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

Ladies and gentlemen, thank you for standing by, and welcome to the Ultragenyx GTX-102 for Angelman syndrome interim data and program update Conference Call. [Operator Instructions] It is now my pleasure to turn today call over to Joshua Higa, Executive Director and Head of Investor Relations. You may begin.

Thank you. We issued a press release and posted slides that detail this interim data and program update, which you can find on our website at ir.ultragenyx.com. Joining me on this call from the company are Emil Kakkis, Chief Executive Officer and President; Camille Bedrosian, Chief Medical Officer; Scott Stromatt, Chief Medical Officer of Neurology; Mardi Dier, Chief Financial Officer; and Tom Kassberg, Chief Business Officer. I'm also pleased to be joined on this call by 2 investigators on our study: Dr. Erick Sell, Associate Professor Neurology at Children's Hospital of Eastern Ontario in Ottawa, Canada; and Dr. Elizabeth Berry-Kravis, Professor Pediatrics at Rush University Medical Center in Chicago, Illinois. I'd like to remind everyone that in today's call, we'll be making forward-looking statements. These statements are subject to certain risks and uncertainties and that 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.

Thank you, Josh, and good afternoon, everyone. We're here to talk today about the GTX-102 midyear interim update and program update. There is a slide show that we'll be referencing during this call. On the second slide is our legal warning, which Josh has just provided. On the third slide is a disclaimer that GTX-102 is an investigational product without proven efficacy and safety by FDA or any other authority, and these are interim data update for an open-label study and specific limitations of that design and ultimate proof of safety and efficacy will depend on the conduct of an adequate randomized controlled study. Move on to Slide 4 are the group that are here today. I won't go over that again, but we have a great team of people to help answer your questions later in the call. On to Slide 5. What we're showing with you today in our release and the slide deck is that we're seeing promising clinical data for GTX-102. We haven't seen clinically significant safety issues, AEs or SAEs that relate to the drug up to 10 milligrams using the new administration strategy. No lower extremity weakness and no pattern of increasing inflammation, which is extremely important. In addition, at these lower doses, we are seeing meaningful clinical activity in multiple domains on multiple measures in most subjects with 3 and more improved domain. So we're comfortable that we're seeing strong changes in results in these patients that are drug specific, and we're encouraged by what we see so far. With the data we saw early on, we proceeded to actually initiate a Cohort 6 and 7, in which we started the new higher doses, that we have shown are not causing any inflammation, the 7.5 and 10 ml doses, we've already cleared that with U.K. and Canadian authorities and in fact, have dosed the first patient with Cohort 7. And so we're moving ahead already with loading doses starting at these higher levels, which we expect would achieve higher accumulation and improved -- further improved efficacy. Finally, with the strong data we have so far, we have triggered the GeneTx acquisition, and we'll talk more about that in a little bit. So let's dive into the story. On Slide 6 is a summary of the protocol. I won't spend too much time. I'll need to, however, remind you there's a Cohort 4, which is young patients; the Cohort 5, which is the older patients above age 8. And that in the Canada U.K. protocol, we're doing a dose titration for all the patients at 2 different ladders, and there is a Trendelenburg and flush component of how the administration is being done. In the U.S., it's only the younger patients that are at the 2 mg dose and maintained without dose titration. On the next Slide 7 is the general outline for dosing, and this is important for you to kind of gain understanding of where the doses are and where are the evaluations that we're going to be talking about. There's a baseline evaluation after which dose 1 is given. During the monthly loading phase, there are 4 doses in general that we planned at monthly intervals. Before the dose 3 is given on day 58, we will check for whether the -- what the clinical changes are seen, and if they're sufficient to keep the dose or escalate the dose. And at day 128, after the fourth dose, we do a major assessment evaluation, which is the major body of the data we're presenting today. Some patients have progressed to the PMD or maintenance dose evaluation at day 170. Between day 128 and 170, there's no more drug given, there's just more time of exposure to the existing loaded drug. Subsequently, they will be getting every 3-month dosing. The furthest patient in the trial has reached actually the second PMD dose, but all patients have gotten through the 4 doses at this stage. So that's sort of the dosing schematic, we'll be referencing those at times as we go through the deck. Slide 8 is the summary of the dose exposure information. I don't want to go through all of this with you, but just to highlight a few of the key points. In terms of the cumulative dose, we have patients going up to 32.4 milligrams and 7 of 14 patients have 20 milligrams or more, which was the level at which the -- the lowest level which safety that was seen before. In addition, we have patients now gotten up to 6 doses without and some of those exposures, 13 of them are greater than 147 days at which we had previously seen a safety event. 8 patients have now more than 6 months of exposure. If you look at the maximum dose exposed through the titration, we have now patients getting 10 and 7.5 mg of dosing at their most recent doses without the adverse effect. So we're pretty comfortable now that both the dose and exposure were showing that this drug at these doses and with this administration strategy can be done with reasonable safety. Now safety details shown on Slide 9. In Slide 9, we show there's no drug-related SAEs and no AEs or SE reports of lower extremity weakness. There have been some unrelated SAEs, which are of no significant consequence. And the most common AEs, you can see they are in a minority of the population, therefore, we're very comfortable with what we're seeing regarding the safety experience so far acceptable. Importantly, when we look at CSF total proteins, you can see for half of the bottom of this page. We're seeing very stable CSF protein. No trend of improvement, no signs of the inflammation that would result in increased CSF protein, which might indicate a problem that was observed last time. There is 1 patient that did have asymptomatic infection leading to a rise in CSF protein, which was transient in the result, but unrelated to the drug. So we're comfortable so far with the safety profile and the lack of inflammation as measured by CSF protein. Now let's go on to Slide 10, which is the efficacy. In the upper panel is the Cohort 4 and in the bottom panel, Cohort 5. 7 of 9 patients show improvement in at least 3 domains. So we're seeing multi-domain improvements across many patients. And what is very notable, of course, is the sleep has very strong effects with 4 out of 6 patients showing a significant effect there. But many of these patients have multiple domains, and we're seeing effects in improvements across many domains in the setting. In Cohort 5, similarly, multiple domains are improving, and this is giving us the same kind of pattern we saw before in terms of a variable multiple domain response across these patients. Now in the next Slide 11, we're introducing it to an Angelman Severity Score. The first score, the change score with a relative score, the severity score is an objective score which looks at criteria of a series of items that will determine where you are on the score. So this is more of an absolute score and more rigorous. When you look at the same set of data in Cohort 4 and Cohort 5 using the objective severity score data, we see a very similar pattern with very strong sleep results and multi-domain improvements across the range in Cohort 4 as well as In Cohort 5. And so we're seeing -- the objective criteria showing us a very similar pattern what we saw with the relative change -- Angelman change score. Now if you look further over time, this is day 128, and on the next Slide 12, you follow the patients that have made it to the PMD visit. There are 4 patients, 3 from Cohort 4 and 1 from Cohort 5. What's very striking here is that there's 1 patient you can see from Cohort 4 that's now achieving multiple domains of 2-points or 3-point improvement. In fact, every domain is improved at least 2 or 3 points on the objective scale. And this is 7-point scale, so these are meaningful changes. And these improvements are greater than what seems this trial at day 128 would suggest that once their patient is loaded, that they can continue to improve over time. And we think this is very encouraging to see that kind of broad effect across multiple domains in -- this patient was actually the 4-year-old. He's the youngest patient in the study, which I think is very enlightening with regard to the drug and what it can do. Now on the next, we're going to go through and deep dive into some of the particular domains and give you some commentary both from the objective scale as well as from caregivers. A deep dive into sleep domain. I think what was important, we saw a number of patients with significant improvements in the sleep score. These are severity score changes. 3 of the patients with 2, 3 or 4-point improvement. And I think you can see from the scale down at the bottom, the kind of things; these are not all the items, these are just 2 of the 7 items; that are being scored in terms of creating a score. But to go from a score of 6 to 2 as in patient 001 did, that is a change from severely impaired to slightly impaired. And if you look at the caregivers commentary, you can see the support for what a profound improvement in sleep is observing for some of these patients that were, for example, not able to sleep and waking up multiple times and now sleeping through the night or having much fewer awakenings. We think this result is very similar to what we saw before. In this study, we have actually more patients with sleep problems, but it's very similar to what we saw, particularly in patient 5, in the prior study with a very substantial sleep improvement. On Slide 14, we'll dive into the behavior of the Vineland score. Most patients have improved in the score, and this is the personal self-care subdomain, not the whole of the Vineland. This is how they take care of themselves, feeding, eating, toileting . The 2 patients achieved statistically significant results in the Vineland score, most of the patients did improve. And you can see some of the verbatim on the side, which tell you a little about how patients are changing. They're using a fork, opening doors, et cetera. We think it's still early for the complex type of behaviors involved in the Vineland but we're encouraged to see the improvement going forward so far. On Slide 15 and some comments from patients around general behavior that is adaptability, interactions with others. And I think what we're seeing here is very similar to what we've seen before in terms of patients being calmer, being more responsive, more interactive. And we're pleased to see these kind of changes happening, which are important for families in their daily lives. On Slide 16, we'll dive into the communication domain, the Bayley-4 expressive/receptive communication improved in most patients. In 3 patients, there was a statistically significant improvement in receptive and expressive in 2 patients. One patient under the receptive in Cohort 1 had an inexplicable drop, even though that patient had a very strong ORCA score, and we believe that was just a bad day at test. Subsequently, at the PMD visit, day 170, that patients showed statistically significant positive response in the receptive along with the other 2 patients. So all 3 that reached day 170 have statistically significant improvement in the Bayley-4. They also showed trend of improvement in the expressive communication. So we're encouraged by what we're seeing here with communication on the Bayley-4. Now if you go on the next slide, the ORCA, which is the caregiver evaluation is a complement to what you see with the psychologist-administered Bayley test. In this core, we see ORCA in that 1 patient achieved 14.2% statistical significance and another 5.7 and several patients with positive scores. And the verbatim here suggests what the patients are seeing from a standpoint of communication, being able to use their device and string of words together to make a sentence is certainly important. And some of them making -- trying to make sounds that is very similar to what we saw before in terms of patients beginning to make constant noises, which preceded their transition to actually saying words. So far, we are encouraged then on the communications domain. On Slide 18, the next one, we touch on fine motor and gross motor. We've seen gross motor improvements, both Cohort 4 and 5 and good improvements as well in the fine motor in the Cohort 5. In Cohort 4, some variation in response in Cohort 4, but we're encouraged with the type of things we're seeing both in fine motor and gross motor, particularly in walking, which I think is really important, both because patients with [indiscernible] especially if they get older can have a great difficulties with walking and you can see those improvements, but also because it tells you that these patients' lower extremities are in fact improving, which further helps confirm the fact that we're not seeing a safety issue with the lower extremities. On Slide 19 is the EEG data. This data was more mixed. It was not as distinctive and not as strong as was seen in previous data. We believe we probably need to have a higher loading dose, and we'll continue to watch EEG going forward. So on Slide 20, Cohort 4 and 5 are entering the therapeutic range with acceptable safety. When we look at what we're seeing, we're not seeing any drug-related SAEs, no leg weakness, with stable CSF proteins. This tells us that given the [ oligo ], the ASO at this dose and with this administration strategy, we can avoid any inflammation that might be a cause of safety event. We see encouraging clinical activity both in the change score and the severity score, but it's supported by a number of different secondary evaluations, which help show these changes can be clinically meaningful for some of the patients. So we think that's very important, especially with multiple domains positive in most patients. It's also important to know when you look at the three Cohort 4 patients that reach PMD that in one of those patients, you see substantial progression and benefit beyond what we've seen at day 128, indicating that they've been loaded with sufficient amount of drug to achieve the knockdown, we believe, which is enabling them to show continued progress over time, suggesting just more time might help these patients do better. When you look at the details of what we're seeing with the patient verbatims and the results of these tests is clearly the pattern response on the same. The types of changes we're seeing are very similar to what we saw before. And so we feel now with a whole different set of investigators in different countries in different patients, we're getting very much the same pattern. With doses now at 7.5 to 10 already cleared, not showing this effect, starting the next phase at those doses, however, I think we're pretty confident that we should be able to achieve a higher load effect and we would expect to see even further enhanced efficacy. That's the story with the data for now, and I'd like to introduce Scott Stromatt, who has joined Ultragenyx. Welcome, Scott. And Scott will be interviewing our guest, investigator guests today.

Thank you, Emil. It's a pleasure to be here. I'd like to introduce our audience to Dr. Erick Sell. He's one of our leading investigators in the U.K. and Canada. He's enrolled the most patients. He's an Associate Professor as Pediatric Neurology. He's at the Children's Hospital of Eastern Ontario in Ottawa. Erick, thank you for being on the call with us today. From your experience and your perspective, would you tell us your view on the safety and efficacy that you're seeing so far in your patients?

Thank you, Scott. Yes, we're seeing encouraging results early in terms of clinical activity across the different domains. And without hesitance, I find that these are related to the medication. On a very positive note, we haven't seen any drug-related safety issues. I personally feel that the therapeutic effects could be enhanced as we have started to go a bit higher on the monthly loading doses and as we continue to increase in the maintenance part of the protocol. So yes, I can see that -- in general, I do see evidence for a therapeutic effect of the study medication, and I haven't seen any side effects related to medication.

Thank you, Erick. Another question. Can you comment on the dosing administration strategy? We administer the drug by LP. Then there's a flush and then they're placed into Trendelenburg position. Can you tell us about how that is tolerated by the patient and the families?

Yes. So in our center, it has been very smooth. We have a sedation done by our colleague anesthesiologist. It's a light sedation usually with a combination of Isoflurane gas plus/minus propofol, depending on the criteria of the anesthesiologist. With that in place, it's very easy to do the lumbar puncture and inject the study medication. At the moment that we're placing the patients in Trendelenburg positioning and give the 10 ml of flush, we then take the patient to a recovery area with the other patients that are recovering from anesthesia, and we haven't had really any issues that were unexpected. Often the patient is going to wake up just before the time of Trendelenburg and the nursing staff at the bed side will calm them. We call the parents. And in that way, we complete the 45 minutes of Trendelenburg positioning. And then the patients quickly are recovering and are following the clinic for the rest of the protocol. And other than one instance where a patient had nausea and vomiting, which was related more to the anesthetic process because the patient had had an anesthetics in the past and had that common issue of nausea, everything has been very smooth.

Thank you, Erick. That helps give us some color and perspective. I appreciate that. I'll turn it back over to Emil to move on to the U.S. slide.

Thanks, Scott and Erick. So I want to just touch briefly on Slide 22 on the U.S. clinical data. I won't go through all of it in detail, but it's the program that we agreed on with the FDA. And on Slide 22, are some panel information using both the change score as well as severity scale and the Bayley in receptive/expressive communication, fine motor and gross motor as well as ORCA. So we think the data are showing, again, support even at 2 mg dosing. And for these 2 patients loaded and these are younger patients only, I'd remind everyone. And the safety, I think, was very similar to what we've seen before. So we're very comfortable with that, and we're encouraged by it [ today ] so far and I'll invite Scott then to talk with Dr. Berry-Kravis, who's been with this program for quite a while.

Thank you, Emil. Good afternoon, audience. Dr. Berry-Kravis is the Professor of Department of Pediatrics and has enrolled the most patients, has had the most experience with GTX-102. So Liz, given the totality of the data that you've seen and we presented today, how does this compare with what you've seen in the original 5 patients?

So I would say, Scott, the response to GTX-102 does vary some between patients, and it's definitely something that builds with time. So in short times, we don't get the whole story. But I would say that the early responses that have been seen in the recent patients treated with the lower doses are quite similar to the kinds of things we saw in the first 5 patients on the higher doses early on. There are small changes when you look at the whole spectrum of development, but they're quite exciting when you work with kids with developmental disorders who just don't learn new things quickly. And when you look at the speed of acquisition of new skills and the rate of development in Angelman syndrome, so the data from these new patents, I would say, mimic the first 5 patients and that there are changes across domains. And there's a confluence of data on multiple measures, showing that all the measures are moving toward improvement in the patients. There seems to be more rapid change in those who are younger, but there's still variability even within a certain age group as we would expect with any drug and any disorder.

Thank you, Liz. When reading the verbatim from the patients and the parents -- excuse me, just the parents and what the patients are doing in terms of communication, sleep, gross motor and other areas, what is your impression compared -- you touched on this earlier. What is your impression compared to natural history? Could you go into that a little bit more?

So I think these patients are actually describing very similar things to what we heard from the first 5 patients, things like using a fork better or sleeping through the night. And they are describing early changes that are clinically meaningful to families. And these changes are, in fact, being identified not just by the families, but there's early movement on a performance-based measure of Bayley and the clinician interviews of Vineland and some have even had significant improvement on the growth scores in particular areas. And some have had improvements in areas that aren't even detected on any of the standardized measures, things like swimming. So the kinds of -- the significant improvement on growth scores on these tests take years to show change in natural history of Angelman syndrome. So these are -- this appears to definitely be a different rate of progress than what we've seen in natural history. And so my experience is that the changes have been surprising. To me as a person who followed the Angelman patients for a long time, the amount of change that we saw in a few months, both with the early patients and with some of the later patients, is certainly more than I've seen in 20 years of trials in neuro developmental disabilities. Because it's proven very difficult to treat the underlying core symptoms in these disorders. So I would say, the early evidence suggests that when this drug is used that even at the current doses, it's changing the rate or pace of development and because rates and paces produce improvements over long, long time, it definitely suggests we need to study the drive over an extended time frame to see how much change is possible, but it's very encouraging.

Thank you, Liz.

Thank you, Scott. Thank you, Liz and Dr. Sell. I think they'll be here for questions with -- in both periods. I think Dr. Berry-Kravis is very right -- the fact is our variability. The patients are all very different. But the fact that we see multiple domains in each patient with its own pattern, I think, is very encouraging. With what we've seen already and what we've known regarding the protocol in Cohort 4 and 5, we did file and have approved moving forward, both in Canada and U.K., on treating a new cohort at a higher dose level and our expectation is to treat 2 patients in each cohort. This is a 7.5 mg starting dose for the young patients, 10 mg starting dose for the older kids. They can titrate further one more step similar to what we were doing in Cohort 4 and 5, but starting at 7.5 and 10. Getting closer to that 10-milligram dose level that we felt was the level at which we've got consistent knockdown in nonhuman primates. The screening began and first patient in Cohort 7 has already been dosed. And so we're off -- now we are hoping to redose some of the original 5 U.S. patients in Canada, and that should start soon. We want to get those patients treated. Again, they're very interested in getting treated and getting started again. Now opening the protocol in the U.S. will depend on submitting and review of the interim CSRs that the FDA has requested from us. And so that process will be getting going because it is our plan to get their acceptance and will begin to open U.S. sites. In the meantime, we continue running in the ex-U.S. study. Now our next program update sometime late '22, early '23, I think we've given you a very [ substantive ] update today. Now Slide 25. We made the decision to acquire GeneTx and complete early that acquisition and take control of the program, and it was because we were confident in the value of the program for Angelman syndrome. We are convinced to what we're seeing in both the safety and efficacy evaluation at this early stage. We're comfortable on the dose range we achieved so far and that we're very close to the dose range that we think will be optimal. Therefore, we're confident in our ability to get there given the safety we've seen so far. We think Dr. Dindot and the fund -- work funded by FAST and patented on the intellectual property of the gene region and the understanding of the molecular biology, I think is very critical and very unique and valuable to us. In addition, we've learned a lot about Angelman syndrome biology. The fact that Angelman syndrome can reverse, that these symptoms can change, and we're seeing that again replicated with different investigators in different parts of the world. I think that's an amazing opportunity when the biology works. We can't fix the biology, but as the biology can revise and improve neurologic function, that's an important element of our decision. The combinations in our data IP and the biology put us in position to drive forward a development program. By taking on the program ourselves, now we're able to impart more effort into the control of the execution of the plan and head towards Phase III planning as we continue to collect more data. But we are excited about what we're seeing so far. Now deal terms shown, we had negotiated with GeneTx for an earlier option of $75 million upfront, which now has a 30$ million milestone for Phase III [ FPI ] and all other details. And ultimately, there are regulatory and commercial milestones and tiered royalties. It has exclusive right to global data and in support of GTX-102, it does include, in fact, additional alternate ASOs that are in hand. So final on Slide 26. We are "All in" for Angelman syndrome. I can't imagine a more powerful place to be where the biology has given us and the molecular science, the insight in how to change the disease that I thought that the medical genesis couldn't be changed. And we're actually going to be able to change the outcome for this disease, and I can't imagine anything more powerful and important for us to do, as a company, to develop the first-ever treatment for Angelman syndrome. We need to continue to optimize the therapeutic benefit, managing the dose administration to get to where we need, but we think we're close, and we're encouraged by what we're seeing so far. Given the scientific basis and the biology, as we've just discussed, we have to see that early option. We're joining now more formally our hands with GeneTx team that basically is a great story of families fighting for their children, investing in science, finding an opportunity and turning it into a therapeutic. And I want to give my tip of the hat to Paul Evans, [indiscernible] and Scott Stromatt for -- and of course, Scott Dindot, who did the original science, to put this together, in the matter of a few years and put forth an opportunity to change the future for Angelman syndrome, and our intend to develop the first approved therapy for Angelman syndrome. So that's it for the presentation, and we are ready for a call. Operator, give the instruction for the call.

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

And just a follow-up on the dosing that you've done so far versus what you want to do. It looks like you've dosed up to 10 mg individually, which I think is leading up to this update you had, had hypothesized could end up being the best dose. So if you were seeing that clinical activity that you mentioned, why do you want a dose higher to 14 mg? And then secondly, on the acquisition, was there anything in the terms which required you to acquire the company at a certain point in time? I guess, would you have to have paid more if you chose to make the acquisition later? And I might have a follow-up.

Very good. Thanks, Tazeen. So on the dosing, keep in mind, we've reached the 10-milligram dose, but that's in during maintenance phase. And what we're really doing is starting at 7.5 and 10 during the monthly loading phase, which should achieve a higher load. We don't necessarily say, we're going to 14. We -- it could go higher, but we think simply by taking the dose levels we've achieved and giving them now monthly will allow us to get to where we think we need to go. And it could be that we're at the dose or could be for one or both parts to go even a little bit higher, but we're right there. And I think the key thing difference is that we achieved 1 dose, but now we need to load over a period of 4 doses to achieve the level of drug we think will give us our efficacy. Now on the deal acquisition, there was an original deal and then we made an amendment, and I'll ask Tom, if you would like to just briefly touch on the rules of the acquisition.

Sure. Thanks. So in April 2022, we entered into an amendment of the purchase agreement, which did provide an earlier option based on interim data to acquire the company. And as you noted, there are differences. In fact, on the early option, there is a lower upfront at $75 million upfront versus $125 million originally set out in the purchase agreement. That would have occurred towards the end of this year. So by exercising early, we will have made a $75 million payment or thereabouts to GeneTx, the shareholders of GeneTx. And there is an additional $30 million milestone that will be due upon the initiation of the Phase III, so first patient in that Phase III study. So a total of $105 million versus the $125 million that originally was described as the upfront in the purchase agreement. So net savings is $20 million, presuming those...

And the original deal that did include time limit.

Correct, yes.

That would have happened later this year. So the question is, do we want to do it now? Or do you want to wait till later year? And our view was better to go now, we had enough data to give us confidence and accept a little bit more risk, but with a discount of the upfront price. So that was the deal.

Our next question comes from the line of Joon Lee with Truist.

Can you elaborate a bit on the numerically superior CGI-I-AS improvement in the U.S. cohort versus the Canadian and the U.K. cohorts that got much higher doses. Is that a function of how CGI is assessed in the U.S. versus Canadian and U.K. cohorts or something else? And what did you see in the comparator arm in the U.S. cohort? And related to that, assuming your next trial is going to be a registration trial after seeing Cohort 6 and 7. How does the data today help you with the powering assumptions and the size of the subsequent trial?

Yes. So I think all the investigators evaluating the CGI will be doing it on the base of their insight and knowledge. And I think when I look at the overall data and you look at all supportive objective data, I think what you'll see is that the ex-U.S. patients at the higher dose have stronger data in terms of the numbers of the other objective criteria compared to the U.S. patients. So I think the data are consistent in that way, and I wouldn't overread the CGI change for interpretation at this point. With regard to powering, it's a little premature to talk about powering. Well, I'll say to you is that there's there is -- I don't think there should be a problem in designing a study that has this many domain improvements to work with. Normally, we're dealing with one thing and other things. Now there's variability, but we are the kings of variability management. That's what we do at Ultragenyx. Everything we do is about variability. In fact, they can talk about variability. So that's something we'll be good at, and I have no doubt that we'll come up with a great power in combination with Scott, Camille,[indiscernible], we'll figure that one out, but we have a lot to work with right now. We need to get the dosing, understand the required efficacy and then work out with the authorities, what's the way to capture the multiple domains of impact [ on a patient ] in a complex disease like Angelman. And that's a discussion I'm going to be very excited to have.

Great. And then can you comment on the -- what you saw in the comparator arm in the U.S. cohort?

Yes. We don't have any comparator arm data at this point in time. So it's not there yet, but honestly, it was something the FDA has to do what we are going to do, but we don't have any data yet to report that. That program is a little bit further behind the others.

[Operator Instructions] Our next question comes from the line of Gena Wang with Barclays.

I have 2 questions. Thank you for sharing all the data. Since you showed all different endpoints, I have -- the first question is for both the doctors. What, in your view, could be approvable endpoint? And what magnitude of change that could be considered as a clinical meaningful? And one question for the company, for Emil. What magnitude of improvement would you be looking for at a higher dose cohorts to expand to the maintenance dose?

Gena, I actually would not put that question to the docs at this point because it's really a speculation of what be an approval endpoint. I don't think it's worth having speculation right now, but there's a lot of things to work with. So you're asking -- the second question you asked was on the higher -- what we expect from higher doses. Well, we would expect to see more patients have improvements in the multiple domains in the communication area as well as in fine motor. We think we're seeing improvements. We think it could be better based on what we've seen before. And so we'd like to see some of those domains to improve further, and we also are going to gain more experience with time that as the patients that are loaded now will have more time, which will also gain just watching them and seeing, are they on a trajectory of gains, that is the patient as they are currently loaded on maintenance, do they keep gaining ground as one patient did. If they do, that gives us more indicator about the time and process for improvement. So we'll learn about dose and how we are at some objective things, and we'll learn a little more about time as well, and how much time we might need to get to an optimal efficacy point. Is 4 months enough? Is 6 months enough? Maybe it needs to be 8 months. But somewhere in there, we'll gain the ground that will help us understand the best choices.

Our next question comes from the line of Cory Kasimov with JPMorgan.

I guess, based on this update, how are you thinking about the potential timing of a Phase III registration study? And I guess related to that, how much do you first need to see from these additional cohorts with the higher loading doses?

Yes. Well, obviously, it depends a little on whether we get to the mark of efficacy we want to see before we decide. We're doing the next cohort. I think we're very close. Is that the Cohort? Do we have to do one more? This year will be spent testing those patients getting -- and if we hit our mark in this dose or the next step, we would expect to be collecting data this year into early next. Our expectation, though, is to be able to talk to FDA once we have enough patient data to kind of say what the efficacy looks like and how we might design the study. We hope to be able to get to that later in the year or early next, and our expectation would be to head to Phase III in 2023. But we haven't really yet specified exact timing for that, but that would be our target.

Okay. And then as a follow-up. In terms of your discussions with the FDA and thin upcoming submission of the interim CSR, is it your expectation that this would lead to better alignment between the FDA and international regulators in terms of protocol?

Yes. Our goal is to gain alignment that the U.S. and ex-U.S. parts are doing the same thing. And we just need to give the FDA convincing support for one that we can dose lower this drug and using this administration strategy and do it in a safe and effective way. And that -- the issues that they have raised before or questions they've had are answered sufficiently that they no longer are concerned about other aspects of the mechanism. We think we have that information. They are just like -- would like a formal report in this division, and that's what we'll provide them. But we are -- we would expect them to get there and would be driving and seeking their agreement on synchronizing the global protocol to move forward so that when we head into expansion of a cohort once we hit it, we'll be using that both U.S. and ex-U.S. sites.

Our next question comes from the line of Joe Schwartz with SVB Securities.

I was wondering if you've done any analysis to understand whether the loading dose or cumulative dose are more important drivers of efficacy. It seems like the efficacy seems to increase over time, and I'm wondering how much of this is due to an increase in dose throughout the trial versus some other factors like accumulating effect for lack of a better term? I guess...

Yes. So with that experiment in humans that we've done so far, right? It's hard to gain that insight, but we have gained the insight in the nonhuman primate studies that were conducted. And in those non-human primate studies, we showed that you could get to the knockdown with a single dose at a high enough level or you could get there with separate dose. So one dose of 3 mg can get you knocked down, but also 3 doses of 1 can get you to the knockdown spread over a few weeks. So it's pretty clear that the drug has a long half-life and that you can accumulate using small doses that are given frequently. So I think in this situation, there's a combination of 2 things. You have to have enough drug, but also in the end have to have enough accumulation in the brain, and that's what we think we can get to. Now that we're achieved, we're hitting the 7.5 and 10 mg dose range, we think that loading at that dose range will get us closer to what we think would be the level that would result in consistent knockdown of the antisense [indiscernible]based on our non-primate data. So that's why we think we're pretty close. So far, I would say, the non-human primate data [indiscernible] has been pretty accurate at what we're seeing. So we feel pretty comfortable that it is predictive.

Our next question comes from the line of Yaron Werber with Cowen.

Emil, I have a couple of related questions. I mean, I guess the first one, from hearing you, it sounds like you're still trying to figure out whether it's going to be a CGI-AS primary or a multi-domain sort of responder score. Is that correct? And do you need to validate an MDR? Or can you do it based on the emerging data from this study? And then for secondary endpoints, I imagine you're not contemplating doing like a full Bayley or full Vineland for rather some of the sub domains, and will FDA or MA require those to be validated ahead of time?

Well, I can guarantee you that they will require whatever we're trying to prove for efficacy and get label that'll want validation, and that's why we have Dr. [indiscernible] and an entire team of people that spend time designing and validating our endpoints. So the CGI Angelman syndrome, either the change score is one way of going. The problem of the change score is it generally is a relative scale and therefore, not intrinsically objective or meeting necessary proven clinical meaningfulness [indiscernible] itself. The Angelman severity scores have criteria in terms of the scoring, and so it actually provides a little more confidence of what is meaningful or not. But we are willing to go whichever way that regulators want us to go, but our feeling is, if you want our clinical insight choice is that we should be using a multi-domain type endpoint because a multi-domain type endpoint will capture the totality of the data, and we capture a variable population of patients and get the most out of them. Within the MDRI, remember that we don't have to validate the MDRI exactly. Within the MDRI are actual endpoints like the Bayley receptive and expressive indication, that would be one domain that we would include. So the MDRI basically includes a series of validated domains, but we have to pick a minimally important difference that will score off of and we have to provide support for that score. So the MDRI is really I would think of it as an analytical tool rather than exactly an endpoint of itself, and we believe with the validation of some of the instruments we're talking about that with support for the appropriate MID that we can get that. Now this is a novel thing. The FDA has generally not been so accepting of new things, but we are also being the company that is push things ahead that are new, including we used the MDRI in our MEP SEVII approval, it was in there as our key clinical secondary endpoint, it was positive. We also used a novel trial design in that program, and we've done a lot on developing other types of endpoints, including Rickets scoring system for XLH which didn't exist before. So I can tell you, Yaron, we're good at that. We will figure that out, and we'll validate whatever we have to use, but our hope would be to use something that covers multiple domains to get the most out of the efficacy and to achieve the most in terms of labeling. But if it doesn't work that way, we end up with the change score as a primary, I'm comfortable with that also that we can hit that and that we'll use the secondary endpoints, individual domain endpoints to help fill out the labeling of the product.

Our next question comes from the line of Dae Gon Ha with Stifel.

Congrats on the progress. I guess one question and one follow-up. Can we go back to Slide 16 and 17, Emil. You briefly discussed it, but just kind of curious on how the day 128 bad day affected Bayley-4 to the different extent as ORCA. Just trying to get a sense for those 2 metrics given it's administered by 2 different, I guess, technicians or professionals. And then secondly, we've also heard through talking a KOL that there seems to be work being done, trying to establish an MCID using the Bayley scale and could be published sometime next year. I guess how -- thinking about the trial plans for Cohort 6 and 7, I guess, how would that kind of factor into your overall endpoint decision, if any?

Okay. Thank you. So on Slide 16, 17, just to be clear, the Bayley is conducted while the patient is in the room with the psychologist and it requires them to perform, they have to actually perform. Whereas, the ORCA score, which is on the next page, is the parent communicating, filling out essentially responding of what they're seeing at home. So that could have been during the last period of time rather than today alone, if that makes sense. So remember, the ORCA is testing -- the family is testing patients in their normal personal environment at home, whereas the psychologists there in the clinic that day, did something happen on the way in. They didn't get the breakfast they want, something else happen. All kinds of things can happen to make that happen, and it's just part of the nature of these type of instruments. They are noisier because little things can cause problems, but when we look at the big breadth of the data, the patients are getting better. This patient had low, but obviously turned statistically significant positive just 6 weeks later and had an ORCA score that was well beyond the threshold of statistical significance. So we think that shows that the patient really is improving, and we're encouraged with what we see there. Now with regard to Cohort 6 and 7, look, the Bayley and a number of these instruments have a lot of normative data. They have a lot of historical use, and there's a lot of data to help support an MCID or the new terminology referred as MID. That's a minimally important difference. So we think it's not that hard for those types of scores. There's a lot of use and a lot of data to help us found what is really important difference. And these statistical significant marks you see on the presentation are already based on threshold that would be considered clinically meaningful. So being statistic significant in this threshold would mean clinically meaningful. So we feel like achieving that is very possible here, but we need to achieve it for the majority, if not all, the patients, not just some of the patients, right? And that's where increasing the dose, we hope to gain knockdown of a similar level across all patients so we see those kind of changes across the board.

Our next question comes from the line of Yigal Nochomovitz with Citi.

This is Carly on for Yigal. Can you comment at all on what you observed at the Day 58 assessment? I guess just curious if some of the changes you saw on CGI had been observed by day 58 or if they didn't service until after the patients have been escalated to the second dose level. And then sorry if I missed this earlier, but what's the expected timing if you have it for adapting the protocol in the U.S. to more closely match the ex-U.S. dosing?

Sure. So day 58, we saw something similar. We were seeing improvements, and we actually talked about that on our Q1 call, which was the first Q. So we saw something similar, but as we noted then, the patients were still titrated. So they didn't achieve the 2 plus 2 main at Day 58. In general, they are titrated. So we begin seeing the improvements really within the first couple of doses, but we wanted to keep titrating until we saw trajectory improvement that felt more consistent with achieving optimal knockdown. For the U.S. protocol, we have to do the work to prepare the CSR and submissions to the agency that will take a little bit of time, and our hope would be to get that started here in the second half so that we can contribute, but it would be hard to predict the FDA and their response in this area. But I do think we have answered all the questions, and I think we have substantial data on the protocol going forward. So we hope to be able to get that to them and get that going later this year for the U.S. patients.

Our next question comes from the line of Maurice Raycroft with Jefferies.

It seems like for patient 003, who's 4 years old, showed the most benefit on CGI. Can you talk more about specific observations you're seeing based on age, kinetics of efficacy in the younger patients, and then what you expect to see going forward in the younger versus older patients?

Yes. So it's a pattern we've seen before that it seems like the younger patients may respond more quickly and potentially at lower doses. What I would suggest to you is, we've created 2 groups in our dose ranges hat vary in 4 to 8 and 8 to 17. What I would say is, those are very [ crude ] dose grouping. The fact is very small or youngest patient had a strong response, just like patient 5 last time had a very strong response to 20. It tells you that when the patients are small, the effect of the drug may be higher, concentration may be achieved, may be higher. It just indicates to us a little bit more about the dose response, and we've seen the young patient having a better load and effect. It just tells us we're very close to entering the dose range. And for a smaller patient, that dose is already sufficient to get them where they need to go. Whereas perhaps in a larger patient we need to get to a different dose. So this age and dose response, ultimately with this and hopefully, when we expand the 20 patients per group, will give us better insight in the right dosing algorithm to optimize for a small patient versus the old patient. Now aside from the dose to loading effect, there's no doubt that younger patients likely respond faster from a biological standpoint than the older patients. And that's been seen even in the animal model, but the truth is, what's encouraging is, we're seeing nice effects across the age range. Even the 13-year old in this study, the oldest one, actually showed some nice effects. But we definitely will have to incorporate both the time of effect and the relative dosing to come up with the best regimen or dosing algorithm for the drug. But right now, that information has helped us see that maybe we're entering the range and we're close.

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

This is Tommy on for Salveen. We wanted to first ask about the patients who are not improving on the domain. Do you know what could be a driver behind that? Maybe it's the low loading doses, duration of follow-up or just the patient heterogeneity. And on safety, could you just provide some more detail on the COVID cases and the upper respiratory tract infection, like how serious were these? And did they happen around the same time?

Thanks. So I'll comment on the nonimproving, and I'll let Scott comment on the COVID and the other cases. With regard to improving, first of all, even in nonhuman primates at a dose level when you're a little below the optimal, there was definitely a variation between animals. So when you're below the dose level, there's probably [indiscernible]and how much drug is distributing and getting where it needs to go and to what degree the biological factor occurred. So it's not that surprising that there is some variation, especially when you may be not at a dose that's saturating in a sense. So we don't see any, let's say, obvious reason why some patients didn't respond. They're in the middle of the range of the others. They're not a particular age or not or a type or character that we can see. So it's still a bit random, but I feel comfortable that if we move in the dose range of being more saturating, we should get a more consistent response across the patients. But I would say in a matter of where the disease is variable, there'll always be some variation in response. No one will be the same, and honestly, it's been like that with every rare disease treatment we've ever done. Now I'll let Scott talk about the COVID cases and [ URI ], and I'm very happy that [indiscernible] actually talking about safety because that [indiscernible]

Yes. Exactly. We had 4 cases of COVID happened at varying points of the trial. Not surprising. I think any of the investigators will tell you that these children are always putting things in their mouth, and so they will be exposed and the upper respiratory infection was a common cold and the same thing, four cases at their end-points in the trial. They were all grade 1. All [indiscernible] is not related, obviously. And the emesis, you can ask about that, but that also occurred sporadically. Didn't occur with every dose. Probably related to procedure. One of the children it was stress response. When they get stressed, they vomit, but it didn't happen with every dose.

Our next question comes from the line of Liisa Bayko with Evercore.

I apologize for the background noise. They're doing like massive construction in the office beside here, and thank you for sharing all this data. There's a lot to digest here. So I'm sure we'll have more questions as this time goes on, but just a couple for me right now. The Bayley-4, you have mentioned, Emil, seems to have more variability than say, ORCA and so I was wondering about the CGI scores. Maybe the doctors can comment like how many of these things are a moment in time? And how might that look from one day to the another? So like how much is above and beyond what might be like background noise or either sort of developmental milestones that may be achieved? Could you just comment on that?

Sure. Well, I think the thing you have to expect about the Bayley is, we're dealing with -- and I think Liz just talked about this variability issue, which is, these are severely impaired children. And now you put them in an unusual setting with a person they don't really know, psychologist is going to test them. And so their reaction in that situation may vary on that day and how that -- so I think even without any change, there will be that variation. And with change, there may be another part on top of it. The CGI has one benefit is that in this case is the doctor and the doctor they may have seen multiple times, they're may be more comfortable with who is doing the evaluation of the patient and having various inputs. And I don't know, Scott, if there's anything you think that would help distinguish the 2 between them.

You want Liz to address that?

If she wants to talk about the CGI, certainly, Liz.

Worked on the instrument and worked first 5 patients in these patients. Does she want to make a few comments?

The reliability of CGI's findings.

Okay. I can go ahead. I think we worked hard to anchor the CGI. So it would have specific kinds of items that would define how problematic the person's symptoms of Angelman are in different domains. And also, when we're doing the CGI, we're asking about the symptoms over the past week or over the past 2 weeks, depending on how it's defined in a particular study. So it's not dependent on how that patient performs in the moment in clinic. And of course, we do take that into effect also, but the CGI is a global composite of everything the family tells us the patient is doing, the videos they may have sent us, what the teachers that they're doing in school and what we see in the patient when we see them in clinic. So as a result, it's not so dependent on performance at that moment of time as the Bayley is.

So maybe -- does that mean that it's a little bit less variable, you would say, than like the background variability is like? I'm just trying to distinguish like single from noise kind of factor.

Yes. The goal of the CGI is always to not be too variable, and so that's why in many studies, we anchor the CGI so that this amount -- this rating is this and that rating is that. And then we do fidelity training with the clinicians that are in the study, particularly for registration studies. So it is, yes, supposed to be less variable when you work hard to get alignment between the clinicians on what the different ratings mean.

Okay. And then on -- can I just ask one more? Just on the same concept of CGI, the severity versus like the [ C ] rating. So how do those really relate to each other? Because it doesn't always track you have a change that may be 0 or minus 2, but then it sort of shows up slightly differently on the sort of minimally versus much improved or no change? How do those kind of relate to one another?

Well, the severity scores actually has objective items with specific score Angelman syndrome that we've set. So that severity score is a summation of those items that comes up with the point score system. So it is an objective scale, it's not a relative scale. That means it's assessing today what these things are you're seeing, but when you look at the other score, the CGIs and interpretation of how they were at baseline versus how they are today, right, and setting a criterion for what's minimal, what's moderate was very much. I think that becomes a little bit more subjective. There are criteria that you anchor on as a Liz has put forth, but you're still trying to compare 2 things. And so that involves some level of judgment when you're making those comparisons. With regard to the FDA, they have accepted the change type 4. That's what [indiscernible] had in their Phase III. However, historically, change scores have been less accepted because they have the issue of recall bias, and there is a tendency to prefer instruments that are assessed in an objective criteria timelines. But I think the fact that FDA has accepted is good. I mean it's a sensitive tool, but the objective criteria are actually harder and more rigorous because you have to hit multiple things in order to make it work. So I'd look at severity being objective and more rigorous and to some extent, a better way to determine clinical meaningfulness. But I think the relative scale is sensitive to showing that something's happening in the trial that's meaningful. The quantification of it is a little bit different. I don't know if you want to add anything to that, Tom?

No, I think you covered it. The change is looking back to baseline from day 128. And severity, you're looking at that point in time what the values are and things.

Our next question comes from the line of Joel Beatty with Baird.

First, for the comparator arm data requested by FDA, how far behind is that? And is that something that would be in the next update in the late '22 to early '23 timeframe? And then also for vomiting and emesis, how severe and how long lasting was that?

I'll ask Scott. Go ahead.

Yes. In terms of the vomiting and nausea, we had one patient actually swallowed in EKG lead, and that caused vomiting, probably he had had some pre-existing bile gastroenteritis. This is one of Dr. Sell's patient and the patient was hospitalized for a while to get the vomiting under control. This is a patient that responds to stress with vomiting. She's had multiple injections. Sometimes she vomits. Sometimes she doesn't. The other episodes of emesis was grade 1. So again, they don't occur every time. They're fairly sporadic, mild. What was the other?

Comparator arm timing.

Yes, the comparator patients have all completed the day 128. The FDA wanted us to look at these instruments from baseline to day 128. Dose at all been completed. Those patients are now being dosed at the 2-milligram here in the U.S. So now that they complete it, we'll collect that data and do that analysis and present that at a later time.

But we didn't have -- the day 128, we didn't make in time.

They weren't available at the time we did the data cut.

Right. So they're in progress, since they are already past event, we just have to it by the data cutoff for you.

Our next question comes from the line of Debjit with Guggenheim Partners.

This is [ Rai Forcets ] on for Debjit. You mentioned earlier that the viral infection could be attributed to the oral behavior of the patients? Is it possible that also a steroid regimen is playing a role in the viral infection? And our second question is, you've mentioned the dose projection kind of based off of safety and efficacy. Does CSF PK/PD plan to your dose projection at all? Have you determined an ASO half-life?

So first of all, the patients are not getting steroids in the protocol, so it's not really related to the viral infection and I think COVID was rampant in U.K. and Canada. So that's kind of what it was. In fact, I'm happy as only a minority, not the majority of the patients have got COVID because they're going back and forth to the hospital. So we don't think that's related to anything from the protocol or a difference. On the dose, we believe the CSF volume probably is a factor in the PK/PD of the drug. That is the amount of drug -- the concentration of the drug in the CSF is what the brain is exposed to. It's one of the reasons why we're thinking that the dose might need to be higher in older patients versus younger patients, and that's partly why we stratified the 2 dose cohorts for starting those. So yes, I think CSF concentration is a factor, and we're going to be paying attention to that we come up into defining a dosing algorithm for different ages in this program.

Our next question comes from the line of Laura Chico with Wedbush.

I just have 2 -- just a follow-up on the last question around PK/PD analysis. Emil, based on what you're seeing now, I do realize you're moving up to a 14-milligram dose. But wondering if you could discuss maybe perhaps how you're envisioning the dosing interval during the maintenance phase? I guess is there a potential for this to kick out and follow a longer time frequency between the 2 doses as you go further out? And then I have a quick follow-up.

Yes. So right now, we're just going to 7.5 and 10. If we don't get to where we know, they will step one more from 7.5 to 10 and 10 to 12. So we're not necessarily going to 14. That would be only if we're not where we need to be. So just to be clear on that point. It's very possible as you get to higher doses loading that you could change the time. We had made the decision early on that trying to change 2 variables at once actually makes it very hard, and it's better for us to figure out the dose during a 4-dose load. Because the [ half way of the drug ], a 4-dose load should get you the steady state. And every 3 months, we think should maintain that steady state, it's not accumulated a little bit. So the goal of this is to kind of make dose the actual variable and not play with the regimen just yet. Once we're get in patients' load and hitting our mark in efficacy in the maintenance phase, in the long run, when they're loaded in the maintenance phase, we can look at whether you can space out the maintenance phase a little further. But right now, our biggest goal is, let's get to the dose that is safe and doing what we need with the assumption that we use for 4 doses, which should get you to steady-state drug levels based on the PK properties of the drug. That's how we're going about it right now, Laura. Hopefully, that answers your question.

It does. That makes a lot of sense. And I guess I do have a follow-up question then just for the physicians. What percent of your patients currently under care for Angelman would be amenable to receiving monthly intrathecal administration? And I guess I'm asking this in the context of how feasible is transportation and just getting them in the clinic. What proportion of your patients could feasibly go through a regular intrathecal dosing regimen?

Yes. Thank you. Erick, do you want to make a quick comment in that regard for your patients?

Yes, sure. So it, of course, will depend on the medical system. So factors are country to country and province or state. I can speak in Canada, there wouldn't be obstacles in the number of patients that require or that are scheduled because there are provincial hospitals of different levels that would be able to collaborate and have all patients dosed under sedation. So yes, I wouldn't -- I don't think that there would be a limit in the percentage of patients that would be able to be dosed, considering the need of sedation because if a hospital is, let's say, saturated or does it...

I think -- excuse me, Erick. I think they were just asking, do you think parents would not want their children to get an LP? Yes. What fraction of patients would want to do this is, is what the idea is...

Sorry, apologies, 100%. So they -- so far all the parents that I have spoken with, they all pretty much would go for it.

Thank you. I don't know what Liz thinks.

Yes. I have about 200 patients on my wait list to get into any ASO trial right now, and I haven't met anyone in clinic who was afraid of LPs when faced with the magnitude of the problem in Angelman syndrome and an opportunity to improve it.

I'm showing no further questions in the queue. I would now like to turn the call back over to Josh for closing remarks.

Thank you. This concludes today's call. For additional questions, please contact us by phone or at [email protected]. Thank you for joining us.

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