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

Good morning, everyone. Welcome to Barclays Global Healthcare Conference. My name is Gena Wang, I'm SMID Cap biotech analyst at Barclays. It is my great pleasure to introduce our next presenting companies Ultragenyx. With us today, we have Emil Kakkis, President and Chief Executive Officer. Emil, before I jump into the questions, do you want to give just a brief overview of the company.

Sure, happy too. Good to see you again, Gena. I think we've been talking for a long time here. So good to see you again. So Ultragenyx now founded in 2010 with just 2 people. We're now a company with 1,300 employees, global commercial. We are doing rare and ultra-rare diseases, multiple modes and approaches to their therapy. We have 4 approved products and 5 indications. They're exceeding $500 million of revenue this year. So we're a global commercial company. In addition to that, we have 6 late-stage pipeline programs. So a very full -- one of the biggest rare disease pipelines out there includes 4 gene therapies with MPS IIIA and GSD1a 2 diseases closer to either unblinding the Phase III or getting to file. In addition to that, we have Wilson and OTC gene therapies. Our big biggest programs, though, are the Angelman program, which we'll certainly be talking about with Gena, which is coming up to a data release of expansion data on our Phase II study for that disorder and our osteogenesis imperfecta program, which is an antibody called setrusumab. And that's -- we released Phase II data last year, and we're enrolling Phase III right now. Heading to -- we'll have some Phase II data later in the year, and we'll begin the potential interim assessments late in the year and next year, hopefully adding to a filing in 2025. So those are the big progress for us. I think going forward or managing our cash position. We're $777 million in cash, and we're managing net cash burn to less than $400 million this year, which will put us on a path to profitability around 2026 with the continued growth in our revenue and discipline on the spend side. So that's us in a nutshell.

Great. Thank you, Emil. So maybe I will start with Angelman. I think we learned that the data will be presented at the AAN as a late-break abstract. So maybe a little bit question around there. So for the expansion cohort, when we look at the clinical trial design, you do have over patient -- over 30 patients enrolled across cohort A and E data from 20 patients from 6 months follow-up. So maybe when we -- at the AAN presentation, like what kind of dose range you already would be testing? And then how -- like how much data you will show like what kind of domain you will show. I know you will not be able to talk about the actual data, but just the amount of the data you will be seeing.

Sure, I have to do that. So the expansion cohorts actually compass 54 patients enrolled in the expansion cohort. And we'll have more than 20 patients, which will have 6 months or more of data, right? There will be some patients with less data, and we'll provide a safety [ appetite ] in all of them. The expansion patients that we're probably presenting are primarily from cohorts A and B, which are the ex U.S. because those enrolled sooner. You remember the U.S. opening took a little bit later, and so the other cohorts are a little bit further behind. We will present the data though, as a group, young and all together, all as one group because I think it's -- we're not really seeing as big an age differential or other basis for it. So right now, we're going to stick to basically presenting the data as one group, which is what we did at the Analyst Day for the patient. So what we are going to be looking at are the same 5 domains with a very similar format to what we presented before at Analyst Day that are -- those domains are Bayley cognition, Bayley receptive communication, behavior, sleep and the Bayley growth motor. Those are the key 5 domains. We'll analyze some other endpoints, but the main focus will be those domains -- in those 5 domains. So you would look for those. We're going to look at them over time from baseline through day 170, and we'll compare them to day 170 in the extension data that we showed last year as well as to natural history where natural history exists, which is primarily the Bayley. So it should follow the same format and allow investors to track what we showed you before and what we're showing going ahead. Day 170 is earlier and our expectation for Phase III is to be running more likely 8 to 11 months kind of timeframe, which we showed you last year was day 250 to day 333. In that timeframe, we think we had -- we accumulated more efficacy while not keeping kids on placebo for too long because these are little kids, and we're doing placebo intrathecal injections. So I think the 8- to 11-month range is what we'd expect FDA. We've had that discussion with the agency. Our expectation is at least in that we had a conversation with the agency earlier in the year is that we will be doing a randomized placebo-controlled trial for Angelman, 1 trial is placebo-controlled, 100 to 120 patients, probably in the 8- to 11-month timeframe. And we're talking through the endpoints, which they were happy with these domains that we're talking about and the endpoints we're using, and they were very flexible in how we approach the use of those particular instruments, which was a refreshing and new. They were actually willing for us to look at even trimming some of the questions within those domains, not actually sticking to the hard choices. So I thought it was -- we've had a collaborative discussion with them and I felt good about our ability to get there. We have multiple domains to work with. The Multi-Domain Responder Index is something we've talked a lot about a way of capturing the data. We could use that as a primary or we use it as a key secondary in either place, it allows us to capture efficacy across multiple domains in individual patients, which I think is added to the value. And we have multiple endpoints to work with that we've shown. So the late-breaker is only a 5-minute talk and poster -- it's like a poster symposium. We would expect to have an associated investor call at the same time, which would allow us to go in a little more detail and talk about regulatory and other issues that are not appropriate for the late-breaker presentation -- that kept you all the questions?

Yes. So, Emil, you said that you will group all the patients together. You do not see the differences in terms of age. What about the loading dose because you do have a different loading dose?

Well, originally, we keep in mind what we started with was the loading dose for the [ 48 ] was 3 and 5, remember, and the older ones were 5 and 7, they are one step further. But as we titrated everyone, we end up kind of finding that the differential doses not -- did not need to be there. We were wondering whether little kids needed were more sensitive or had more safety questions at a higher dose, but it turned out there's not that much difference in the dose effect. And so all the patients at the young and older actually at the same dose, the same loading dose. And so we -- what has also been happening as we -- originally, everyone thought that older patients wouldn't respond, right? The young ones respond and the old wouldn't, but that's not really true. The old ones respond too. And so our view is that they all are responding, and I'm not seeing what we call as distinctive in age difference. And maybe in a larger group, we'll see it. But right now, the main way we're going to present is all the patients together. If there's any value looking at subsets, we can certainly do that. But the big picture is how do these patients look. And we've been impressed with how older patients do respond sometimes in different ways, but they are responding too. So -- and if you look at the extension data we showed you, you couldn't tell which one was old, which wasn't young on that list, who responded, who didn't? I couldn't really tell. So -- I think that's a good thing because it means out in the marketplace, all those kids are addressable, right? treatable. So we think younger patients could respond even better but we don't want to put young kids into a randomized controlled Phase III trial when we haven't studied them before. So our plan is to -- the randomized trial will help cover 4 to 17, the majority of the market, and we'll look at young patients in an open label format and maybe look at other genotypes, et cetera, in a basket trial, it's kind of a supportive trial for expanding the label. But I don't think as a parent, I can't see putting people's kids 2-year-old in an intrathecal placebo-controlled trial. I think one placebo-controlled trial to prove the drug works. And once proven, we can then interpret data we see in open-label format for young or others as being rational and valid. So that's our strategy going forward. Our expectation from this data is we collect it, maybe some more extension data end up going to FDA for end of Phase II mid-year to finalize. So at AAN, we're going to tell you about where we're at in the discussion, but finalizing the trial design and endpoints will occur in end of Phase II meeting. Because the FDA invited us to an early meeting, which was special, come before you had Phase II data to talk about endpoints. I think they're trying to help us be successful. And that discussion helps lay down a foundation for how we can get to end of Phase II and finish the discussion on endpoints and trial design midyear. And our hope would be then to start a Phase III later in the year.

Okay. So I have several questions here. So when you -- like the younger patient -- I mean, 4 to 17, right? So that range, you do not see any major differences. What is the maintenance dose for these patients? Do you have a range or you were on a fixed dose?

We have put the dose out before that an extension trial. They're in the 10 to 14 range, most of patients heading to 12 to 14 and ultimately pushing to 14. That's 14 milligrams every 3 months. We haven't put out what the loading doses are. We've been holding that closer. At some point, we'll put out what those loading doses are. And so -- but everyone is at the same dose or loading right now.

I see. And then there is no concern on safety for the younger patients, then it will be higher. If consider the [indiscernible].

Well, I think it's -- at the dose we have chosen, we are not seeing an issue.

Okay. Good. And those dose you will disclose at AAN?

At some point.

Okay, at some point. Okay. And so then I just want to be clear. AAN, I think as April 17 that we write the -- so you haven't shared the data with FDA after AAN, you will go to FDA and share and then get the final Phase III design sign off...

Right.

And it get started. Okay. Good. So then going forward for the Phase III, so are you thinking about just one dose for the loading dose and also for the maybe higher dose for the maintenance dose?

Right. So for the 4- to 17-year-old group, we're thinking one loading dose and then 10 to 14 maintenance dose, and there would probably be -- we'll step up to it. What I would say to you is -- our expectation is that different patients may need different doses. But so far, what we're learning is they're close enough to the same dose that we could -- in the Phase III trial, it's simpler to treat everyone the same. But in the post-blinded period, we will allow some level of titration to occur to individualized dosing. I think in all neurology, there's individualized dosing is probably required. It's a strategy we've used in other rare disease programs where you do kind of a standard dose to get through the randomized period where it's harder to adjust. And in the post-blinded period, then you can allow individualized dosing decisions that will help expand the range of what's on [ label 1 ]. And this is how Kuvan was developed by the way. There's one dose in Phase III and then 3 different doses used in the extension period, and the label includes all of that in the labeling. It's just very hard to individually titrate in neurologic disease in the middle of a blinded study, right? You can't make a decision. So -- but we feel pretty comfortable that this dose range is showing a consistent effect. And therefore, we feel comfortable that we can get a good effect across all patients. And -- but there may be some patients that need somewhat more drug, let's say, and/or a different regimen. So believe open the idea that you can individually titrate dosing regimen.

Okay. And then the Phase III or in the future, that will still be the full-loading dose and then every quarter, quarterly dosing afterwards.

The expectation it's going to be 3 or 4 loading doses and then every 2 to 3, depending on the regimen. We're looking at different regimens actually in different cohorts slightly different variations. But essentially, within the same timeframe, what we showed you, it's the same number of doses. So through day 250, it's a total of 6 doses being administered.

Okay. So how soon do you think that you will start to see clinical benefit or separation from the placebo arm?

Well, based on what we've seen so far, we think within 6 months, you'll see separation, but which day 170 is that point. But if you looked at the data, there's still a lot of separation you can gain by going a little bit longer -- and so that's why the trial more likely to be 8 to 11 months. The drug that you give in the first loading period is clearly having effect for 2, 3 months longer. So you can -- if you go another couple of months, there is significantly more change. And so I think by the 8 to 11 month range, you'll have a bigger separation from what you see in natural history. Now the question on how much the placebo will be many have asked this question, like what because I think people are thinking back to the [ AVID ] trial, whether it was a placebo effect on CGI score. But keep in mind that in that trial and in some other trials and then Angelman people are using or mixing deletion patients with missense with UPD, and their developmental trajectories are different. We're focused only on deletion purposely to eliminate that variation. The deletion patients natural history is very flat. They do not [ going round ]. This will make for a cleaner assessment, right? Therefore, we think placebo will be -- and because of the timeframe 8 to 11 months, we think it's well past the timeframe for what a placebo effect would happen, would occur. And therefore, we're not really concerned about it. We think that the amount of change from baseline, let's say, in daily cognition, many of the patients were in double-digit 10, 20 point changes when 5 or 6 points is considered clinically meaningful. So the magnitude of the effect is really substantially past what you see. And certainly, I've never seen the Bayley score improved by that much in any program. We've used in a lot of programs where drugs work, but we don't see the Bayley change that much. So the imaginative change, I think, is so far beyond what you see with natural history that we're not so concerned about placebo effect with this particular design.

Maybe one last quick question regarding Angelman program, your latest thoughts regarding competitive landscape.

Well, what we've said from the beginning when we got into this program, there was already Roche and Biogen Ionis working. And clearly, they're advanced companies with a lot of ASO experience and neurology franchises. So we wouldn't go into that competitive environment lightly. In our diligence, when we looked at Dr. Dindo's work, which is the basis for [indiscernible] is doing, his work with a tour to force of molecular genetics. It was clear to us that he understood something no one else understood and that the use of ASOs to the very 5-prime end and the antisense meshes were far superior than messages in the patented region that Roche and Ionis are working in. The 5-prime end gets you a transcription termination effect, which is really important and actually reducing the antisense transcript, and that effect is best when you're at the very 5-prime end. So we had a patented region that was unique and that was more potent. And I think that's played out true. The Roche program was at much higher doses, though not disclosed. Their drug concentration was like 50 per ml. So we knew they were dosing in much higher concentrations probably more than 100 at some point but they didn't see efficacy adequate for their use. We're dosing in the 5 to 14 range, right? And we're seeing substantial clinical benefit that 10- to 20-fold difference is the potency difference that was apparent in nonhuman primates. The Ionis program is using an MOE chemistry, which is a safer chemistry than LNA, but it's also less potent. And in our hands, we make an MOE and LNA in the same sequence. The MOEs are 1/10 to 1/100 as potent. The LNAs are stiffer that gives them maybe more inflammatory effect but they're also way more potent. And so the real question in drug development, always listed therapeutic window, how much drug you have to give and what's the safety profile. We think the potency puts us in a position that we can be underneath the toxicity issues and achieve an excellent therapeutic benefit. So we appreciate that MOE thing. But the Ionis program is, we believe, working within this patented space they have, which is where Roche was. So our expectation is that they're using much higher doses than we are. I think the Roche Ionis programs were having effect. That's what we hear. But I think the question is, can they match what we're doing. And I think right now, we are -- I feel confident we have the best and most potent ASO.

Okay. So now switch gear on setrusumab, which also is very important asset now we potentially have the data first interim by the end of this year. So maybe Emil, like -- what is the rationale selecting say the event rates for the first interim? And the offer you allocated, I think very small half of [indiscernible]. And then how confident do you think that, that will be able to hit?

Well, in our regulatory discussions, the FDA wanted clinical fractures, excluding fingers and skull. So there's a certain definition of what clinical fractures are for the primary endpoint. Based on that, we have a fracture frequency, which we showed you the analyzed fracture was -- we showed you last time, it was around 0.7 or something, but we showed a 70% -- 67% reduction in that fracture rate. When we look at that data and we look at the fracture frequency, we are seeing at baseline in the early period of the trial. And in addition to what the reduction in fracture as it became evident to us that we didn't need 195 patients that 150 should be plenty, and open the door that we might hit persuasive statistical significance early. But in our regulatory discussions, the FDA has clearly wanted at least a year of treatment. However, our view is if you hit 0.01 p value before that point that you should have the right from an ethical standpoint of ending the study early. And so our proposal is to do an interim assessment when we had what was equivalent to about 60% of the fractures required and which comes expected to be sort of around the end of the year. But in order to compensate for the fact that we'll be doing less that some patients will have less than a year of exposure. We made the stringent -- a very stringent threshold of 0.01, which I think is possible. I think it's less than 50% it would hit, but I think it was possible of hitting. If it doesn't hit there, there's another interim, which we'll spend some alpha and then a final assessment of 2 more ways. But we wanted to open the door if it's really off scale that we could argue to FDA that we're going to end the study early because it's not ethical to continue and the result is so extreme. If you look at the data, it looks like half the fractures we show you were within the first 3 months. And after 5, 6 months, there are only really a couple of fractures in that group of patients. So it's very possible if they have 8 months or so of treatment, right? There'll be enough time for separation of this -- of the 2 groups, and they hit that kind of an endpoint. What will happen, though, is the interim will be assessed by the -- which is the clinical fracture endpoint will be assessed by DMC. We won't know unless it hits, if it hits then will be informed that it hit, if it didn't hit, we won't be told. If that one doesn't hit, there will be another interim a few months later. And then a final in approximately 18 months from the start. But that's kind of the timing expected. So we think between the end of the year and sometime in 2025, we should have the Phase III results. Our goal is to find -- finish as soon as we can with the power we think we have, we think we can possibly do that early and get to a file early. The Cosmic study is running right along parallel. Both studies are nearly finished enrollment. We're moving along very well. The Cosmic study, which is about 60 patients in younger kids -- is looking at total fractures -- we'll be evaluating that alongside the other but that study is randomized, controlled against bisphosphonates, but it's open label. So the DMC can actively look at fracture frequencies between the groups. If the group separate to a strong degree like hitting a 0.01 p value, they potentially get in the study too. There's no formal interim in it, but if it is of an extreme level [ statistical ] significance occurs, then it wouldn't be faired a little kids who are 2 or 3 years old to keep them going on an inferior treatment, right? So that program should move along in parallel, we'd hope that we'd be finishing both around the similar timeframe and be able to file with both studies.

Okay. So maybe the enrollment status for Orbit. I think I previously said 1Q '24, we don't have too many weeks left?

We're pretty close, on track, yes. We're doing well.

Okay. And then for the say filing with data from Cosmic would be including the filing package also with the Orbit?

I think it's important to get Cosmic data in there. We think the younger patients will respond even more rapidly than the older ones even more profoundly. So we think that data will develop even more quickly and will allow us to show differential very quickly because [ bisphosphonates ] doesn't work very well. And so we're pretty confident we'll have enough data, whatever we do. And to me, in the long run, treating 2- and 3-year old, this is going to tell you more about the long-term future. And I think with type 3 and type 4, if you can stop their vertebra from collapsing and their bones from being deformed. You take kids who would normally end up being wheelchair bound and disable their whole life and turn them into kids they're still walking and running around. We've certainly seen, as we showed at Analyst Day, a kid with Type 4 who has been on the drug for 17 months, who went from wheelchair and walker dependent to being running around and playing with kids on schoolyard falling and not fracturing. So the potential to be transformative is there, and I think it's going to be a very exciting program. And since we're very well developed as a company in the bone space. I think it's going to fit our commercial capabilities extremely well.

Okay. I think last question very quickly on Wilson program. I think you also have quite some progress there. Maybe the status of enrollment -- and then the latest thought on potential accelerated approval based on the biomarker give Peter Marks now very open or flexible.

Well, for Wilson gene therapy, we've fully enrolled the last [ co-patient ] cohort 3 was enrolled in Q1. And we'll release data when that patient has gone enough time to be able to look at chelator titration. So it's more like 6 months after that point. So think of it as midyear. What we showed is cohort 1, 2 out of the first 3 patients have come off their chelators fully and doing well. And we'll look for how many patients able to come off their chelators and still maintain low urinary copper. It's still early yet, but we're encouraged that the drug is active, and we're seeing not only the copper coming down in some of those patients, but serum copper oxidase activity, the ceruloplasmin oxidase going up, which is an indicator that they're distributing copper through the normal pathway. So that's where we are with Wilson. We're excited about the potential. I think that fixing the underlying defect will have a profound effect, I think, on Wilson in the long run. But we'll expect to get data midyear for that one. Now with regard to the regulatory pathway, I don't think the FDA has called an accelerated approval for us because there are multiple programs approved on copper, I believe they're considering that already validated what they asked us to do was to look at removal of chelators and still maintenance of the effect, right? Because they want to make sure that the gene therapy by itself is getting rid of copper. But I think in their mind, copper has already accepted endpoint. So they didn't really mention accelerated approval for that. But I agree, and we'll hardly support Peter Marks view that accelerated approval for these genetic disorders that affect biochemistry, it is a safe place for the accelerated approval. It's a place where this will work, and it's the right answer. Not only is it correct, it's actually better that using the biomarkers, the primary disease and the situation for these inborn errors will be far more accurate than any clinical endpoint. And so for MPS IIIA program, the heparan sulfate, we've had a big Reagan-Udall Foundation conference where all the companies presented. I think that is my hope will be the watershed event to change FDA's mind about use of those type of biomarkers. I think it's for OTC, it's ammonia that's an accepted endpoint from the past. And I think that the FDA would do well to start approving some of these inborn errors based on these biomarkers because I think it will be not only be right, it's actually the best way to develop and to allow for improvements going forward. So Wilson, I don't think will be considered accelerated approval based on our discussion. But I do think it's a situation where prior treatments have give us -- afford us maybe a clearer path than we would have for diseases where we're doing the first ever treatment.

Okay. Great. Well, thank you very much, Emil. Thank you, everyone.