Good morning and welcome to the Ultragenyx GTX-102 for Angelman Syndrome Phase I/II Interim Data Update Conference Call. At this time, all participants are in a listen-only mode. At the end of the prepared remarks, you will have an opportunity to ask questions during the Q&A portion of the call. If you would like to queue up for questions, please hit star 1 on your telephone keypad. As a reminder, this call is being recorded. It is now my pleasure to turn the call over to Joshua Higa, Vice President of Investor Relations. Thank you. We've issued a press release detailing these results, which you can find on our website at ultragenyx.com. The slides we are presenting today will be available for download in the IR section of our website shortly after the presentation. Joining me today are Emil Kakkis, Chief Executive Officer and President, Eric Crombez, Chief Medical Officer, and Howard Horn, Chief Financial Officer. I'd like to remind everyone that during today's call we'll be making forward-looking statements. These statements are subject to certain risks and uncertainties, and our actual results may differ materially. Please refer to the risk factors discussed in our latest SEC filings. I'll now turn the call over to Emil. Thanks, Josh, and good morning, everyone. Glad you could join us for the call. We're excited about the data we have on Angelman Syndrome and to provide an update on this important program. If you move through the forward-looking statement, slide three is just a disclaimer. Of course, GTX-102 is an investigational drug, and we're showing open-label data. Ultimately, we think proof of efficacy and safety will require randomized control studies. Slide four: We're going to talk today about expansion cohorts A and B data, up to day 170 as the main body of the data. We'll talk about what's new. It's 24 patients' worth of data through day 170. We'll also give an update on dose escalation cohorts four and seven up to day 758. Then we'll talk about regulatory discussions and phase III planning as well. Let's move on to slide five. GTX-102, as you know, is a devastating neurodevelopmental disorder, a fairly large, rare disease population of around 60,000 commercially accessible territories. And we have identified in our developing what we think is the most potent science behind the ASOs with the targeting of a region near the five prime end, which is, we think, the most potent and the greatest potential in being able to prove Angelman Syndrome. And I think today we'll show you data to help support that conclusion. In slide six, I'll just give you the high-level key takeaways, and then we'll start diving deeper into the study design and the results. Cohorts A and B show rapid and clinical means improvement across multiple domains, and we're very encouraged by what we're seeing. When you look at the improvements, they are consistent or exceeding what we saw before with dose escalation cohorts four to seven at the higher doses used in cohorts A and B. Cohort four and seven continues to show increasing clinical benefit over time, suggesting that the effect we're seeing on the brains of these kids continues to gain ground the longer we treat. And with this information, we are well set up to start and begin, and we have begun, the phase III planning process for this program with our goal to initiate phase III before the end of this year. Slide seven just talks about the domains. These are domains you've seen before at our analyst day. What we have, basically, the Bayley will be showing you data for cognition and receptive communication. The ASA is used for behavior, sleep, and gross motor. Now, gross motor, in the past, we've shown you Bayley data, but we don't have Bayley at day 170 in the way this was done to help reduce patient burden because Bayley is a more burdensome assessment. But we'll show you the ASA for gross motor for comparison so you can see the improvement in gross motor. Those are the assessments. And for some of these, we have natural history, and we'll show a comparison to natural history where we find them. On slide eight is just a little bit more about the natural history data we have. The original data we've shown you before is from natural history study I, which is shown here in older studies. It included 63 patients which had the deletion type of Angelman Syndrome in the same age range. So we're only looking at natural history patients that match the ones in our trial. We also have data now from natural history II, which is distinctive in using the Bayley-4 directly. This is a relatively small set through day 365, but we'll continue to increase as time goes on. I think if you look at the natural history sets, you'll see that the data are essentially the same. We don't see a meaningful difference between them. Those will be the comparisons we use where we have data on natural history. Now, let's go on to slide nine. The Angelman Severity Assessment we use frequently. The ASA score is designed with Angelman in mind. There's very specific 68 questions that are very responsive to the changes and problems that occur in Angelman Syndrome. This is why these tests are, I think, more sensitive and more responsive. Most of the patients are really probably in the four to six range, somewhere at three, but they're mostly in the moderately severe range. These are deletion patients. Based on this, changing from markedly to moderately, moderately to mild, a one-point change here represents a significant change in clinical state for these patients. That's the one-plus change that we set as a clinically important change for these assessments. Now, on slide 10, we'll just touch on the exposure. We now have 74 patients on drug. What's important, we've been treating these patients in the cohort four to seven for nearly 2.5 years, and the cohorts A and B for up to a year. This combination, the number of patients treated for so long, tells you a lot more about the safety of the program, that you can treat this many patients for that long in dose ranges from 3.3-14. We'll be showing you today 24 of the 34 patients that are in cohorts A and B, so the majority of these patients are in that cohort through day 170. Now, let's just talk about who we enrolled on slide 11. The cohorts A and B, you can see shown here, both age and sex are very similar and very comparable to what we've shown you before in cohort four to seven. The populations really look the same. What's different is that we've enrolled now at 19 sites across seven countries. We've gone across different languages and different settings, and we're going to show you that we can replicate the same effect across the world, which is a great setup for phase III. On the bottom of this slide, you can see the dosing regimen for cohorts A and B. There's a monthly dose, and then it goes every three months for dose five and six. We mark on here day 170 evaluation and day two through four, which are ones that we'll be talking about as we go forward. There are other points of evaluation, but not all endpoints are evaluated every single point because of the burden of doing that many tests. So that gives you an idea of how to anchor what we're going to show you in the data that comes. So now, let's get on to slide 12. This is where we get into the actual data. We start with cognition by Bayley-4. What we're showing is rapid and significant improvement. If you look at the left panel, you can see the blue dot. At day 170, we're actually crossing already the clinically and statistically significant change with very consistent tight error bars showing the improvement on par with what we saw before, replicated now in all these sites. If you compare that to natural history on the right, you can see that there's barely any movement in the scores, showing how much different it is. Now, if you look at the cohort four to seven data, you can see the patients continuing to gain ground across now almost three years of time with many folds above the statistics threshold. So this shows that whatever we're doing, we're turning on these brains that continue to work over a long period of time and gain ground. This is very encouraging in our view that the cognition continues to gain ground as these kids' brains start to develop. So the data shows we are matching what we had before and doing it consistently. Slide 13, we talked about behavior. ASA score for behavior. In this case, you can see the blue line, which is the cohort A and B, came down very far, in fact, much further than we did with the ASA in the cohort four to seven, in fact, doubling the decline rate in that day by day 170. And if you look at that line where we hit at 24 patients, it would have took us about a year to get there before cohort four to seven. So we've accelerated that by half, essentially, in time. If you look over time, certainly in the cohorts four through seven, they continue to gain ground and get better over time, but the speed and rapidity we're seeing of this effect exceeds what we saw before. On the right, to give you an understanding of what you're seeing in the ASA score, you can see the basis for the 68 questions: excitability, impulsivity, compulsivity. These are all characteristics of abnormal behaviors. And you can see the response rate. These are the CGIC scores at day 128. These are whether a patient feels that they're or the doctor feels that they are better by plus-one or better. And you can get an idea of what patients are responding to and how they're changing. It just gives you comfort and understanding what these ASA scores mean by looking at what we're looking at when we make that assessment. So we're excited about this behavior improvement. It's a very important one for patients, as we'll hear about later. Let's go on to the second measure for behavior. We added another measure, a corroborative measure called the ABC score. And we're looking at one particular domain within that score, the hyperactivity noncompliance score. And what we're showing you here is the same pattern of response. By day 170, a very significant response in this hyperactivity reduction, hyperactivity, which continues in a few patients that actually made it to day 254 as well with very substantial improvement. And this one does have natural history data showing far exceeds that of natural history. So this is just another data with a validated endpoint showing the same behavioral improvements that we saw with the ASA, giving us comfort in the value and reproducibility of that data. We go on to slide 15, where the sleep is assessed. Again, the blue dot in the left panel comes down all the way to greater than minus one mean. So the mean is already a plus one score or a decline of one score already. And that is on par with equal to wherever we got to with cohort four to seven already at day 170. So we achieved a level of sleep improvement that exceeds or matches what we saw through the entire day 158 with the cohort four to seven. Cohort four to seven have a rapid sleep effect. It is sustained over time. If you look at the anchors, you can see what this represents: night awakenings, caregiver disturbances, the amount of sleep, etc. These are different ways that the sleep is improving for patients. But it shows you a broad array of changes that are improving in sleep. And the rapidity gives you comfort that the cohort A and B dose regimen is giving us a very substantial improvement in sleep, which is a fundamental improvement for patients. On slide 16, we have the receptive communication by Bayley-4. Again, rapid improvement observed on par with what we saw before with the Bayley-4 receptive. And what you can see there in this case is that the error bars are very tight on that blue dot compared to the red dot. What that tells you is that we're seeing a more consistent pattern of response across patients at this higher dose, meaning patients may have been less responsive or moving now. Consistent, giving us a tighter error bars. It just gives you a sense that we are having a more consistent pattern of improvement. When you look at the natural history, of course, this far exceeds what you see with natural history. And over time, receptive communication continues to gain ground going above the statistical and clinically significant threshold as well. But we're encouraged by seeing a replication here of the receptive communication improvements in our A and B cohorts. On slide 17, we'll go on to the gross motor. Now, here we're assessing by the ASA rather than by Bayley because we didn't do the Bayley at day 170. We will have Bayley data later. On the left, you can see that the blue line and blue dot go well below what we saw in the cohort four to seven, showing a more rapid and more significant improvement in gross motor severity than we saw before, consistent with what we believe is a really important effect on how patients are walking, as we showed you before. If you look at the dot, if you look across, the improvement we're seeing here at day 170 is on par with what we saw maybe with a year of treatment in cohorts four to seven. This is showing a more substantial effect at an earlier time point. When you look at the Bayley on the right, which has now some additional data, you can see the patients continue to gain ground by the later time points across statistical thresholds. But the Bayley is not really designed for Angelman. It's a general test. I think the Angelman severity scale is telling you about the things that affect Angelman. These patients are improving rapidly and give us great confidence in the importance of what we're seeing. So that's the main body of data. I'd like to go on slide 18 and just give you kind of an overall comparison now. If you look at all the ASA scores and compare cohort four to seven with cohorts A and B, we can see whether you're looking at sleep, behavior, communication, or gross motor. On the left panel, you can see the sleep has a higher percentage of responders of at least a one point change or better in every single measure in every single domain. The communication domain, which is one we haven't mentioned before, an ASA just shows the level of response, both in receptive and expressed communication, are improved in these patients. We're comfortable then looking at this that the dose we're using is exceeding what we saw before with the four to seven cohort. If you look on the right, we even have, with sleep, 10 patients that already have plus two improvement in their sleep, which is a profound improvement in sleep and very encouraging. It's this very early day 170 time point. So I think what we can show is that at day 170, the new dose regimen is exceeding what we saw before in cohorts four to seven. I think that's very clear, and we're confident about it. If we go on to the next slide, this is the multi-domain responder index. Because we didn't have Bayley data, we didn't include the Bayley in this particular version of the responder index. We used the other four domains that you've seen before. What you can see on the left panel, which is day 170 cohorts A and B, is that all but four patients have a responding domain by day 170, which I think is phenomenal, the consistency of that response. The majority of the patients have two to four domains already by day 170. And we would expect this to continue to improve over time. You can see a wide variety of different domains and different patients and different combinations. But it gives you a sense for the breadth across these patients in their response and achieves a p-value with a plus two median domain number achieving success and a p-value, of course, very small, less than 0.0001. On the right, we show you the comparable. But this is for day 338 for cohort four to seven, showing you similarly a plus two domain median but at day 338. So it just shows you that we're achieving the same level of response essentially at half the amount of time in the new cohort. And it just shows you how powerful and broad the effect is across individual patients. That's easy to see when you're looking at this heat map type approach. Now, let's go on to safety on slide 20. I think the data we've seen show we have an acceptable safety data set and that there were no unexpected serious adverse events. So we're very comfortable with what we're seeing with safety. With regard to lower extremity weakness, there's one patient from cohort seven that we'd reported previously. That patient has recovered completely. There were two new patients that are more recent. They haven't reached day 170 yet at the time of the data cut. One patient was very mild. It was actually not recognized. It resolved within a week and was recognized retrospectively as having it, so it was very mild. One had moderate symptoms. Both resolved and are staying in the steady. We're comfortable with what we're seeing here. We think it's a much more minimal effect and very manageable. When we look at what's happened with patients that have had lower extremity weakness in the past, five of the original patients from cohort three are all being redosed multiple times without recurrence and having good clinical effect. Cohort seven patient also has been redosed multiple times now. That's a kid with scoliosis. He's receiving maintenance therapy now without recurrence. The two patients from cohort B then, they would expect them to continue dosing as they move forward. We've discussed the whole safety profile at our meeting and subsequently with the FDA and other regulators notified. There were no issues raised by any regulators. No additional actions requested in terms of evaluation analysis. I think we're past now this issue being something of great concern. We have to keep our eye on it. But we think it's manageable. And I think the new approach to dose administration, Trendelenburg and flush, have resolved this issue. And we just need to keep our eye on it. But we think it's not a block to therapy. And all patients have been able to get back redosed successfully without recurrence. So that's the safety data. Now, to try to get deeper to the clinical meaningfulness, I'm going to hand the call over to Eric Crombez, who will go into some of the physician and caregiver feedback. Eric, go ahead. Thank you, Emil. Next slide. We also see support for the clinical meaningfulness of the data we just reviewed from feedback from the physicians and patient parents participating in the study. In this slide, you see comprehensive improvements across all five domains. And just to highlight a few, you see learning as part of improved cognition. Patients are calmer and better behaved, which helps with school and going out in public. Disordered sleep can have an outside impact on not only the patient but also the entire family. We are hearing that improved sleep is one of the most important things to some of these families. Development of language helps with understanding and following directions. We are also seeing signs of development of expressive language. The improved motor skills provide better and safer movement. Next slide. Equally important is the feedback from caregivers. Again, you see how patients are learning new skills, which helps with independence. The improved behavior is providing a level of freedom that they have never had before. The new language skills also improve the quality of the way these patients are interacting with family members. From all of this, I find the most compelling feedback to be from a parent who can now imagine her daughter living with assistance in her own home and contributing something of value. Next slide. We would like to now share a few videos from a patient enrolled in one of the expansion cohorts. While each child is different, the progress that this patient has made within the first 170 days of treatment is remarkable and does generally represent the progress we are seeing in the expansion cohorts. We will start with a video taken before enrollment in the trial and then a couple of videos highlighting the development of new skills. Before we start playing the videos, I would like to remind the audience that the videos should not be captured or redistributed without express written consent. Operator, please start the video. This is the before video. This test of puzzle solving evaluates cognition and fine and gross motor. Here, you see the patient struggling to hold onto a ball with a hole in it and inability to place a ball on the peg. You see that the patient is getting support to stabilize her trunk in a standing position so she can focus on using her arms and hands. When the therapist lets go, the patient leans forward against the table. The only way for her to place the ball on the peg was for the therapist to guide her hand in placement of the ball. Here, you see the same patient threading beads onto a string as a test of cognition and fine motor function. You can see that she understands the task. She has the ability to independently focus on a task without redirection or hand-holding. She makes repeated attempts without frustration. This is learning. The pincer grasp and finger isolation to pick up the beads are a new skill. This skill is also important for eating with utensils. She is showing evidence of improved cognition by independently focusing on the task without redirection. Angelman patients typically cannot walk down stairs without support. Here, you see that she is focusing on the stairs and exercising caution. This improvement in balance and coordination prevents falls. Reciprocal stepping, step over step, instead of stopping at each step to put two feet down for balance is a new skill. Angelman patients have difficulty walking, especially on uneven surfaces, and have frequent falls. Now, you see her showing sufficient balance and coordination to navigate uneven surfaces. She is focusing on the task and exercising caution. She is cognitively aware of the branch and risk of falling and actively trying to prevent it. She slows down, thinks, and steps over it. To catch the ball, she is showing focus. She is taking direction, and repeated attempts are evidence of learning. You also see sufficient trunk stability to focus on her arms, which was absent in the first video. The ability to attend to a task, focus, and follow instructions is one of the most notable cognitive findings in these videos. Progressive increase in balance, coordination, and trunk stability to allow for performance of more complex gross motor activities highlight the overall progress that this patient has made within the first 170 days of starting treatment. These patients are generally representative of the progress we are seeing across all cohorts. Emil, I'll turn it back to you to finish out the presentation. Thank you, Eric. I think the video is pretty compelling and how fundamental changes in life that we're seeing. And to see that in a lot of patients and the feedback we've gotten, it's just been tremendous. We're excited about what we think this can do for Angelman patients. We'll go on to slide 25 and talk about regulatory processing. We've had FDA interaction, including a meeting earlier in the year. We'll talk a little about that and where we're going from there. At the meeting, we talked about the study design, randomized placebo-controlled trial around the size and those elements. The FDA was aligned with those elements. I think they were pleased that we were taking that direction. I think it's very important to recognize that with these types of endpoints, it's very important to have a placebo-controlled trial to be able to interpret and use these endpoints, not just for regulatory approval but for reimbursement for these kids in the long run. Doing one randomized trial, I think, is essential to get to where we need to go. With regard to the endpoints, all the endpoints that we need are contained in our study. That is, the FDA did not ask for any other endpoints, any other measures. So I feel comfortable that we got the data set we need. And they were happy with the set of endpoints we're measuring. They gave us some flexibility with those instruments. But we're confident what we have is what they're looking for. With regard to MDRI in our discussions, we think we'll probably put the MDRI as a secondary endpoint rather than primary with that discussion. But they have flexibility with including it in our data set, which will help support the efficacy of the drug. So we feel good about that. We'll actually bring the data we have today back to them and kind of come to that final conclusion about endpoints. But we're very comfortable. I think the FDA has been very collaborative and the other authorities equally. So we feel good about being able to get through this without too much difficulty given the data we have we've shown you today. From a safety perspective, there were no issues raised, no additional actions. I think they were very comfortable. In fact, we spent very little time talking about it. I think they're well past it as we are. It's a manageable thing. Patients return to therapy. And it's a reversible thing. That's not going to be a problem in the long run. So with these data today, probably a little bit more extension data for some of the patients. We'll end up getting the FDA mid-2024. And after that discussion, settlement of the agreement of the design, we'll let you know. Our goal would be to get to a phase III start right later in the year. On slide 26 is just the phase III design. It's pretty, I would say, pretty basic design, a randomized controlled trial of 100-120. The size of the trial is really driven more off of having enough safety data and to make sure we're powering all of the endpoints. We don't really need that much. But it's what I think is an appropriate amount for this size of a population. The length of trial will be decided. But it's likely between 254 and 338. If you look at how the data improves over time, that's an amount of time which we think will get us substantial improvement. With regard to the phase III patients, we are going to focus on the deletion four to 17 that we've been treating so far. What we expect is that the other types of patients we'd include in another trial, not a randomized trial but an open-label trial, to help assure we have the broad label for its indication. But one randomized trial, we think, will be enough to prove the efficacy of this treatment. And let's move on to slide 27. There've been a number of people who've asked us to try to give us a sense of power. And what we can tell you is based on if you look at the Bayley-4 cognition score, the magnitude of benefit we're seeing could readily support a phase III, all right? At day 338 or day 254, the first two rows in the table, you can see the current data compared to the pooled natural history data. You can see the power to do that assessment with 120 patients enrolled, 108 completers, is well over 99%. That's way overpowered. It does not need to be that great. That just gives you a sense for the level of improvement with the variation. If you take the Cohort A and B data we showed you today at Day 170, even there, we're already powered to see that amount of change. We don't think Day 170 is long enough. It at least tells you how much more power we have than we require. In the last case, Day 338, if you take the natural history data and just multiply it by three, make it threefold higher just as a safety assessment, we still have more than 90% power to detect the change. So the point with this study is to say we can readily power a study to achieve the measuring efficacy of the magnitude and variation that we're seeing in our data today. So we're very confident about that. Very good. Let's go on to slide 28. What we're showing you today are strong and consistent results for GTX-102. Positive data from the new cohorts across many sites around the world in seven countries, rapid and sustained clinical benefit corroborating what we saw before and exceeding with regard to the degree of effect we see by day 170. So we're very comfortable with the trajectory of these improvements and what it looks like going forward. With these strong data and the excellent situation with regard to safety, we're in the position of and we have then initiated phase III planning. We'll expect to get to the FDA, get the end of phase II meeting completed, and also speak to other regulators to head toward phase III initiation by the end of 2024. GTX-102, we think, with these data, put us in a position of them being one of our major products going forward and one of five approvals we expected in the next two to three years for Ultragenyx. That puts us in a great position having a major product that will have, I think, a transformative effect in Angelman Syndrome based on the data. We still have to get through a phase III trial. But we are very excited about what we're seeing so far. It tells us that we have a drug that will do something very important for Angelman patients. That's the data for today. We now can open the call to questions. Thank you. We will now be conducting a question-and-answer session. If you would like to ask a question, please press star one on your telephone keypad. A confirmation tone will indicate your line is in the question queue. You may press star two if you would like to remove your question from the queue. For participants using speaker equipment, it may be necessary to pick up your handset before pressing the star keys. We ask that you please limit yourself to one question and one follow-up question. One moment, please, while we poll for your questions. Our first questions come from the line of Anupam Rama with JP Morgan. Please proceed with your questions. Hey, guys. Thanks so much for taking the question. Congrats on the data. The behavior domain is one that, based on prior data, we thought might take a little bit longer to see a benefit. It seems like you're seeing better effects than what at least we had anticipated. What do you think is contributing to this? Well, it's interesting. I think it's a combination of things. But of course, improvement in sleep could probably have an impact on behavior as well. I think that a lot of their behavior is essentially hyperactive and neurotic. And the cognition improvement combined with the sleep probably is combined to give them an effect on behavior. It's very much in line with what you see with the sleep as well. And so I think they're probably somewhat related. When you make patients sleep better, they will improve their behavior. But we're encouraged with the rapidity, particularly at this dose level, of what we're seeing. Thanks so much for taking our question. Thank you. Our next questions come from the line of Gena Wang with Barclays. Please proceed with your questions. Hi. This is Gena Wang. Thank you for taking our questions. So for the SAE on the three patients, one from cohort seven and the other two from cohort A and B, may I ask how many doses follow-up did they start showing this symptom? How long did they take for them to resolve? And how did they resolve? Thank you. Yes. Generally, these things happen within around four doses, around three, four doses, around that point. That's where we had seen it. For cohort seven, within a few weeks, he was better. For the mild patient, it was like a week. It wasn't even recognized upfront. It was recognized afterward that maybe they had some weakness. And so for the other one, it was just a few weeks. It was actually relatively short. So these are resolving much faster than we saw in the original. It's a much milder effect than what we've seen before. So it's within a few weeks. And it's after some doses. What's interesting, you can put them back on treatment and give them multiple doses. And they're fine. So it's very important about the technique and how it's being done to make sure that that's being done well. But we feel comfortable if you can redose all those patients and do it safely and show the benefits of the drug. I think we're just confident. It's just one of those things we're going to have to manage. But it's not something that stops therapy with GTX-102. Okay. Yeah. I think since you modified your protocol, so there's nothing wrong happened during that administration. And also for higher dose, how many doses have you already dosed? Because you mentioned you don't have this symptom right now for the higher doses cohorts. Well, the dosing for cohort A and B is the higher dose that we're going to be using going forward. So the doses we administer here, yeah, are the dose we're going forward with. So yeah. So we have the exposure in these well, it's really like 50 patients. There's 24 today. But we are looking at safety across A through E cohorts. So it's more on par with like 50 patients worth of patients being treated at this dose level. So the incidence of this is relatively rare. And since it's resolvable and reversible, they are able to redose. And I don't think that I just don't think it's the same problem we thought we'd had. I think the dose administration strategy allowed us to get better efficacy at a lower dose and reduce the risk of this problem. Thank you. Thank you for taking our question. Thank you. Our next questions come from the line of Tazeen Ahmad with Bank of America. Please proceed with your questions. Hi. Good morning. Thanks for taking my questions. So Emil, can you just clarify what additional data you're waiting for before you go and talk to FDA? Secondly, what preliminary, if any, discussions have you had with the agency on this specific data update? And then lastly, can you provide any feedback from physicians about the profile of the drug as it stands now? I think there might be some confusion about whether or not expectations were to completely eliminate lower extremity weakness. And since you haven't, how does that impact the overall profile of the drug in doctors' views? Thanks. Great. So we're not really waiting for more data because the process of requesting a meeting is initiating now. The FDA has not seen this data. This is just data we're releasing, right, at this moment. We just got it ourselves. So they haven't seen it yet. But we'll be sending this in. But it takes time to request the meeting, get to the meeting, schedule, etc. In that interval, we may collect some additional extension data that we'll bring to them at the time of the meeting. But we're proceeding to request the meeting at this point in time. And they haven't seen it yet. With regard to the MDN profile, we've not had any issue with doctors in dealing with it, including some of the doctors that had the event and put people back on drug. I think people are pretty comfortable with it. They understand it's reversible. It's certainly much milder than we saw before. It's certainly a concern. But I haven't had anyone say they don't want to be part of it. In fact, we have mainly doctors trying to call me to get into the study or patient groups repeatedly from various countries trying to get us to bring the phase III to their country. At this point, those changes are not a concern compared to what the benefit people are seeing. We're comfortable with the safety profile. We haven't had any feeling of a concern. Okay? Thank you. Our next questions come from the line of Yaron Werber with TD Cowen. Please proceed with your questions. Great. Thanks so much as well. And nice job in a really tough disease. Maybe I have a couple of questions, really, Emil, just on the endpoints. So it looks like the powering you gave us talks about Bayley-4 for cognition. Is that you're thinking as a primary and then a host of secondaries? Or are you thinking about potentially going something a little bit broader as a primary? And then maybe secondly, the one data we haven't seen sort of is the ASA, the ASA overall, or the CGIC. Just any thoughts on those as endpoints? Or you're going to sort of dive more into the subdomains? Yeah. So thanks, Yaron. The Bayley-4 cognition is one of the ones that has the most broadest importance and impact. It continues to gain ground. We are looking at that as a potential primary. We'll still have that discussion with FDA to verify. That's one that seems to make the most sense. We could look at some others. The idea is if that were primary, then we'd put some of these other endpoints as secondary. With the MDRI as a way of integrating a per-patient basis response to help support the meaningfulness of that, that's our view. Right now, the ASA overall, what I would say to you is the domain-specific assessments are a lot more powerful than the overall. The overall basically dilutes out things. It doesn't give you as precise a result. We will use the CGIC. We used it here at day 128 to show you that's the change scores, excuse me, the CGIC's change scores. But the FDA, although they allowed that in the Aurora phase III, change scores are not anchored to specific data and are not usually preferred when we talk about phase III studies. So the Angelman Severity Assessment, the CGIS, what used to be called CGIS, is anchored by specific questions. So it has more of a fundamental basis. So we'll focus on ASAs. We might do CGI change scores just plus one, plus two here and there in the program. But with regard to endpoints, we'll focus on data that is driven off of hard findings on the patient rather than the physician's opinion of how they're doing. So that's the big picture right now, Yaron. Thanks. Next question, please. Thank you. Our next questions come from the line of Chris Raymond with Piper Sandler. Please proceed with your question. Thanks. Yeah. Just two questions. So I want to understand maybe the plan with younger patients. I know you guys have talked about in the past not wanting to run or not being feasible to run a placebo-controlled trial on younger patients. But any sort of plan to address that group of patients? And then I know you mentioned, Emil, you'd have an end of phase II meeting here soon. What's the timing of when you'd have certainty around the primary that you could communicate with us? Thank you. Yeah. So I'll be able to second question. The end of phase II is like I said, mid-year. And it depends on when FDA scheduled it. We will request it. And then we have to get to that. So it's about mid-year. Around that time, we'd expect from that meeting to come to a conclusion on primary endpoint. If we are fully resolved, we'll put out a notice on where we're at on that interaction. We definitely want to treat younger patients. It's just as a matter of policy, we don't put little kids in phase III trials. We haven't studied them before. We do need to understand dosing in the younger patients. In the long run, treating kids from when they're one or two years old could very well be substantially better than treating when they're older. We will want to do that. It'll be an open-label format in this situation with the randomized study showing you the cause and effect of the drug. We think in this younger patient study, we can show that in deletion type or other types that the improvement we're seeing, we can be confident is a real drug effect. So our expectation is to be able to do an open-label trial. There are also the other genotypes that we want to look at. And those could be put together in a basket-type trial that we would conduct in parallel with phase III. We do want to make sure we get the dosing right in much smaller kids too. So that would probably involve some dose titration to get to the right number. Thank you. Thank you. Our next questions come from the line of Yigal Nochomovitz with Citigroup. Please proceed with your questions. Yeah. Hi. Thanks, Emil. I just had a question on the MDRI. I recall at the analyst event back in October last year, the company was positioning that potentially as the primary endpoint for the phase III. And then you mentioned that FDA was more comfortable making that a secondary. Could you just explain in a little more detail why the FDA was pushing back a bit on MDRI as the primary endpoint? Well, we've been talking to MDRI. I've been talking with FDA and various parts of the FDA about MDRI for a while. It's a new approach. It's an individual patient response kind of based approach rather than a continuous variable. I think the FDA expressly said they'd prefer just a continuous variable primary endpoint type analysis, more traditional. I think they're just more comfortable with that. I think the MDRI will be in there and allow us to support the efficacy. And we're comfortable with having it in there as a sort of alternative to looking at cognition, for example, if that were primary. Cognition involves improvement in a number of different parameters. So it is a little bit of an integrative type of endpoint already because receptive communication, expressive, attentiveness, behavior, all that stuff play into the cognition assessment. So I look at that as being broader to begin with. But MDRI as a key secondary could very well provide the support we need. And I think from my own perspective as a company, we're trying to advance the science of rare disease drug development. And putting the MDRI in there will give us a chance in a large randomized study to show how meaningful it is as an approach. Putting it secondary just gives us a chance to get the benefit of it without creating any challenge for the FDA at this point. But we're excited about what we've got because we can make this work almost any way. You can pick practically any endpoint we have and be able to make the study work. So we have a lot of options in front of us. That's not always true. We have multiple endpoints we could use to get a successful phase III trial. Okay. Thanks. And then regarding the dosing, you've quoted the range many times. Is there a point where you're going to be a little more specific regarding exactly which doses within that range were used and how many patients per dose? Yeah. We'll get more specific at some point in time. We're keeping it more secretive now for competitive reasons. But we've given you the range. So obviously, the loading dose is within that range. And what I can say from that is that this ASO is very potent, far more potent than those being tested by competitors. And that's the key thing, that we have the right science technology approach to how you're going to turn on the UBE3A gene. And the fact the dose range is so low, I think, is what makes this drug special. Thank you. Thank you. Our next questions come from the line of Dae Gon Ha with Stifel. Please proceed with your questions. Hi. Good morning. Thanks for taking our questions. I guess on the question of endpoints specifically, if we think about the natural history that you presented today, there was a natural history 1 as well as natural history 2. Some discrepancy there. Natural history 1 seems to be more of a derivative/extrapolation versus natural history 2. I know in the powering calculations, you kind of assume three times just for the sake of powering. But how should we think about the natural history more, I guess, relevant to this case? Is it more natural history 2 or natural history 1? And then I've got a follow-up. Yeah. I think, Dae Gon, it actually really doesn't matter. The natural history 2 is a smaller set of data. But they're flat for both on cognition. What we're doing with natural history one is that you ran the Bayley-3. And so Pearson, the company that makes the Bayley, has a formula to transform Bayley-3 scores into the Bayley-4 scores. It's the same set of questions. So it's really quite homologous. The only difference is whether there's an extra little tidbit of whether a skill is developing. They already have a transformation. What we're showing you, whether you do it directly or you do it that way, if anything, the direct Bayley looks flatter than the NHS one. We're comfortable with either way you go. It's not going to change anything, whether you're doing transforming Bayley-3 or whether you're doing Bayley- 4 directly. By the time we get done with phase III, we will have a natural history done with Bayley- 4 that's more extensive as those patients, the 40 patients, get through the full year. But I'll say to you, in a randomized control study, it won't matter. We're going to be comparing to our own control. But I think what you can see from all of that is whether you do Bayley- 3 and transform a Bayley- 4, you're getting the same result. These kids don't get better on their own. Okay. Thanks so much. And then second question, I understand sort of the lack of clarity on dosing just for competitive reasons. But any more detail you can provide on the safety issue? You mentioned sort of three to four weeks for the cohorts A and B. But what doses were sort of triggering that? And any additional mitigating strategies that you can implement going forward in phase III? Thanks so much. Yeah. Cohort A and B, C, D, and E are actually all using the same dose. So we set a dose based on what we saw was great efficacy in the dose titration cohorts. We came up with a dose based on those cohorts. They're all receiving the same dose. It's all the same. There's no distinction between the different patients. The fact we see this kind of a little bit here and there randomly and then it seems to resolve and you redose suggests there's a little bit more of a procedural thing than it is the actual dose choice. I think the dose choice is a safe dose choice. We start a little bit lower. They titrate up into maintenance when they go less frequent. The dose goes toward the higher end of the range for the maintenance period, which we said before that the maintenance period is in the 10-14 range. We're comfortable with the dosing. The dose is safe. There's no distinction between having or not having effect. It's just an individual thing. Sometimes it may relate to a particular site or other technical issues. But right now, we're comfortable that it's resolvable. And we get put people back on treatment. And it doesn't recur. Thank you. Our next questions come from the line of Joon Lee with Truist Securities. Please proceed with your questions. Hey. Some awesome data. Thanks for the updates. Did the FDA indicate any desire to coordinate the approval endpoint with Ionis, who may also have data very soon? I mean, if Ionis proposes a different endpoint, how would that connect? Which drug to use? And on MDRI, what exactly is the FDA wanting to see before allowing you guys to use that as an approval endpoint? I have a very hard time understanding their pushback. Thank you. Yeah. Actually, FDA has never mentioned Ionis once. They've not given any indication there's anyone else to coordinate with. I'd look at what we do as a company. We have a whole department that works at endpoints. I think we have a very sophisticated, deep team on this kind of work. We've had our meeting with COA and FDA. It was a very good collaborative meeting. I feel like the endpoints are in our decision to make. I'm not concerned about Ionis at this point. Regarding MDRI, I've been talking to FDA. It was in our MEPSEVII program, you may remember. We did use it there. We just haven't run it in a big randomized study yet. I think it's just a new thing. They're just not sure about it. I think they see the value. I've had many people inquire about it. But putting it primary ends up for them is they're just conservative. And it's okay. We're going to put it in as a key secondary. And that'll put us in position being able to get the benefit of it in our program. So I'm not concerned about it. And as we get it in there and they see how it performs, the truth is, if you want to know how our kids are doing, you look at the MDRI table. And you know how all these kids are doing in one visual look across multiple domains. It's so powerful. Eventually, we'll get them there. But I've been pushing as an individual in my career, whether here, BioMarin, or at Every Life Foundation, we've been pushing the envelope on how to do good drug development. MDRI is one of the things I think is highly valuable. Biomarkers, primary endpoints, I've been pushing on as well. The FDA takes time. They have the U.S. health in their hands. I understand. They want to work quite carefully. We think in the long run, the MDRI approach will become a standard way of approaching these highly variable diseases. With time, the FDA will learn the true value of having an all-comer study and being able to capture efficacy across multiple domains. So we just will continue to press ahead on making that change in the long run. But it'll be in there. That'll allow us. They were flexible having it in there. That gives us a chance to prove its value, and this shows what it can do. I think this will be. I think Angelman is the perfect disease state for this endpoint given its heterogeneity and broad set of issues and broad set of ages that we're looking at. So we're excited about getting forward with it. But I'm not surprised. The FDA is conservative. And they should be. They are responsible for everyone's health. And they want to make sure they get it right. Thank you. Thanks. Next question. Thank you. Our next questions come from the line of Kristen Kluska with Cantor Fitzgerald. Please proceed with your questions. Hi, everyone. Thanks for taking my question. Can you talk about your expectation on a placebo performance relative to what we see from natural history? And I know the primary isn't finalized yet. But we have seen data from these scales that you're considering for other studies and indications. So do you have a sense of that and how it may impact powering calculations? Thank you. Yes. Well, I ran the Bayley, by the way, in a lot of studies in my career. Others, like Alex Greiner, who heads our endpoint development strategy group, we run it a lot. Open label, blinded study. It really doesn't move, right? It really is hard to move. And Angelman patients don't change. So we feel comfortable that we're not going to see a lot of movement in the score. It's really hard to move the Bayley. This is why we're excited about moving the Bayley because she and I both have never seen it happen. And even the FDA was impressed that the Bayley is moving because it's not a very sensitive test. But to get it to move means you have to have a lot of effect. So we're comfortable that the assumption that we put forth are right. I threw in as an example that you could triple the size of the natural history benefit. And you're still well plenty power. So I think we have a lot of room here in terms of power and being able to demonstrate a difference with placebo in this trial. So we're not concerned at this point. Okay. Thanks. And then do you anticipate reporting further updates from these current cohorts of patients? And what else more is important to learn from longer-term data at this stage? Thank you. Yes. We would expect to report more data. This is an interim look. We'll expect to report when all the patients have gone through and with longer-term data. And that could be later in the year. We haven't set an exact timeline yet for that. But we'd expect to have another bite of this data where we have all the patients through certain time points. And we're looking forward to that because I think getting more patients through more time will help solidify the confidence we have in the program. Thank you. Our next questions come from the line of Liisa Bayko with Evercore ISI. Please proceed with your questions. Hi there. Thanks for taking the question. I just wanted to drill down a little bit more on the radiculopathy you're seeing. How is it different from what you saw in those first five patients several years ago when you were first testing GTX-102? And how are you more comfortable with what you're seeing at these dose levels versus before when you did have to reduce dose? And then how are you thinking about incorporating that into the kind of trial design if you do have to sort of pause therapy and redose patients, etc.? How does that factor into your thinking on patient number, etc.? Thanks a lot. Yeah. Well, Lisa, so in the original five patients, four of them had gotten to 36 mg. And one had gotten to 20 mg. So much higher dosing. And we weren't doing Trendelenburg and flush. At least Trendelenburg was not being done reliably. I think what we're seeing here is much milder, much more rapid. There, it took a few months for some patients to resolve fully, many within weeks to months. But here, we're talking about weeks. So it's much shorter. The degree of impairment is much less. So it's a lot more transient. In fact, one of the patients, it wasn't even recognized that he had a problem until in retrospect, it was gone back and realized. So what we're seeing is something much milder and not as significant as it was before. It's still important. But it's nothing like what it was before. Regarding phase III, we're looking at the incidence of this as being relatively minor. And so it's less than a few percent. So we don't think it's going to have a fundamental impact on the phase III. But we are looking at that in our design. We assumed 10% dropout. But keep in mind, if a patient has an AE in a study, they're still going to be counted in the study, right? But a lot of the data, those patients, when they get back on drugs, still have improvement and clinical benefits. I don't think they're going to detract from the data significantly. We're paying attention to it in our design. When we come out with that later in the year, hopefully, that will help explain our approach to managing that. Okay. That's really helpful. Just one additional question. Just historically, if you were looking at the natural history, how well is that tracked with what's been seen in randomized placebo-controlled trials? Has it been pretty predictive? To give us some context for any examples of that, that'd be helpful. Thank you. Right. Well, a lot of people look at the Ovid randomized controlled trial where they had a CGI, a global impression scale score that had a 0.7 change in the placebo group. That's a physician opinion change, right? This endpoint is a psychologist evaluating very specific questions in the kid. It's a very different thing. It's not based on opinion. It's based on what the kid does. This is why I think it's a lot more rigorous and harder to move. And therefore, I think you'll have a lot less background change because it's very hard for kids to do something they've never done before and can't do randomly. So just the rigor of a psychologist test to come out the PI, very specific set of questions, means the natural history, I think, will be more consistent with what you see in a trial and that we wouldn't expect to see a lot of placebo or bias going on. Now, the design of the trial is also not going to be 12 weeks or very short, which was what happened before. The trial will likely be 250-338, which is well past the period of when you normally think of placebo effect or bias being optimal or maximal. I think those are factors that will help mitigate that kind of impact on the trial design. Given the maximum amount of power we have, I just don't think there's going to be an issue even if you had more signal than you thought in the natural history or in the control group. Thank you. Our next questions come from the line of Joseph Schwartz with Leerink Partners. Please proceed with your questions. Hi. Thanks very much. I have a question on safety and then one on efficacy. For safety, I'm just wondering if you could talk a little bit more about the drug administration technique and how important you think this is for avoiding SAE now. How consistent has this been across study sites? What will you be doing in phase III? Sure. For safety, the drug administration method is basically the patients are done. The tap is done. Drug is pushed in. And the flush. And then the patient is put in Trendelenburg within a few minutes. The technique is standardized across all these sites. The fact that we have now 12 sites across seven countries and now with the US, eight countries, being able to do it, I think it's not hard to do. And most of the sites we're using in this phase III trial are actually being essentially prepped to be phase III sites. So that should give us some consistency, Joseph, with what we're doing now and what we expect to see, right? So a lot of these sites will become the phase III sites as well. So that helped us essentially train them, get them optimized technique. It is not an unusual thing, by the way, to do Trendelenburg or flush. So I think we've not seen a problem in applying that consistently. Okay. Then on efficacy, how much will the overall efficacy assessments change between phase I and II and phase III? Are there any additional levels of blinding or any other considerations that we should keep in mind with how efficacy is going to be evaluated in the next trial? No. It will be essentially the same in terms of the endpoints. I mean, whether we change any endpoint or not, I think it would be very minimal, if any. But we have the flexibility to do so. But we're going to be running these endpoints the way we run them. And so we wouldn't expect any other levels of control. It's just a traditional randomized two-group comparison. So it's pretty plain vanilla in terms of design and execution. As a habit, as a company based on our own internal policies and RARE pearls, we always run phase III the same way we run things in phase II in terms of the endpoint, how it's conducted, and what we do to avoid any changes which might induce a new variable. This is part of our policy, how we run phase III in RARE. Thank you. Our next questions come from the line of Maury Raycroft with Jefferies. Please proceed with your questions. Hi. Congrats. Thanks for taking my questions. Kind of a follow-up to some of the others. Just wondering if there's anything else that you need ahead of the end of phase II meeting. And what exactly do you need to align on the endpoint for the pivotal with FDA? And then separately, what are gating factors to redose the two patients in cohorts A and B? Sure. So we'll probably include a little bit of longer-term data when we finally get the FDA. But the data package we have now will be the main data package. A few patients getting through day 254 or 338 will probably include it at that time as closer to the time that we get there. But there's nothing we're really waiting for. We'll be able to get that short-term data I mean, longer-term data at the time we go to the meeting. Now, in terms of getting settled on the endpoint, I mean, I think we have all the information that we need. There was really no fundamental problem with what we're doing from the agency in terms of what we're assessing, how we're approaching it. So we feel we just need to go out and propose, "Here's what we're going to do. And here's why," and show them how the power works and what the efficacy looks like when we get there. So I don't really expect there to be difficulty in coming to that conclusion based on the discussions we've had so far. The other question was on. I'm sorry. What was your second question? Redosing the two patients in cohorts A and B. What are the gating factors for redosing? Yeah. The gating factors, generally, we'll wait to see if there's any protein elevation for the protein from what happens is the drug can have some local inflammatory effect. We see some protein elevation. So we wait for that to resolve. And then they get redosed. Got it. Thanks for taking my questions. Thank you. Our next questions come from the line of Salveen Richter with Goldman Sachs. Please proceed with your questions. Good morning. Thanks for taking my questions. Two for me here. One is on timelines, just an understanding here of whether the trial would start by year-end. And alongside enrollment, should we assume about two years till data? And then with regard to ex-U.S. regulators, is there any synchronization with regard to what you're discussing with the FDA as well? Yeah. So the timeline would be to get the phase III started. We've designed the phase II, as you can see, with many more sites than we really needed. The purpose was to get them set up to be phase III sites, to treat a couple of patients, get the method set up. The hope of that is to be able to activate those sites more quickly for phase III than before. It's part of our learning in the post-pandemic world. By setting them up, having contracts, understanding methodology, they'd be better set and trained essentially for phase III. So a lot of those sites will be the majority of the sites for phase III. There may be a few extra. That'll help get the sites up and enroll quicker. I don't think it's two years from data. I don't want to set a timeline precisely today. But our goal will be, if this drug is working, we'll want to get it enrolled as promptly, get the sites up and enrolled as promptly as possible, and complete the study and clean and lock it as promptly as we can. So I don't want to set a timeline yet for that. But it is important to move this along. It's so valuable for patients and for us to get this done promptly. Thank you. Our next questions come from the line of Jeffrey Hung with Morgan Stanley. Please proceed with your questions. Hi. This is Michael Riad on for Jeff Hung. Thank you for taking our question. Has there been anything to suggest age-dependent differences in clinical benefit, specifically between cohorts A versus B or C versus D? Just trying to figure out, based on the disease progression, is there an optimal age for preventing skills from getting lost versus helping to regain them? Thanks so much. Well, I think it's a really important question. When we first started, I think a lot of people felt that the older kids wouldn't respond. And what we've been surprised with, they actually do respond. And I will say to you, I can't tell you a distinctive difference in response. We have patients responding across the age range. And that's encouraging because I think people thought the older kids couldn't respond. But they actually do in very important ways. So right now, that's why we're including that whole age range. We started wide thinking if we had to, if some didn't respond, we'd narrow it down. But in fact, up to 14, 17 are responding. So then the question becomes, "Well, what about adults?" And someone asked earlier about, "What about younger?" Younger, for sure. But I think you have to look at some adults. I think if you made kids calmer, slept better, more functional, cognitive, that's going to affect someone at any age. What I will say to you is that the theories about human development based on mice have not translated well. And the fact is that human brains are more plastic, more evolved to be able to adapt and change than what any mouse model. And so the mouse model, they would suggest that after a certain age, the benefit would be minimal. It's not really turning out to be true. And I think that's a great thing for Angelman patients. It means we probably could do more. I think there's no doubt treating early would have a better developmental outcome. I think that seems very likely. But right now, we're not seeing a distinctive difference, which is why we haven't shown you A versus B, right, including A and B together because there's really not a distinct difference between the groups. Thank you so much. That's helpful. And then based on that, there wouldn't be any sort of age preference for recruitment for the phase III? It's going to be the same age range. Yeah. And we'll probably stratify the ages just to make sure they're equal between the two groups, that there's not a shift of all young in one group and all old in the other group, certainly. But there's not a great deal of concern about it. You'd like to see an even spread of age across the two groups. Thank you. Appreciate it. Thank you. Our next questions come from the line of Whitney Ijem with Canaccord Genuity. Please proceed with your questions. Hey, guys. Just quick one from me. Have you looked at the EEG data from this data cut? And is it consistent or anything of note there relative to previous updates? Yeah. We didn't present EEG data because the EEG in this part of the expansion study was more focused on, "Have U.S. sites?" It was just not done in all the sites. And so there isn't much EEG data at all from the ex-U.S. sites. So as we get the U.S. sites reporting data on the other cohorts, then we'll have more EEG data. Great. Thanks. Thank you. Our next questions come from the line of Ed Arce with H.C. Wainwright. Please proceed with your questions. Great. Thanks, Emil, for taking our questions. And congrats on the data. Just one question for me. I'm focusing on the cognition as the likely primary endpoint here. And given the administrative strategies used in the more recent expansion cohorts versus the others, where a lot of these domains actually improved, I'm wondering if you can give some thoughts around why cognition with the Bayley-4 was consistent but didn't really improve versus the prior cohorts. And would you expect these strategies to yield some improvements over longer term? Thanks so much. Yeah. It looks like the Bayley-4 is a more difficult-to-move instrument. It's just more rigorous. It's more constrained. It doesn't match Angelman severity, Angelman disease state very well. So we see it operating the same. But the magnitude of effect is still large enough for us to use it as a primary. The ASA certainly is responding more quickly because it's really honed. But if you go long enough, cognition is getting to where it needs to be. And we feel comfortable that that will be true. It may be that for cognition, because it's complex and there's a lot of features to it, that it takes more time to improve, to readily improve. And you can see that curve. It continues to gain ground over months, right, over even a couple of years. So we think that's what you're looking at. You're looking at a complex neurological process moving along. Whereas if you look at a specific thing like sleeping or behavior, those changes are very specific, narrower. And you can detect the change quicker. But the integration of them into a cognitive outcome. They take more time. But we're not concerned about it. The fact it replicated using multiple countries and a broader number of sites tells you that the data is solid. It's not like one site or another site or a country having an effect and the others not. So we feel good about cognition being a reproducible and a highly recognized endpoint. Great. And then perhaps just one follow-up. Given the correlated effects of sleep on behavior, I'm wondering if you've had any specific discussions around that endpoint with the FDA and how you think about it in the overall design for phase III. Thanks. Well, we've talked about all of those endpoints. Behavior certainly is very well respected. There've been other drugs approved on behavioral-type endpoints. So that's certainly something the FDA has seen. Sleep is an area of more complexity for the agency in sleep endpoints. I think in this case, when we talked about it, we were talking not about subtle changes in sleep but profound changes in sleep. And FDA says, "If you're talking about these profound, big changes, they appreciate how that's important because it's an effect on caregivers and everyone else." So the ASA score is really looking at profound changes in sleep, not subtle changes. It's not like 10% more deep sleep or something subtle. It's really, "I'm not sleeping at night. And now I'm sleeping through the night," right? It's that fundamental. And I think that kind of sleep change the FDA is comfortable with. It's not planned to be a primary. But we do think it's an important secondary. And it's important because it responds very quickly, profoundly, and probably is part of contributing to the improvement in brain function that you see through other endpoints as well. Thanks so much. Thank you. Our next questions come from the line of Jack Allen with Baird. Please proceed with your questions. Great. Thanks so much for taking the questions. Congratulations on the data. My questions are somewhat in the same line as the most recent questions asked. I was wondering if you could dive a little bit more into the sleep data and the effect seen there. It seemed to level off with a longer-term follow-up from the cohorts four to seven. I guess, how are you thinking about the impacts on sleep and the long-term ability to continue to improve sleep? Yeah. Well, the truth is, once you start sleeping well, it's kind of hard to keep sleeping more well, right? So when you have kids now sleeping through the night, then they've gotten better. And so then it's a little bit harder to distinguish the difference. So they're getting a rapid response, kids sleeping through the night. And that's staying there. So that's why it's, I think, it's not getting better. I wouldn't necessarily expect it to be better. But what's most important, if you look, is the patients who have really poor sleep have the biggest response to sleep. And we get to patients who have mild or minimal sleep problems. I think once you're there, there's not much more you can get. Does that make sense? It's like a floor effect. You can't get much better in sleep than a certain level of improvement. So we're not concerned about that. I think it's back to sustained is important over a long period of time. I think the data we're showing now, the cohort A and B and the new dose and the regimen, shows a more reaching as good as they get for the entire couple of years in the cohort four to seven, we're achieving in day 170. So we feel good that this rapid effect is showing the drug's effect. And we'd expect it to be sustained and flat. Actually, we're not necessarily expecting sleep to continue to improve over time. It should get very good and stay good. That'd be our expectation. Got it. That makes a lot of sense. Congratulations again on the data. Thank you. Thank you. Our final questions will come from the line of Yaron Werber with TD Cowen. Please proceed with your questions. Just maybe thanks for the follow-up. Just, Emil, on the deletion, so it looks like you're going to be looking at the same 15q11-q13 deletion class I, II, and III. Can you give us a sense of what percentage of patients does that include? And can you get a broad label? Is the label going to say children, essentially, with Angelman? And what's the strategy for adults as well? Thank you. Yeah. So thank you, Yaron. Our strategy on labeling is to anchor the data set with the 15q11-q13 that represent the majority of Angelman patients out there. That will be the placebo-controlled proof of cause and effect for the drug. We'll then fill in the labeling with open-label arm studies. This is something we've done in the past with multiple programs, right? So this is a way you can do it. You don't need randomized studies in young or in adults or in other situations. We think we can do this if we have one high-quality randomized study in the most severe population. Our expectation is to do a basket study where we look at young patients. We will look at patients with the missense and the UPD type as well. We should include another arm of that study, which would have adults as well. The idea would be to show safety in all these populations and to provide support of the efficacy data, which we think, from my history and working in rare diseases, would be enough to get the full label. That's our approach we expect to take. You can run them in parallel? Or are they sequential? Thank you. Yeah. We would run in parallel. We'll get the phase III up and running and then start the basket trial with a separate team so they can run. We might put the basket trial at sites other than where the phase III is to not cause interference. There's a lot of people, a lot of patients with missense or EPD, other types that are pinging us frequently because they want to get in on the story. But we decide to keep phase III narrow. It's our habit. It's part of our RARE pearls. You don't study in phase III people you haven't studied in phase II. And so we don't want to broaden the genotypes and create noise in the system in the middle of a randomized-controlled phase III. Since we're studying the severe patients in phase III, we think that is a rigorous test of the drug, right? It could be with missense patients. They might have a better effect because they're, let's say, closer to the edge of being a normal, right? They have some activity, not none. So in EPD, there's two chromosomes that turn on. So that could be also a benefit. But we'll study those. And we'll be running that in parallel. We'll start the phase three first as first full core, probably set up a separate team to work on the basket, include some other countries in the basket trial as well to get broader international exposure. And there's a lot of interest internationally getting part of that program. Thank you. We have reached the end of our question-and-answer session. I would now like to turn the floor back over to Joshua Higa for closing remarks. Thank you. This concludes today's call. The slides from our presentation have been posted to our website. If you have further questions, please reach out to us via email at ir@ultragenyx.com. Thanks. Thank you. This does conclude today's teleconference. We appreciate your participation. You may disconnect at this time. Enjoy the rest of your day.