur in the future. And with that, let's go to the next slide. Good afternoon. Thank you for joining all of us today for our investor conference call to review the top line results from our Part 1 interim analysis of Study 102 evaluating our gene therapy SRP-9001 in patients with Duchenne muscular dystrophy. By now, you will have seen our press release announcing the results. In a moment, I will turn the call over to our Chief Scientific Officer, Dr. Louise Rodino-Klapac, to review those results. But let me start by making a few introductory comments. You will hear today that Study 102 has largely confirmed our hypotheses on the performance of SRP-9001. We strongly hit statistical significance in all biological measures, including expression, genome copies, localization and CK levels. We continue to see a favorable safety profile with no new safety signals, and it is clear that the therapy is providing a substantial benefit to patients. But with that said, it is an understatement to say that I was initially not only surprised but deeply disappointed that while we saw separation between active and placebo, at every time point, we missed statistical significance on our primary functional analysis for SRP-9001. But upon review, the likely reason for this became immediately apparent. While you should view these analyses as tentative until we conduct a deeper review, it seems evident that our miss on statistical significance was not a problem with the overall design of the study, neither the study size nor the end point nor the time line nor our assumptions in how the DMD patients performed in the chosen 4 to 5 and 6 to 7 age groups. And importantly, we do not believe it relates to therapy performance. On its face, the lack of statistical significance appears to relate to improbable, bad luck in the randomization process for the 6- to 7-year-old cohorts. Let me explain. As we've talked about often, as an organization that understands the disease course of Duchenne, we identified that there would be a difference in the predicted trajectory of the 4 to 5 age group and the 6 to 7 age group. So we took many steps to ensure we accounted for this. Indeed, age was the stratification variable we used for randomization. And we stratified for 4 to 5s and 6 to 7 age groups, and we set up separate prespecified statistical analyses for the separate groups. And we further worked to reduce variability by placing a ceiling and a floor on the entrance criteria. With respect to the 4 to 5 age range, our plan worked. In the randomization process, the active and the placebo groups had nearly identical baseline characteristics. And thus, you will also see that the active therapy cohort did exceptionally well versus the placebo cohort, and we robustly hit statistical significance, a p-value of 0.017. Now some have voiced concerns that the NMR study might have been too modest. But I would point out that we achieved robust statistical significance with only 16 participants, achieving NSAA numbers that are very similar to what was predicted by our prior proof-of-concept study, 101. But as impressive as the statistical significance was for the performance of the 4- to 5-year-olds on therapy, unfortunately, there was an even more impressive statistically significance between group difference in this trial. Through the vagaries of the randomization process, the baseline characteristics for the 6- to 7-year-old placebo cohort on NSAA were very substantially different and importantly far milder than the baseline characteristics of the 6- to 7-year-old active cohort, which were more severe. Indeed, the difference has a p-value of 0.004, meaning that random chance of this occurring was about 1 in 250. As Dr. Rodino-Klapac will show you in her presentation, this imbalance occurred in favor of the placebo group in each and every baseline functional characteristic. This means that for the 6- to 7-year-olds, we had 2 very different and noncomparable treatment populations: The active group, by random chance at the bottom of the entrance criteria, would be predicted to be moving into the potentially significant decline phase of the disease; while the placebo group, at the ceiling of the entrance criteria, would be predicted to be in a much milder phase of the disease. And so when the baseline characteristics were well matched in the 4- to 5-year-olds, the therapy performed as predicted, and we robustly hit statistical significance and a clinically meaningful benefit over placebo. But where the baseline characteristics between the active and placebo in 6- and 7-year-olds resulted in materially noncomparable treatment populations, we could not show a between-group difference on function. And this is the reason we missed the overall statistical significance on the primary functional measure. We are very frustrated by this randomization issue. But having now treated over 50 patients, the fact remains that the results of this study further bolster our confidence in the transformative potential of SRP-9001. And we owe it to the patients that we serve not to be taken off track by a random error, but instead to take the wealth of information and insight that we have from Study 102, inform our program going forward, refine our protocol for the next trial and continue to advance SRP-9001 with a sense of urgency, which is exactly what we will do. We will learn from this trial. We will keep moving forward to gather the evidence necessary to bring these therapies to patients that are waiting around the world. And with that, let me turn this call over to Dr. Louise Rodino-Klapac. Dr. Klapac?

Thank you, Doug. I'd like to start out by reminding you of our study design for our SRP-9001-102 study. It's important to note that we're reporting results from Part 1 of the study. The study is ongoing and remains blinded. Because of that, we will be providing top line data and will not be sharing individual patient-level data. There's additional data that we cannot share today, but we will in the future. Regarding the study design. It's a 41-patient, double-blind, randomized, controlled trial. As Doug previously mentioned, the randomization was based on the age stratification between 4- and 5-year-olds and 6- and 7-year-olds, and that was later used as a prespecified subgroup analysis. In Part 1 of the study, subjects were randomized 1:1. And half of the subjects received SRP-9001, whereas the other half received placebo. Following Part 1 at 48 weeks, subjects were crossed over to either placebo or SRP-9001. So today, we'll be presenting 12 weeks biopsy data from Part 1 as well as 48 weeks functional data. Now it's important to note that we'll be receiving additional data in the near future, which will include biopsy data from Part 2, which will be 12 weeks post-crossover. The study will remain achieving -- or monitoring for safety and functional outcomes for an additional 48 weeks. And then following Part 2, subjects will cross over into an open-label extension study where we'll be gathering additional safety data. In terms of the primary end points, primary end point for expression was at 12 weeks, as measured by Western blot, and then change in NSAA total score from baseline to week 48; for the secondary end points, which included micro-dystrophin expression by immunofluorescence and percent positive fibers and other timed function tests. So now turning to the expression. We met the primary end point showing significant micro-dystrophin expression at week 12 by Western blot with a mean of 28.1%; in terms of looking at immunofluorescence labeling, a mean intensity, this is the amount of micro-dystrophin localized at the membrane, for a mean of 63.7% and also a mean of 33% dystrophin-positive fibers. In terms of vector genome copy number, we had a mean of 1.56. We did see a range of expressions. As you will recall, we use clinical material from Nationwide Children's Hospital for this trial. At the start of the study, we used Nationwide's method for titering, and this is a qPCR assay with the supercoiled plasmid standard. We then developed a robust titering method at Sarepta that is more accurate, precise and well-controlled using a linear standard. When we went back and we retrospectively titered the clinical lots used, there was some lot-to-lot variability due to the previous method used at Nationwide. Preclinical studies did not predict this variability would lead to differences. The dose used in Study 101 was 1.33x10^14 vector genomes per kilogram using the linear standard, and all lots are now titered using this method. And the dose will be consistent through this for the remainder of the 102 study. Thus, we will not see variability in Part 2 or other studies moving forward. And we will predict to see an overall higher mean for expression. But this will have to be confirmed. On the next slide, we are showing representative image of our micro-dystrophin expressions. It's labeled by immunofluorescence. And so what you're seeing at the top, this is a baseline biopsy, we co-labeled all of our biopsies using merosin, which is a marker for the membrane and then stained for dystrophin. So at the top, we see a biopsy labeled -- labeling the membrane with merosin. But for micro-dystrophin, we see nothing. So there's no expression at baseline. On the bottom panel, we see, again, membrane labeling and then robust micro-dystrophin labeling, as you can see in red. As I mentioned, this correctly localizes the membrane. And you can see that a merged image are clearly -- they're co-localizing with the membrane marker as well. So now turning to function. Next slide, please. NSAA, or the North Star Ambulatory Assessment, was the primary functional outcome measure. To remind you, this is a 34-point scale, including 17 different primary motor function measures. Here, what you'll see is the NSAA plotted from baseline to 48 weeks comparing SRP-9001 treatment, you can see by the red dotted line, compared to placebo in the blue solid line. And what you'll note is that at every time point post-baseline, there was an improvement in the NSAA in the SRP-9001-treated group versus placebo. However, it did not reach statistical significance at 48 weeks. But if you look just at the 9001-treated group from baseline to 48 weeks, we did see a statistically significant improvement. The treatment group showed a 1.7-point increase compared to baseline for a p-value of 0.009. And when you look at the placebo group, we saw a 0.9 increase compared to baseline, and that was not significant. So now let me turn to the next slide for the prespecified subgroup analysis in 4- to 5-year-olds. You will see a clear picture. You'll see that comparing baseline to 48 weeks, there is separation between the 9001-treated patients and placebo patients at every single time point, including 48 weeks where we see statistical significance. So just to orient you on the graph, the placebo is shown on the bottom line. SRP is the gold dash line on the top. And so what's important is that there was a 4.3-point improvement in the NSAA from the SRP-9001 group from baseline and a 1.9 change in placebo, and that resulted in a delta of a 2.5-point difference between the group. And this was significant, where p was equal to 0.0172. Now we do not expect [ management ] to see the same result in the 6- to 7-year-old subgroup, and the next 2 slides I'll go through will address why. So next slide. So here, we're looking at the baseline functional characteristics of the 4- to 5-year-old group. You will note that the placebos were very well matched at baseline, and all measures -- in all measures, there were no significance between the group. So let me just take a minute to orient you on these measures. For NSAA, again, this is a 34-point scale. So the higher the number, the better the individual is doing. Whereas all the time tests that you're seeing, 100-meter, 4 stairs, time to rise and 10-meter, you're looking for a reduction in the amount of time. And so patients that have a lower number are doing better because they're taking less time to do some activity. But in this case, in the 4- to 5-year-old group, you're seeing that these patients are all very well matched between 9001 group and the placebo group. You see no significant difference between the groups. Now when we look into the 6- to 7-year-old group on the next slide, we see a much different picture. In this case, at baseline, they're highly favoring placebo in all measures. As Doug mentioned, there's a significant imbalance between the 2 groups at baseline, except for one. The difference between the 2 groups is significant. So let's take a little bit of time to walk through this in detail. So focusing on NSAA. You'll see that the SRP-9001 group has a mean baseline of 19.6, whereas the placebo has a mean of 24. That's a 4.4 difference between those 2 groups. For all of the time tests you see, the placebos have a significant reduction in the amount of time it takes to do that activity. So they're doing much better at baseline compared to the 9001 group. Again, these are all significant with the exception of the 4 stair climb. So in summary, these baseline characteristics were highly imbalanced between the 2 groups and very likely contributed to the lack of statistical significance that we saw in the 6- to 7-year-old group. If we just focus on 6 to 7s, in particular, the micro-dystrophin-treated group, we know from natural history that the 6- to 7-year-old age groups, as I've mentioned, are generally declining. And those with the low baseline function, specifically in the 9001 group, would be significantly declining over a year, and yet we only saw a very small point for a decline. So it's difficult to demonstrate a treatment effect with the significant imbalance between these 2 groups. So we'll be looking at this in greater detail. But at this point, the imbalance between the 2 groups was quite clear. Now on the next slide, we're going to turn to safety. We continue to see a very favorable safety profile with no new safety signals identified. The most common adverse event was vomiting, and the majority of AEs were mild to moderate. There were 4 patients with a total of 5 SAEs. Three patients were in the treated group and one in the placebo group. There were 3 cases of rhabdomyolysis, in which case, 2 were in the treated group and 1 in placebo, and 2 cases of transaminase elevations. There were no adverse event-related discontinuations and no deaths in the study. And importantly, no clinical complement activation was observed. So on the next slide, let me take you through the next steps. So we're going to continue to advance the Part 2 crossover phase. As I mentioned, we'll be collecting biopsy data at 12 weeks to assess expression and other biological markers. And then we'll be assessing longer-term assessments of functional measures, including the NSAA for -- of the Part 1 patients that received SRP-9001 as well as the placebo crossovers. Importantly, in our 103 study, this is our open-label commercial process material study, we have enrolled and dosed 11 patients, and we're expected to have biomarker and safety results from that study in Q2 of this year, 2021. This will be leveraging the learnings from both Study 102 and Study 103 to inform our future clinical development plan, including a Phase III. So next slide. In conclusion, we've identified importantly that there's no new safety signals that have been observed. We're pleased that we met the primary biological end point of micro-dystrophin expression at 12 weeks. Our total NSAA score of treated patients versus placebo demonstrated a positive increase for every single time point, but yet did not read significant -- statistical significance versus placebo at 48 weeks. Importantly, we had a prespecified analysis in the 4- to 5-year-old age group that showed a significant improvement in NSAA versus placebo at 48 weeks. And the imbalance in baseline functional characteristics in the 6- to 7-year-old group contributed to the lack of statistical significance on the functional end point. We've collected a wealth of data in this study, and we'll -- that will continue to support future clinical development plans. We still have data to come in this study, and this will guide our development as we move forward with our 9001 program. And with that, I'm going to turn it back to Doug for questions.

Thank you. Operator, let's open the lines for Q&A.

[Operator Instructions] Our first question comes from Brian Abrahams with RBC Capital Markets.

I just wanted to expand a little bit more on this idea of the imbalance in baseline function. I guess how much do we know from historical natural history data about the degree to which baseline function should predict the trajectory of 1-year progression? And if you were to adjust or match the baseline NSAA within the 6- to 7-year-old cohort, would there have been a clearer delta favoring the drug?

Louise?

Sure. We do know from natural history that in the 6- to 7-year-old age group, we would typically see a decline in this group from 0 to 4 points. And also, the baseline does predict -- the particular baseline does predict the trajectory to some extent. So in this case, we really -- it was clear that the significant imbalance made it virtually impossible to see a treatment difference in this group. Gilmore, would you like to add to that?

Yes. No. Yes, thanks very much, Louise. Thanks, Brian, for your question. I think Louise has summarized what we would predict. I think your second part of your question is around the sensitivity analyses, and we're obviously doing further analyses. This is the top line data that we're giving -- getting to you fresh, but we have a lot of other prespecified and obviously post hoc analyses that are ongoing.

Our next question comes from the line of Tyler Van Buren with Piper Sandler.

I guess my question is, is there any chance that you go to the regulators and see if there's some sort of filing that could occur based upon the prespecified 4- to 5-year age group? Or without a doubt, based upon the data, do you think you need to generate more data in Study 103? And what type of data do you think you need to generate to eventually file?

Yes. Thank you for that question. First, of course, we have had no additional discussions with any of the regulators about this. I think as it stands right now, we are going to focus on the completion of Study 102. Remember, this is, in fact, an interim analysis. This trial remains blinded. One of the reasons we're not going into the depth that you might otherwise expect at the end of a trial on patient-level data is that we have an ongoing trial, and we'll have a readout on that. And it will be interesting because that readout of the study, which will occur when all of the kids have had 48 weeks, including the crossover participants, will be a different sort of analysis than this context of looking at the therapy on active versus placebo at 48 weeks. We'll do trajectory analysis and a number of other analytics that are prespecified there. We're also focusing on Study 103 and getting the expression level information out of Study 103 to see the performance of our commercial material. That will be coming up in the not-too-distant future, which will tell us a lot about the therapy. And then, of course, very importantly, right now, what we are focusing on right now is to take the learnings out of Study 102. We've seen a lot out of this study. And again, I'm not going to ignore the fact that I'm extraordinarily disappointed that this imbalance caused us to miss stat sig. But inside of this study, not only are we more confident than ever in the performance of this therapy and the ability for this therapy to have a transformative impact on the lives of patients, but we have a wealth of information out of Study 102 that no one has ever received in a trial such as this with Duchenne muscular dystrophy. We have dosed over 50 patients already. We're going to take all of that learning and all of that information at a granular level and ensure that the Study 301 that we're going to launch this year is going to maximally have the best chance of not only resolving this issue but having the strongest probability of success. And that's really what we're focused on right now.

Our next question comes from the line of Alethia Young with Cantor.

I hope you guys can kind of work through something for at least the kids. I guess I just want -- I looked at -- quickly at that NSAA picture you put out by time point. In between 12 and 24 weeks, it seemed like in the placebo, there was a huge jump up with wide error bars. I mean, can you just talk a little bit about it? And then after that, it seems like you might have hit, if it's been 12 weeks. So I just wanted to get your perspective on that.

Yes. I'll just say real briefly on that 24-week time point, you'll see the strange increase or a momentary increase for the placebo group. We can't say a ton about it yet because we haven't done all of the analysis and analytics. What one can point out there is that, that was in the middle of the most difficult period of the pandemic. And if you look at the numbers below the graph, you will immediately see that there were more out-of-window and missing time points. So that may very well just be an artifact of the pandemic. We don't believe it actually affected the 48-week time point itself, but it might have been an aberrant leading on the 24-week scale. Louise, did I miss anything with respect to that?

No. I think you captured it.

Our next question comes from Gena Wang with Barclays.

Just 2 quick ones. The first one, just wanted to ask if -- did you see higher protein level at age 4 to 5 compared to 6 to 7? And also, Doug, based on the current data of different -- of the subgroup, is it fair to say likely that you would have to run a Study 301 in order to get approval? And that based on the data we've seen so far, how can you -- what could be the study trial design you will have in mind? I remember in the past that we discussed like we've tried to minimize the North Star. So I'm a little bit surprised to see even in the placebo arm, we saw 24 for the age 6 to 7. So like going forward for the Study 301, what could be the best design enrollment criteria to minimize the mistake or like unfortunate events we've seen so far?

Yes. So a couple of things on that. First, we've got a lot more analytics to do. But on the face of it, there's nothing about protein levels that explains this. This result is explained entirely -- well, at least based on what we know today, entirely by what you see in the 6- to 7-year-old baseline characteristics. But the difference between those 2 groups impacted the active versus placebo for the 6- and 7-year-olds. And when you add that to the entire group, you missed statistical significance. As it relates to the design of the next study, we are going to learn from this study and address it. And there may be a number of ways in which we can address the next study to increase the probability of success even further. But one of the -- look, you're exactly right. The NSAA number for the placebo arm on the 6- to 7-year-olds, to use the colloquialism, is ridiculously high. It is, in fact, at the very top of the ceiling for the entrance criteria. And I will also say that ironically, when you look at the p-value, it is -- we are struck by lightning. It is a 1 in 250 or 4 out of 1,000 probability that, that would have occurred. But indeed, it did. We are going to address that issue. And frankly, notwithstanding the fact that it is an extraordinarily improbable thing that might occur, we will ensure and find ways in the entrance criteria and the design of the next study that we address this issue so that we don't have that issue for the next study. And then beyond that, I think on the face of it, if this -- that alone would drastically increase the probability of hitting statistical significance. But we have other learnings inside of our Study 102 that's going to help us inform the next trial.

Our next question comes from Anupam Rama with JPMorgan.

Doug, just following up on your comments right now. I don't think you've ever disclosed the floor or ceiling on NSAA for this trial. You just said that it was basically at the ceiling for the 6- and 7-year-olds in the placebo. I mean, like how close to that ceiling? And like in your next study, is it likely that you're going to kind of decline that ceiling to kind of maybe elicit a better treatment outcome?

Well, it's ironic. Let me be defensive for a second and say that I don't think on the face of it, it wasn't an issue with too wide of ceiling and floors because -- and the reason I'll say that is because if you look at the 4- to 5-year-olds, it worked brilliantly. In fact, those children had strikingly similar baseline characteristics. And they were, of course, subject to the same ceiling and floors as the 6 to 7s. It really was this crazy, improbable -- best we know, we don't think there was any kind of inherent flaw on the randomization. There was this very, very improbable randomization that occurred with one group at the very top of the ceiling and one at the floor. And while we haven't -- while I still will not provide the exact details of the floor and the ceiling for this call, I will tell you that you can backward engineer this because the placebo group in the 6 to 7s were right near the top of that ceiling, and the active group was right at the floor of that NSAA floor for the criteria. So -- but with all of that said, we're not going to simply say that was improbable, and we're not going to make changes, and we're just going to continue to forge forward. We will find a way through different kinds of stratification to ensure that we do not replicate this problem as, frankly, on the face of it, as improbable as it was, we are going to address these issues, though, that we are very, very confident in our next trial, the baseline characteristics will end up very similar between the treatment arms and the placebo one.

Our next question comes from Joel Beatty with Citi.

It seems that there was lower dystrophin expression in this study compared to the previous study. Do you think that, that's a real difference between studies or a difference in measurement techniques? And are you considering any differences in dosing in DMD or limb-girdle going forward?

So a couple of broad strokes on this. One, there were some changes in the way that the measurements occurred, but I don't -- we don't think on the face of it that, that played a material impact. One of the things that Dr. Rodino-Klapac did mention in the -- in her remarks was that the Nationwide Children's Hospital, in their manufacturing process, used a form of titering called qPCR. That titering does result in more lot-to-lot variability than what we currently use. We've come to a different titering process, much tighter, a linear titering process that has been applied not merely to our commercial process material Study 103, but also to all of the children and crossover we used linear titering. So while, again, I will harken back to my forward-looking statements and say it is dangerous to make predictions about the future, we do anticipate that we will likely see higher mean expression levels with the tighter titering range.

Our next question comes from Ritu Baral with Cowen.

When might we have the functional data from the crossover patients? And given that at least half of those placebo patients started off with really, really high baselines, what could you potentially show? I'm interested in any statistical analysis from baseline that might be robust enough to help you design 301. And do you need that before you start 301?

Yes. So well, we don't need it before we start 301. I want to start with that. We don't need this -- the crossover done until we start 301, and that would be our plan. We would have an analysis of all of the 41 patients, including all of the crossovers, at the 48-week time point, which would really be essentially at the very end of this year would be the analysis. And while I don't want to predict in advance, and we've got a lot of work and thinking to do, we do have prespecified analyses on all of the patients. That will be, in the many regards, different than the approach that we took in the Part 1 analysis because, of course, we're moving into a phase where all of these patients will now be on therapy. And so we have to use other measures, and some of those measures include the fact that many of these patients had significant run-in periods and then we have trajectory. So we will see -- as opposed to simply looking at baseline characteristics versus post-therapy or placebo characteristics, we'll be looking at trajectory analysis as well as natural history analysis. And there are a number of other analytics. So I don't want to predict in advance that it will make a significantly meaningful difference, but there is a real possibility that one of -- this analysis may wholly or partially address the fundamental flaw that occurred in the randomization of the 6- to 7-year-olds. So there is that possibility, but that will not -- that analysis will not occur until the 48-week time point for the last patient, so that will be really at the end of this year. And then just so we're very clear about that, we are going to take a deep dive into Study 102, and we're going to inform Study 301. We are not going to wait for that secondary analysis before we commence Study 301.

Our next question comes from Vincent Chen with Bernstein.

This morning, have you had the opportunity to assess what the expected change in the natural history would have been for the 6- to 7-year olds in the placebo group? And I guess the -- what the expected change in the natural history would have been for the 6- and 7-year-olds in the gene therapy group? You have -- I'm curious, what does it look like? And what's your best guess for what the effect size is achieved and what it would have looked like in the 6- to 7-year-old cohort if you had been able to adjust for this?

We're just in the beginnings of that analysis now. So I think everything that we could say right now would be fairly speculative. We do know on whole that 6- to 7-year-olds would have been in -- beginning to be in some decline right now, some serious decline. And then in fact, given that these children on the active group were on the very severe end of what one would have expected for a 6- and 7-year old, we're going to do a lot of patient-level looking at natural history. But I -- our presumption is you'd see a significant decline in these patients. And on the active therapy, you didn't see that among these patients. But again, we had that versus a comparator with very mild patients that didn't act consistent with the severe 6- and 7-year-olds. So we're going to do all of that work, but we would have expected, broadly speaking -- even on the mean, we would have expected 6- and 7-year-olds to be in decline right now. And of course, the kids on therapy were not in any steep decline right now at all. They were very stable. And in fact, these kids were on the more severe end. So very likely, although when we look at the patient-level data on natural history to confirm this, we might -- we very will -- might likely have seen an even more significant decline in this group.

Our next question comes from Matthew Harrison with Morgan Stanley.

I guess I want to go back to the comment you guys made about the titering process. 30% dystrophin expression is a lot different than, I think it was, 73% or 74% in the earlier studies. And so can you just talk about your confidence that the material really is the same and that some of the issues might not have been material-related?

Yes. I don't -- well, first of all, I don't think that it was material-related. The lot -- just so we're clear, this material was the same sort of -- type of material from the same manufacturing process at Nationwide Children's Hospital. So 101 and 102 that you've seen so far all came from the same process. There is -- that, and Louise will note, there is a titering -- there's variability in the titering that Nationwide historically used, this qPCR process. We identified early that we needed to move to a very significantly tighter process, which we did, which we use for the commercial process material that we use as well as for the -- even for the Nationwide material at the crossover patients, number one. Number two, I would remind us that actually, on the first 4 patients, if you -- Louise, you're going to correct me when I get some of these numbers wrong, I'll give them directionally, if you omit the fourth patient, which was a very high responder, you were sort of in the low 50s for Western blot, I believe. And so we're sitting here, what is lower than those 50s. But we do believe that when we tighten up the titer, what we'll see out of Study 103 is something that's getting closer to what we saw before. But of course, we're going to wait and see that in the next couple of months. With all of that said, I do want to be clear. If you're looking for this as an explanation, really the explanation here is the differences, and the obvious one is the differences in baseline characteristics. We took -- unfortunately, we took, in the 6- to 7-year olds, just through random -- the randomization process, we took a very severe group of 6- to 7-year-olds that ought to have been in steep decline, and they got an active therapy. And we compare them to a very mild -- these are just different populations. Every single solitary functional measure at baseline is significantly different and milder for the placebos than for the actives. Even the 4-stair climb, we noted that the 4-stair climb didn't hit statistical significance. But if I'm not mistaken, it was 0.09, and it was nearly statistically significant even in the 4-stair climb. And those kids in the severe group that were on active in the 6 to 7s, given that 4-stair climb, one of the things we already know is that is correlated with a significant -- that the slowness of that stair climb is correlated with a significant loss in the 6-minute walk test and then in ambulation. So I think that you can look at this statistically and say that, that 6- and 7-year-old imbalance appears on its face to explain why we wouldn't hit overall statistical significance. What is surprising -- that's -- that imbalance is unbelievably surprising. I do want to linger for a moment on the fact that in the 4- to 5-year-olds, which was only 16 children, they were well matched, very, very well matched. I'm very, very pleased to see that the randomization process worked there. And we saw a very significant delta across all of the time points between the placebo children and the active children. And even at 16 participants, we saw very robust statistical significance, 0.017, below 0.02 on statistical significance and a very significant difference. I think it's about 4.3 NSAA points versus baseline. So we do think -- we're going to do a lot more thinking on this. We've got a lot more analysis to do. We're going to do everything we can to greatly increase the probability that this therapy that we're very confident about shows itself in the next study statistically. But we're very pleased to see that in the 4- and 5-year-olds and do believe that if we get the balance right, that we're going to see a very positive next study.

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

I don't know if it's best for Doug or maybe for Louise. But understanding everything that you've said about natural history, progression, I'm just wondering, just based on the data that you've seen so far, your thoughts about whether the use of micro-dystrophin might just end up being better for younger cohorts. That the older you go in the age group, it's more important to have a full-length dystrophin presence. I'd just like to hear your thoughts on that.

Yes. I'll touch on it briefly, and then Dr. Rodino-Klapac can provide some comment on that. When you first looked at the data, you could -- you might wonder about that because you see the 4- to 5-year-olds that looked like they're performing very well versus the placebo versus the 6- and 7-year-olds, even though on the face of it, it really makes no mechanism of action-related sense that a 4- to 5-year-old would benefit from this truncated dystrophin, but a 6- to 7-year-old would not. And then at least from our perspective, this puzzle is resolved almost immediately as soon as you look at the baseline characteristics. And I think more -- Occam's razor would say more likely than not, if you want to understand why these 2 groups act differently than one another versus their respective placebo arms, if you look at the baseline characteristics, you realize that you literally, with the 6 and the we -- unfortunately, we and the 6- and 7-year-olds literally compared different patient populations. I mean, to use this sort of -- to use a metaphor, it would be like we had looked at a 5-year-old versus a 9-year-old with Duchenne. Nobody would imagine that, that would have been a successful approach. And that, in a sense, through random bad luck really, that's what we did. We took kids that had more than 4 points, almost 5 points or so of delta between them on NSAA on mean and comparing them to one another. And unfortunately, not surprisingly, we didn't see a statistically significant difference between those groups. I think more likely than anything else, that explains why we're seeing such a stark difference in the performance of kids that are 4 to 5 and 6 to 7. But Louise, perhaps you can provide additional color on this. Or if you disagree with me, certainly do that.

No. All I'll add is that the mechanism of action of micro-dystrophin, there was nothing that's specific about 4- to 5-year-olds versus 6 to 7 in terms of the way the micro-dystrophin would act. So it's what we predicted on its function. In the 6 to 7s, we have more data coming. And I think this is something that may -- obviously, the baseline characteristics influence this significantly, but we may see over time play out in 6 to 7s as well as the crossover. So the data that we'll be collecting will be important to inform this group. But as far as mechanism, it will -- there's no difference in terms of age.

Our next question comes from Colin Bristow with UBS.

To piggyback off a couple that were already asked. So for patients that were dosed early in the trial, is there anything you can share about the NSAA trends you're seeing beyond the 48 weeks? I think this was asked, but I didn't hear a response, the difference in expression, if any, between the 4- and 5-year-olds and the 6- to 7-year-old groups. And then I just wonder if you looked to Phase III or maybe even reconsenting current patients, is there any value at looking at the biopsies at a later time point just to get sort of greater insight into expression turnover and I guess how ultimately this correlates with efficacy? Would you consider this for the Phase III? That's it.

On the last point, I will note that with respect to our current trial, we don't have to reconsent the patients thankfully because the study remains blinded, and we will have additional biopsies in that study. But with that, Louise, do you have any comments on some of those questions?

And I think to your point, we do not have any functional measurements beyond 48 weeks at this time. And we're certainly looking -- we're doing additional analyses. But at this time, there is no indication that protein impact and the results or differences that we saw between the age groups.

Our next question comes from Debjit Chattopadhyay with Guggenheim.

This is Aaron Welch on for Debjit. So maybe I missed this earlier. But was the change in NSAA scores from baseline in the 6- to 7-year-old patients, was that reported? And is it possible to only enroll 4- to 5-year-olds in Study 301?

I'll let Louise answer the first question. The second question is, we certainly could create a study that was narrowly focused going on 4- to 5-year-olds. I -- we are still doing more work on this. I suspect that, that will not be the conclusion that we come to. And that, in fact, if we are correct in our analysis that the imbalance in the 6- to 7-year-olds explains why they [ will build ] the hit stat sig, then we can address that issue directly and still enroll that patient population. They should -- if the results we've seen so far both in the 4 to 5s and the preclinical and the 101 all bear out, they should certainly benefit in the same way that the 4- and 5-year-olds from this therapy. And we frankly think they are today. They're all benefiting from it. Louise?

Yes. So the 6- to 7-year-old group, they're essentially stable with only a 0.4-point decline. And as we mentioned, we would expect in this age group, especially with the severe or low baseline scores, to see a significant decline.

Our next question comes from Brian Skorney with Baird.

Not to harp on sort of the disparities between what you're seeing on the biomarkers, but maybe just trying to contextualize it a little bit in terms of the differences and tying that to any differences you're seeing in the safety profile. I mean, when I look at the percent positive fibers, that almost looks even more disparate. I think it was 81% in the Phase I versus 33% here. So it's only 1/3 of the positive fibers are actually showing dystrophin expression and similarly, the big differences in vector copy number. When you kind of -- when I also -- I guess just a straightforward question is, are you -- what is sort of the immunological markers that you're monitoring? I mean, in the 2 cases of rhabdo, is there any association with complement activation at all? Is there anything in terms of immunological markers that would indicate that you're seeing lower expression because the immune systems are responding to this more aggressively?

No. As we've said, Brian, we have seen no clinical complement continuing. One of the things, again, I would say is, and we do, do differently, I'll let explain Dr. Rodino-Klapac explain it. We did do a different -- a more automated approach to PDPF, but she can talk to you about it. I'd say with respect to the Western blots, we're seeing, excluding the fourth patient in 101, 28% to 53-or-so percent. And we do expect that when we evaluate the 103 data with a much tighter, a much more predictable and tighter titer, that we will see something that's going to start looking more like the data that we saw out of 101. Dr. Rodino-Klapac, do you want to comment on this?

Yes. Just on the PDPF assay, we're using an automated method. But we did do studies, and the 2 methods are concordant in terms of expression. And so like Doug said, we would expect on all measures, not just the Western blots, in future studies that we would expect to see higher numbers with the tighter titering method in the future studies.

And we did see very high intensity as well, I would know. We saw, I think about 60% on intensity.

Our next question comes from Danielle Brill with Raymond James.

I guess, Doug, I'm curious if there was a specific reason why you didn't stratify by NSAA score to that baseline? And then I might have missed this. But did you see any correlation with mean dystrophin expression and NSAA scores?

Yes. To the best of our knowledge right now, no to the second question. On the first question, it was our -- we did a lot to keep this study tight, frankly. We had a ceiling and a floor. We had frankly done a number of things in advance that we got right. I mean, we had identified what I think so many people hadn't and the literature hadn't fully identified before, which is that 4- to 5-year-olds and 6- to 7-year-olds act differently, and we stratified on that basis. We did put -- we did try to tighten up NSAAs between putting a ceiling and a floor. And I do -- again, this is -- I'm getting very close to being defensive about this, and I apologize, but let me say that it works. With respect to the 4- to 5-year-olds, it works, and that's why I think you see a nice balance. It didn't work in the 6 to 7s. I will note that the -- I wish we had done something different, but I will note that the probability that what we got would happen randomly was 4 out of 1,000. I mean, it should not have occurred. It was, in a very real sense, very, very bad luck, very significant bad luck. With that said, we're not going to allow it to occur again. And we're going to address this issue and find a way to ensure that this doesn't occur, even though, frankly, statistically, it ought not to occur another time.

Doug, could I just add, just with regards to the question about multiple strata? We actually considered a number of stratification approaches. We even considered using 2 strata, but owing to the size of the study, realized that, that was just not feasible. Owing to the patient numbers, it would actually make running the study extraordinarily challenging. We actually then also looked at age versus NSAA as the stratum and, for the reasons that Doug and Louise have outlined in their previous answers and Louise's presentation, determined that age as an overall surrogate was the better stratum where we had to choose a single stratum for randomization. Thanks.

Our next question comes from the line of Hartaj Singh with Oppenheimer.

And I echo Alethia's comments, tough day for DMD patients today. I just wanted to ask a question on safety. You've mentioned that you got 4 SAEs in the treated group, 1 in the placebo. Can you please let us know which ones are related to study drug, if any were? And whether they were probable or likely related to study drug of those 4? And then why did you get so much vomiting? I mean, it's almost 3x as much in the treatment group versus the placebo group.

Dr. Rodino-Klapac?

Let me answer that question about vomiting. This is -- we saw this in the 101 trial as well as our limb-girdle trial. This is not atypical of gene therapy trials that I've seen in others. So this is not unexpected, not as far as planning. I did miss -- you broke up in the first question. Doug, I don't know if that -- did you -- I don't know if you heard the...

He wanted to know if we have -- yes, he wanted to know if we have a view on high [indiscernible] with respect to the SAEs.

Well, I think I actually asked if there is relatedness to treatment, where the SAEs related to treatment, like how do we -- sorry. I don't know -- yes, I apologize. I mean, just if the SAEs were related to a study drug. Off those scores, which one were related and how likely were they related?

So these were -- as it related to rhabdo, there's obviously a potential side effect from -- indication for DMD. So it was not unexpected. There were 2 in the treated and 1 in the placebo arm. And the transaminase elevations, again, were expected from previous trials. And so those were designated as related.

Our next question comes from the line of Difei Yang with Mizuho Securities.

So just 2 quick ones. Number one is that, is there a option left to use the new method to analyze those floor patients, Study 101 patient sample, so that we can compare the expression data apple-to-apple? And number two is that, would you have the opportunity to talk to the FDA before starting Study 301?

So two things. We don't -- there's no reason to believe that the methods between the 2 approaches are substantially different in that there are no reanalysis of Study 101 that would render a different result. Study 101's -- remains very accurate in its expression. We certainly will be talking to the division before we commence Study 301. We'll do a couple of things with respect to Study 301. We're going to take all of the learnings from 102. We're going to think about the results from 102 to enhance the probability for Study 301. We can get that done very rapidly. That will all go into our thinking for Study 301. And of course, we will meet with the division and talk to the division before we commence Study 301.

Our next question comes from the line of Salveen Richter with Goldman Sachs.

Just following up on Anupam's question here. Could you narrow the target NSAA range for the older group versus the younger group, just given the slope of decline?

The older group and the younger group had the same ceiling and floor entrance criteria for NSAA. We will do additional work on looking at the natural history to get a better view on what -- how aggressively, for instance, we would expect this -- the patient populations in the 6 to 7s would have trajected both on the placebo with the much lower NSAA -- with the much higher NSAA scores and the active group in the much lower and more severe NSAA scores.

Our next question comes from the line of Joseph Schwartz with SVB Leerink.

I was wondering, have you looked at -- have you looked to see whether the less severe 6- to 7-year-old patients within the SRP-9001 arm performed better on NSAA than those who were more severe? Or that the more severe 6- to 7-year-old patients treated within the placebo arm performed worse than those who were less severe? Just to confirm your hypothesis that the baseline NSAA levels are so predictive. Are there any analyses like this that you can do to substantiate the hypothesis that an imbalance in disease severity amongst the older patients caused this aberration?

Dr. Rodino-Klapac?

Yes. We'll certainly be doing additional analyses, sensitivity analyses. I will mention that we're going to have a low end when you start to look at these things. And so we'll have to be very careful about the conclusions that we draw from that. But certainly, we will be looking at the data from every angle possible. Any additional data that we get from Part 2 will also be informative as well.

Our next question comes from the line of Tim Lugo with William Blair.

Could you include another biopsy at the end of Part 2 for some more expression data which may be informative? And you also showed time course data for NSAA. But was there anything positive to glean out of the time course data from the 4 steps ascend or time to rise or 10-meter secondary end points?

Dr. Rodino-Klapac, do you have a view on any of this?

Yes. On the biopsy, so we are taking a biopsy in Part 2. As I said, I don't know if you were asking for an additional biopsy. But right now, there is a biopsy at 12 weeks post-Part 2. As far as the other secondary measures, we're not providing additional data at that time point. We're still going through that, and the study is still blinded. And so we'll continue to look at all of those end points and follow up in future medical meetings.

Our next question comes from the line of Gil Blum with Needham.

Maybe a bit of a naive question, but it seems like the patients on the placebo trended upwards on NSAA over time. Is this normal in either age group?

Yes. It would be normal in the 4- to 5-year-olds. You would see -- you see just about what we saw on this trial in the 4- to 5-year-olds, you could see a modest increase in NSAA over that course. And one of the things we had often talked about as we design the trial, so if you look at the literature, you will often see kind of 4- to 7-year-olds treated very similar. And one of the things that we had noted and made a point of discussion was that actually, there's a difference. The 4- to 5-year-olds can benefit about a point or so, and that's about what we saw over the course of this trial, and the 6 to 7s will start declining. And so what we saw in this trial was the 4- to 5-year-olds acted as we would have anticipated. The 6 to 7s in the placebo group were very mild relative to what we would have expected, and we didn't see that decline, of course. And then unfortunately, the actives were very, very aggressive. And what we would have thought we would see in natural history is that they would be -- would aggressively decline. And we didn't see that either, of course. And I -- we think that's because they're on therapy, and they're benefiting from the therapy just like the 4- to 5-year-olds are. But we're going to have to do more -- obviously more work on that, and it will help inform Study 301.

That concludes today's question-and-answer session. I'd like to turn the call back to Doug Ingram for closing remarks.

Thank you all very much for joining us today. I want to thank as well our participants in the Study 102 and their families for their willingness to play a significant role in advancing this therapy. We are -- I'm not going to hide from the fact that I'm disappointed that we did not hit statistical significance in our primary efficacy end point. I am, nevertheless, both frustrated but also pleased ironically that it does appear, at least on the face of it, that an imbalance in the 6- to 7-year-olds explains this. And that when we adjust for that, as the 4- to 5-year-olds did because they naturally had a better balance between the 2, we see in this therapy what we would have anticipated, a therapy that brings a transformative benefit to patients with Duchenne muscular dystrophy. We're also very pleased that we not only saw good expression, and we actually think we'll see even better expression with the better titering that we have and the tighter titering, but we saw good -- we got good expression. We've got obviously good genome copies per nucleus, and we saw a statistically significant benefit on CK as well. And we're going to take all of this data and all of these learnings, and we're going to prepare our next study, Study 301, on that basis to enhance the probability of success of that study and continue to move with a sense of urgency to bringing this therapy to patients waiting around the world, as we are convinced that SRP-9001 will play a transformative role in the treatment of Duchenne muscular dystrophy. As we track across this year, there will be a number of moments with additional readouts. We'll have a readout on Study 103, which will provide data both on expression as well as other biomarkers and [ safety ]. And then of course, in the back half of this year or into early next year, we'll have actually the Part 2 readout of Study 102, and we'll commence Study 301 this year. So thank you all very much for this. Thanks for spending time with us this evening, and we'll continue to update as we progress throughout 2021.

Ladies and gentlemen, this concludes today's conference call. Thank you for participating. You may now disconnect.