Ultragenyx Pharmaceutical Inc. (RARE) Earnings Call Transcript & Summary

I am Yigal Nochomovitz. I'm one of the biotech analysts here at Citi. I think, Emil needs no introduction, but I'll give a brief introduction, CEO of Ultragenyx, founded the company, I guess, in 2013. Was it?

2010.

Yes. So welcome. Great to have you back, this time in Miami. Hopefully, we'll be treated to better weather than up in New York.

So well, maybe we can just start with maybe a question that you don't usually get from the analysts, which is why did you found the company? What was your thinking back in 2011 or whenever it was originally started? You were BioMarin. So tell us why you felt we needed another a new rare disease company, and what are you going to do differently?

When I left BioMarin in 2009, I began having large numbers of the families trying to develop their own drugs come to me, ask for my help trying to develop their drugs and it became clear to me there's a whole lot more to do and that I couldn't really walk away from it. Although the financial crisis had just happened, right? And things were tough, financing was terrible. I said there was probably room for another rare disease company to get going. And I found an asset that made sense as a simple product we could put forth. But starting the company, I felt like there was an opportunity here with all that's been learned about rare is to establish a company that would have -- be a next-generation rare disease company that would take all the learnings and design a company to operate in those -- in that way. And I formed the company in 2010, me and the secretary, a couple of million dollars and started the process of putting together a portfolio. The idea is to build a next-generation company that had developed the right strategies around picking assets, potent biology, bad diseases of varying size and market, but ones where we thought there would be a dramatic benefit. And usually, it's disease that had no treatments at all, that would be really highly focused on how we run development, adaptive designs, efficient designs, time line as being a key part of how we control and execute, developing new endpoints, having a department that ran the creation of endpoints, and pushing toward approval very quickly. And finally, how we'd go about commercializing, doing a more effective way of managing patient find, which is a key part of rare disease commercial launch. And then thinking through pricing and reimbursement strategies to optimize the patient experience and the physician experience as well as assured that we got majority access of patients on drug. Putting those elements together, we were able to put -- we thought to put it come together, take the best of each of those strategies and establish a new paradigm for how rare disease drug development should happen. I think between 2010 and 2014, we put together a portfolio of 5 programs, we've got Phase II clinical data. And between 2014 and today, 2024, 10 years on the public markets, we have 4 approved products. We have a fifth product that we acquired, sixth product in advanced stages, potentially 3 BLAs in the next year. And in the history of rare disease biotech, that's far past what anyone has achieved, and we've done a number of programs in the 3- to 5-year time line. Some of the gene therapies have taken longer. And we're also doing multiple modes, best mode for each disease state and all the strategy that put us in a position of not only having a company that's having $530 million to $550 million revenue, heading to profitability in a couple of years, but also the company that sets standards in regulatory policy, comp use, compassionate use, commercialization strategies, and I think being the true leader of the future rare disease medicines. So I've been very proud of what we've accomplished. And I think we're at a great inflection point right now with both UX143 for setrusumab for OI, coming to a Phase III interim assessment soon with 2 gene therapies coming to file for 3 BLAs potentially in the next year. Add to that Angelman Phase III is an extra. Combine that with other gene therapy programs and a global commercial enterprise that could take advantage and sell those programs globally. I think we're as well set up as any rare disease company that's out there.

Okay. That's a good intro. So you mentioned OI, so maybe we can start on that one since there's been a lot of questions there. The first of 2 interims is coming with end of this year or early next year. So just to give a bit of background, what's the standard of care. You have the 2 Phase III trials, Orbit and Cosmic. What are the goals of each of those? Maybe just a quick on the mechanism on anti-sclerostin and then we can talk about other things.

Well, osteogenesis imperfecta is a defect in collagen, and it can be either a deficiency, giving your type 1 osteogenesis imperfecta or an altered type of collagen that interferes with the triple helix of collagen, which causes bones to have become dysregulated. Now the collagen, it's like the rebar bone with cement, the bricks are the hydroxyapatite crystals, the glycosaminoglycans is the mortar and the rebar the steel, that gives it tensile strength is collagen fibers. So the idea is if the collagen fibers are weak, then the bones are fragile. It turned out though that humans are not like cement, their bones. But what's actually happening is the defective collagen or the defective bone is inducing bone dysregulation. The bone is maladapting to the defective collagen fibers by trying to resorb the bone and lay down new bone to fix it. The net effect is that you have too much resorption, not enough bone production and the bone mineral density of OI patients is low. And we're changing the paradigm and thinking that what's really happening is that the low bone mineral density is a bigger factor in the fragile bones than it is with tensile strength. There's some technical reason why we can explain that. With that in mind, then the question is, how would you fix a patient who's not making enough bone? And setrusumab and anti-sclerostin, it binds the sclerostin, which is part of the WNT system that regulates bone production, how your body dials up or down bone production. Your body says, "I want less sclerostin and more bone production or I want more and less bone." So it's how your body tests the dial. What we're doing with anti-sclerostin is change the dial from excess resorption in OI patients to be pro production and less absorption. The net of that is bone mineral density improvements that are dramatic. For example, in a 5- to 12-year-old, 29% increase in bone mineral density, a change from minus 2 Z score to maybe minus 0.58 Z score, nearly to normal within a period of 1 year. So a very dramatic effect on bone regulation in these patients. The current standard of care for OI is bisphosphonates. The bisphosphates affect only the resorption side of the equation. They don't improve bone strength. And what's very clear in the animal models and humans is while bone mineral density may increase with bisphosphonates, bone mineral density is not well placed. It's there, but it's not where the bone is weakest. When you use an anabolic agent, the bone is laid down where the bonus is weakest. And so when you look at animal model or human, that the increase in bone mineral density achieved with an anabolic agent correlates with both strength. And in animal models of OI, you can normalize bone mineral density and normalize bone strength. So in an animal with a mutated collagen, that should be like a type 3 patient mutated collagen in the bone, if they have bone mineral density that's normal by an anti-sclerostin mechanism, their bone strength on a three-point fracture breaking is the same as normal. That was the study that said, "Wait a minute, do we really understand OI? Is it really a mutated collagen or is it because they don't make enough bone of the right bone [ laid down the road ]. So that led us in this direction to pick up the asset and to do the work, put it in the children. And what we've seen is a dramatic improvement in bone mineral density and very substantial reduction of fractures, which was shocking because this is a paradigm shift. It's changed when everyone thinks they knew about OI. And the good thing is that if this is the problem of the way, we actually can fix this because fixing collagen expression in the osteocytes in the middle of your bones is a very hard thing to fix. But if we don't have to fix it, we can fix it just by making more bone, we'll get the majority of the benefit by doing what we do with anti-sclerostin. So it's that insight that put us in a position to have a transformative therapy for OI. We've been very encouraged by what we've seen in Phase II data and Phase II extension. And we're coming ahead to the interim, the end of the year, early next year for the first interim, 0.01 p value. And if that doesn't hit, a few months later, we'll have the last patient in with an analysis sort of towards the middle of the year for the second interim and the final will come towards the end of the year, would be the 3 steps. Why are we doing interims? Well, our view was that if the fracture reduction is as great as we've seen it, and if the fractures were increased at baseline, the possibility of hitting a 0.01 p-value was there early on. And we felt with that kind of powerful change that we would push for ending the study earlier, patients shouldn't continue on placebo, children on placebo for no benefit if you've already separated. So we're able to get the regulators on board of that. If that doesn't hit, we won't hear anything, get toward midyear, where we'll have a 0.01 p value test. At that point, everyone would have at least 12 months of therapy, which was a more of interest for the regulators. And we could then -- that's probably the most likely place to hit result. And if not there, there's additional 0.04 test towards the end of the year at 18 months. We're encouraged with what we see. I think the biggest factor of whether we hit first or second interim will depend on the fracture rate and a little bit about how much the fracture reduction is. At the fracture reduction level, it is 70%. It's going to be more driven by the fracture rate and the variation between patients and how variable it is. But so far, we've been encouraged what's going on in the long term, the data and who we're excited about having an opportunity to treat OI. For us, commercially, it fits very well what we've done with Crysvita. We know these doctors, 90% overlap in the KOLs. And I think given this biphosphates are off-label, their fracture effect is only 20%, that a drug that has a 67% reduction will be dramatically better and would become, we think, standard of care if studies are positive and drugs approved.

Do you have any sense of the -- on a blinded basis, what the run rate is on the fracture rate or you don't have access to that?

We don't have the blinded data. We wouldn't put that out in any ways. When you're running a randomized control Phase III, it's really not good form to constantly be looking at the data. I think FDA looks at that as suspiciously. So we're not reporting and I haven't seen it. Our expectation is the fracture rate is at least as high, if not higher than it was before. The 0.7 rate is what we estimated before, but we know there's more type 3s and 4s in this study than that we're in Phase II, right? And those patients would have higher fracture rates. So the expectation is the least is higher or higher in the Phase III. Is the question is how much variation is there? We think from what we've seen is that within a few months, the fracture seem to stop or go down dramatically. So we think the groups are going to separate within 3 months, let's say. But how separated will depend how much variation is there, right? So that's why we won't know whether it will be the first or the second. But second is the most likely at 0.01 p value, I think we're a very good chance of hitting that, but we have a shot at hitting the first one.

And then as far as the disclosure, my understanding is that if it hits, obviously, you'll tell us. But if it doesn't, then you just -- you don't comment. So I guess the thinking will be if you don't comment after a period of time, we'll just assume that we're on to the next one?

Yes, it has to do with impacting the conduct of the study and so this is why it's not because we're trying to be difficult. We want to minimize the impact and press releases have potential impact. If the study is ending, we won't announce the result, we'll simply say the interim was hit and the result won't happen, and this is by agreement with the FDA until we complete all the last assessments, which takes a couple of months, clean and lock the data set and then we release it, right? So it's a few months of time before hearing an interim was hit and actually seeing the data. During that time, we would start preparing. We already have begun preparing what we expect to be a BLA for the program. And we'd expect to file it sometime in the second half, but as early as we can.

And can we just quickly go through the differences between Orbit, which is main one and then Cosmic, which is I believe you've indicated in the past that you might get the Cosmic first, but Orbit is going to basically drive the filing?

Orbit is going to drive the filing. It's the main placebo-controlled trial, 159 patients. But the Cosmic study is a head-to-head study. So Orbit is a placebo-controlled trial of sclerostin, 2:1 versus placebo. The Cosmic study is a 1:1 randomized and involves 2- to 4-year-olds primarily, although there maybe some 4- to 7-year-olds depending on the location, the 2 to 7 range, so mostly younger patients. That study is randomized controlled against bisphosphonates. And the idea in that one is to be able to show what the relative effect. It's not blinded because blinding the bisphosphonates, depending on which one they happen to be on was too difficult to do. So it's an open-label study, but randomized. We are not looking at the data repeatedly of not seeing the data. The study though, in the young patients, we expect them to make bone mineral density much faster than the older kids, right? So I expect them to respond more quickly. In fact, if an Orbit hits an interim, then we'll look at the Cosmic study. Our expectation in agreement with the FDA that we'll file with both sets of data, I'd expect Cosmic to hit if Orbit hits, right? It is a smaller study, and it is against bisphosphates, but its younger patients will have more rapid effect, and that population has a higher fracture rate than Orbit, right? So higher fracture will give them more power to show a difference.

Okay. And then I get this question often from people that are new to the story. I'm sure you get it too. Obviously, Amgen has romosozumab and Evenity is the trade name. They have expressed no interest in OI and there's some ceding of the rights to you in some respect. Can you kind of walk through that and just explain why romosozumab is not a threat?

Romo or Evenity is Amgen's brand for osteoporosis. I think they're shifting their focus on osteoporosis toward anabolic agent and Evenity is their flagship product for that particular indication. That's an indication they see as treating millions of patients potentially. That's not what OI could be, right? Opioid is a very tiny thing. They're having to do some studies in OI as part of a PIP commitment or a pediatric investigational plan for Europe, but the dosing they're doing here is much lower. It's efficacy and bone mineral density is much lower than ours. Therefore, we think they're going to see some data, but it won't be as effective, and I don't think it will matter. So we don't look at them as a competitor. To substantiate that further, if Amgen did give us a nonexclusive license to the epitope patent, which was a point of potential contention on FTO, but we were able to get that. And I think they wouldn't have given to us if they were interested in the indication, right? They wouldn't have given up the patent position. That patent does expire. It's a fee-based, not a royalty patent. I think is an indicator of their interest in OI versus ours.

And then the blueprint for the launch, assuming you hit your endpoints in Phase III, you mentioned a lot of overlap in the prescriber base with x-linked hypophosphatemia. So can you just talk a little bit more about how that would play out?

Yes. So the prescriber base for the XLH is 90% overlap with OI. So we have an extensive experience and relationships with those doctors now. And so I think we're very well set up. We understand them. We understand their clinics, where they are, how they operate. What we know from a lot of the major centers is that they have 50% more OI patients than they do XLH, something like that. Some have twice as many. There's more OI patients than XLH. Now the frequency of the disease are not so different, but at these centers, there's clearly a lot more OI patients, probably because they're more severe, and therefore, they're going to more specialized care. What that means then for launch is that we should be able to grow faster because we have now these key opinion centers that have a highly enriched population for OI. Our experience with them and reputation we established in our work with Crysvita and XLH, we'll carry forward to the policies and approach we take for OI. They were highly appreciated. The plan gives co-pay support for patients. If patients have insurance issues, we will get them a free drug and take care of them in a way that allows us to grow the business. We did that successfully with Crysvita. Our reimbursement people were excellent. We got reimbursed. But in the meantime, we take care of them with the full-service approach they deserve for a terrible disease, and that was highly appreciated and we didn't let people drop. I think doctors these days are abandoned a lot by the system and left to struggle. And what we want to do is make sure that we're partnering with them and our patient services hub ready to help solve problems, get things done for their patients.

Okay. Let's shift over to gene therapy then, but before we get into the detailed programs, just at a high level, how do you think about which diseases make sense for Ultragenyx to pursue. You've got some good ones here, but what are the criteria that you use to evaluate whether to go into a certain rare disease for gene therapy?

For gene therapy?

Yes.

I think the whole critical thing here in our view is that it needs to be an urgent indication that's severe for which there really isn't a good solution, the gene therapy provides the first really good solution. And when we acquired Dimension, they were working on heme A with Bayer. But the problem with heme A, while there is a lot of patients, there's a lot of products. There's an established base of products, a lot of complexity. So creating a better financial model maybe easier, but it's not the same dramatic situation as you were with perhaps a smaller indication that's more severe and urgent. And if you look at success of gene therapy today, it's pretty obvious to me what the successes have been urgent, no other solutions, right? So Zolgensma for SMA; Duchenne -- the Sarepta gene there before Duchenne, very urgent serious diseases, Hem A, Hem B more difficult. Sanfilippo IIIa is small, but it is urgent. There are no good solutions and there is -- the patients are going to die and they have a narrow window of time to get treated. It's very similar to Duchenne in that regard, right? There's a narrow window if you don't treat by a certain point in time, you may not get optimal result. So we think that Sanfilippo will be the kind of urgent indication that we'll do well. And it's not big, but we were able to manage to get through with a relatively small set of data closer to what Zolgensma had with mechanism, IV, AV9 that's very similar to Zolgensma. So I think it will do well, but it's a much smaller product than, of course, Zolgensma. The second program we have with DTX401, GSDIa. These patients are living every day with the fear of dying with hypoglycemia crisis. They take starch every few hours to remind them what they have to do. It's very urgent. They have -- are living with a gun to their head. And if they forget to take their starch, they could die. It's a stress factor. It's running around on a treadmill. These patients are very urgent in our trials. People want to get off that treadmill and get on something that gives them control of their life. And the gene therapy, we've been showing, I think, a very good effect, and we've seen considerable urgency and clinical benefit in people's view of how reducing starch dependence and also what I would call the risk of dying. These patients during the night, their glucose will not bottom out. They will stabilize during the night after the gene therapy instead of crashing as we normally do. And that alone gives you the confidence in saying, if I forget to wake up or I stop waking up for taking starch, I'm not going to die, right? If I go exercise, I'm not going to drop and have a seizure because I forgot to take something that my body is now going to start adapting to my glucose needs. That freedom and that sense of comfort is extremely important in the story, and that's what these patients are seeking in our trial, enrolled very quickly. We had a lot of demand. Certain countries that took too long to enroll the trial enrolled, like France. The France regulatory system took too long, the trial ended enrollment, and we told the French site it's too late. We are closing the site. The patient community found out about this. They were so angry at the regulators and called the regulators. The regulators had called us and said, you have to open the trial again for the French patients. And we know it's too late, you took too long. And -- but I've never had patients get so angry at the regulators to call us, to demand that we open the trial, right, at their country. But the next trial that went through France, it went very quickly, by the way. So I actually think there was some net learning. They said, we're going to -- we made it competitive. They said if it enrolls, it's enrolled.

So just -- so people understand, with the standard of care with the cornstarch, are they setting alarms in the middle of the night to wake up? Like what if they don't take it? What is the risk of dying on a given night or any night, just in general?

Usually, each family will set 2 alarms. There's 2 alarms going on to wake up: parent, child if it's a young child; the older child, they have 2 alarms. They routinely set up 2 alarms in case there's a fail. It's like a backup. Every year, this is something Dr. Weinstein -- David Weinstein, who did a lot of the work and is a major GSDIa physician would tell us that every year at the GSD conference, there would be kids who died during the year because they forget, didn't wake up for their alarm, who are doing well, they missed the alarm, both missed, the kids died -- in the morning, kids died. But you don't recognize when your glucose has no support, you crash, you go all the way down to 0. And they essentially, it's like giving someone insulin, you crash your glucose to 0, and they cannot recover it and then he will die.

So it's more of a deterministic thing, like it will happen.

Yes, it will happen. And so people learn because every year, they go to the conference, they find out, "Oh, my god, another person is gone" right? They hear about the cases and these are cases that we're doing well. So all it takes is one slip up. If you talk to a parent talking about their kid with GSDIa and they talk about this, they start -- you can see them start to tremble as they're talking because you can see they're getting like a PTSD response. You imagine you're a parent every year, you're worried about your kid doing every day. It is horrific. And imagine that you give them something where that fear is taken away, but they're not going to crash and burn. And so it's a little bit less about how much starch you're taking than it is, do you have a safety net for your biology now. And that's what I think the GSDIa gene therapy provides.

So you will have a pre-BLA meeting coming up. What are the goals there? And what -- I mean, the label seems pretty straightforward, but what do you want to achieve at that meeting? And what's the market look like for this product?

Well, we're comfortable we have the data to file, but normal pre-BLA meeting is just going through the checklist of things that we expect to have. We'll talk through the Phase III data that we'll have. We'll also talk through crossover data. We decided to move our manufacturing internal to our plant in Bedford away from the vendor, gives us a 40% cost reduction, and it's just a better way to start the plan -- with the plan. Because of that, our [indiscernible] delayed our filing a little bit, but that gives us the opportunity to collect more crossover data. So we'll talk to them with what crossover data we'll have at that time. It'll likely be additional 96 weeks for the treated patients plus -- sorry, additional 48 weeks for the treated patients plus a significant amount of crossover data for placebo that will help put the picture of durability and the treatment effect because all patients in, in the Phase III will have been treated at that point. We'll have significant data. We did put out data in our Q3 announcement, showing that the crossover patients who are now getting their glucose data so that they could reduce their starch more accurately rather than being told about their starch. They achieved a 61% reduction within 30 weeks, which is essentially double the reduction we saw in the Phase III at that time point. It just shows you when the patients have the knowledge of what's going on in their body, then they'll -- they can reduce their starch more.

So that additional [indiscernible] they know there's a back step?

Well, they know what their sugar is. And what was happening during the trial is they had a CGM monitor on them. It was going to the cloud. The cloud would send a signal to a doctor who is unblinded. That doctor would get the cloud saying, "Oh, they're hyperglycemic. They're at 160." He would call their PI and say to the PI reduce their starch by 25 grams or something, 30 gram. That he would then call the patient and say, "Hey, reduce your starch by 25 grams." Patient says, "What's my value?" We can't tell you that. We just wanted to do this on faith. The patients are thinking, "Well, I felt low, though. Why should I do that. I'm not going to do that." So they didn't realize the reason they were low is because they had been high and then they had insulin reaction and it dropped them low. And so they feel the low, but they don't feel the high, so they were reluctant to follow instructions without knowing. This is what we learned early on with CGM monitoring is that a lot of patients were complaining of lows, but if you looked at the hour or 2 before, they actually were hyperglycemic. And what was happening, they're getting a surge in insulin production and a crash. And so once you learn that, you know to react to the peaks, right?

I see. Okay.

Yes. So these patients are all hyperinsulinemic, it's one of the other learnings. So you have to consider the context in understanding how to manage them. So we're comfortable how it's working, and we're excited about it. And I think when you look at kids during the night, you see their glucose come down and they flatten out. 4:00 in the morning, they're turned flat. Their body is putting out glucose. That to me is it, that's everything. That's what I look for. Because I know that kid is not going to die. That's what the gene therapy is about, not let kids die in the middle of the night.

They're not setting the alarms anymore.

They don't have to do.

Okay. Let's keep going because we've got 10 minutes. So Wilson's. What's the path? You've shown some nice data already. People weaning off the chelators, some fully, some partial. What's the strategy there? And is there much of a competitive dynamic, it doesn't seem like it, but apart from the chelators?

Yes. So we've put out data on 3 different cohorts at 3 different doses, 5 patients each. We had 6 patients wean off their chelators, which is encouraging. We had signs of detoxification of copper as well as improved copper distribution, but it wasn't as potent or as complete and consistent. We want to see the majority, the vast majority of patients get off chelators and have strong biomarker data. So we felt like we probably needed to go higher on the dose. This is a transporter, not an enzyme. So the transfer needs to be in as many cells as possible in order to pump out the copper. Clearly, it also -- the effect takes months. It could be during those months, the copper is getting moved out and it may take time. But our decision was we need to do another cohort and go up in dose to try to optimize the effect. So that's what we're doing. We're doing another cohort. We didn't want to jump into the investment in Phase III unless we had a gene therapy that was going to clear the need for chelators in the majority of patients. Otherwise, a little hard to justify why you're doing a gene therapy. But we are encouraged that the drug -- this can work and that we have an opportunity's then to change the future for Wilson's disease patients. Currently, the 2 competitive programs had been the Alexion Chelator, which they had dropped, now got picked up by another company, but we think the chelators are an old story, and I don't think a higher affinity chelator is going to actually change anything. The second was the [indiscernible] gene therapy program, and they have announced that they're dropping their program going forward. So I think right now, it's all ours to get...

Which tissues typically are those the storage tissues, the excess storage tissues and where are you seeing expression?

Well, the liver is the place that fills up with copper and becomes toxic. And then as it fills the liver with copper, the copper oozes out and the free copper ends up accumulating the neural tissues and cause the neurologic symptoms. So those are the 2 primary places. The copper is -- free copper is going everywhere, but there's also the problem of copper not being distributed correctly because the total copper level in these patients in the serum goes down actually. It's very confusing, but the reason it goes down is that they're not loading copper on ceruloplasmin in the liver. What liver normally does is take the copper, put it on this protein, ceruloplasmin, and send it out to distribute to the brain, but that's not happening. So the actual total copper -- most of your copper in your bloodstream is bound to ceruloplasmin, not free. Free should be very low. And these patients have a very high free level, but it's still a small percent of the normal level of copper, but the ceruloplasmin of copper has gone down dramatically. So they're not detoxifying copper through the bile. They're also not copper distributing correctly, both things are wrong. For the first time with gene therapy, we can actually fix both parts of the copper story. Both the detoxification story, enhancing delivery to the bile and eliminating excess copper, but also the ceruloplasmin distribution story. So that's why gene therapy could be a superior way of doing it. And we know, for example, some patients who have CNS symptoms that when they get a liver transplant, their CNS symptoms get better. Now that's not likely copper toxicity. I think in that situation, it could be copper deficiency because copper is needed for a number of CNS enzymes. And I think what you have in Wilson's is a very complex thing of copper toxicity and copper efficiency combined back in the same patient at the same time. So we're learning a lot about Wilson's. We'll go to Cohort 4, optimize the dose and get data on that and then would proceed to Phase III. We've got plenty to do right now, and I think the big catalysts for the company are the OI program, UX111, UX143 for filings and then Angelman Phase III getting enrolled.

One more on gene therapy, and then we can quickly touch on Angelman. OTC, you spend less time on that. The Phase III is underway enrollment. When is the data coming because we don't really talk about that one too much?

We don't. We have a lot to talk about. OTC is enrolling. We're trying to finish the enrollment this year or early next year. But we haven't talked about the time line. That patient -- from last patient in that study takes 64 weeks plus data analysis, right? So even if we finished enrollment in Q1, it's a year and a quarter plus, right, before the data come out. So right now, it's a little further out. But I do think it's going to be important for that disease.

Okay. And then Angelman's quickly, and then we'll do some commercial. So you picked [ Bayley 4 ] raw. Just explain why you didn't do the Bayley growth scale value, the GSV and your competitor picked one piece of that -- the one piece of that domain, the expressive communication, I believe, so you didn't do that. So just explain -- and of course, you're focusing on the deletion as the primary?

So the folks in the Bayley, I mean there's always a negotiation discussion with the FDA. Among different types of endpoints, they prefer an endpoint that has an independent valuator, let's say well-established validated tool and the Bayley score has been around for a long time. It's designed for normal people. It's not designed for Angelman. So it's not a perfect test. It's not perfectly sensitive. But it's a validated tool, has a third-party evaluator, a psychologist evaluator. And so it's an in-house tested thing rather than a patient-reported result in general. That's for at least the cognition that very few PRO, patient-reported segments to the Bayley for cognition. So that's why the Bayley and the Bayley cognition, we've shown a very nice effect in the cognition. We think it's a little bit of a synergy of the improved receptor communication of the patient, the cognition in patients and also their behavior. They're calmer. They more listen so that the psychology valuation is better, because they're actually paying attention. So when you look at Bayley 4, you're actually looking at the benefit across a few things. Domains helping that performance. That's why we think we're seeing very good effects. The Bayley 4 expressive communication moves less. It's not well designed for Angelman. It's in there. We are evaluating it as well, but we didn't think it was -- made sense as a primary because of its potency. Now with regard to Bayley scores, the GSVs, growth scale value scores are adjusted for age and sex and other things. So it's a normalization strategy. FDA doesn't like data transformation, right? Any kind of transformation of data, FDA is suspicious of. So they want to stick to raw scores. And we put out at our fast our comparison of raw to try to give people comfort that the raw score, if anything, is more powerful than the GSV score. So it doesn't matter. You could do it either way. But there was apparently some anxiety about going to raw scores and we just wanted to eliminate the anxiety. It's perfectly fine to do raw. So we'll be doing raw scores for Bayley 4. I'll point out though that we did get acceptance of FDA to have a key secondary endpoint, the multidomain responder index, which has 5 domains. We think that's a better representation of Angelman than just one domain. The 5 domains also give us dramatically more power, and we will allocate 0.005 alpha to that. The FDA has agreed to that. 10% of alpha will be allocated to that. That will give us another way to succeed in the trial. It's been an extremely powerful way to analyze Angelman. I want to press it ahead because I think it's a new way forward in complex multi-domain neurologic disorders. We have to find better ways of assessing them than rigid scores that are relatively insensitive and don't handle heterogeneity very well.

Okay. And then just very quickly, obviously, there's one competitor that's relevant. You have the...

[indiscernible].

Okay. We can discuss, but you got the IP from Texas A&M from the inventor there. So just -- it would be very interesting to just talk about quickly the strength of that IP and the reason why you think it is the superior molecule?

Well, the work that Scott Dindot did helped us find a lot more of the molecular genetics of the Angelman locus that then was understood from the mouse. And he determined there was a region closer to the [indiscernible] antisense which was a better targeting place for knocking down the antisense and knocking down this transcription. And that potency showed us both the non-human primates in vitro and it was that data that got us convinced that we should work on that molecule. Because normally, I wouldn't go head-to-head with Roche and Biogen, Ionis on neurology. That wasn't -- didn't make sense. But I strongly believe that sometimes there's one person, smart person that can be 10,000 people in the company, and I think Scott is one of those guys. He did really excellent work. And what he did is turned out to be true. That is the molecule is more potent, by several fold more potent, and it is performing well. And the science has proven case, and we feel we're in the lead. I think we're -- I think we have -- I think our future is in our hands with this trial and getting it done. And the exciting thing is that when we started the program, we had no idea whether you could actually cause the development of a trial to start to accelerate. That's not been done. Child development had not improved. We're mostly trying to prevent declines or loss. But here's a situation where you actually cause cognition to occur for a patient who here gibberish for language, now to understand 1 d 2-step instructions and respond to them, to express their needs when they couldn't express before. These changes are fundamental, opening our minds to the dynamic and placid nature of the brain. And that fact, the neurons are there that we tweak them with the molecular scalpel and turn on gene expression in a very specific way. I think it's a powerful result, and I'm excited to be part of it. And I think Scott Dindot work is going to turn into a paradigm shifting view of human biology in the brain and maybe some amazing things we could do with some other diseases, but we're excited about it.

All right. Thank you so much, Emil. Appreciate it. I'll have to save the commercial discussion for the earnings call. It's going well.

Yes, it's going well. Okay.

All right. So thanks again. Appreciate it.

Thank you, Yigal.