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

All right. Good morning, everybody, and thanks for joining us for the 45th Annual TD Cowen Healthcare Conference. I'm Yaron Werber from the Biotech team, along with my colleague, Jaena Han. And it's a great pleasure to moderate the next fireside chat with Ultragenyx. Really need no introduction, with us today is Eric Crombez, Executive Vice President and Chief Medical Officer; and Josh Higa, Head of IR in the audience as well. Eric, great to see you. Glad you got to the elevator craziness to get here. So we appreciate it.

So there's lot to talk about. We're going to focus on OI then Angelman and we definitely want to talk about the pipeline as well. Maybe let's start by -- you talked about doing the second interim analysis for the Orbit study based on the April data cut, which is the 12-month data cut. And at the same time, you're going to do now an interim analysis, which I think is also based roughly about 12 months into the COSMIC study, but that's going to be your first interim, and you're only going to do that interim if the interim and Orbit works, right? At that point, you'll trigger the interim analysis on COSMIC. Do I have that correct? Is that how you're going to determine?

So a little bit different for this interim analysis. So for an analysis 1 that read out and we disclosed in January, Orbit really led. So we needed IA1 for Orbit to hit, to trigger the analysis and readout for COSMIC. The difference -- we're calling this IA2 technically, it will be the first interim analysis for COSMIC but we're doing both interim analyses for Orbital and COSMIC in parallel and we said we would release that data midyear.

So you will do the interim analysis from COSMIC regardless of whether -- I mean, you're going to do the second IA for Orbit but you only release that if it hits, obviously. But with COSMIC you announce regardless, you'll do that IA regardless?

Yes, I think we'll decide depending on how they hit. I mean the truth is, obviously, we've worked through scenarios. We do think there is a really good likelihood of success for both studies. So yes, our lead scenario is both will hit and we'll make that information available. But if one hit and one doesn't, we'll definitely work through that communication plan.

But if -- okay, so let's play it out if Orbit hits, I mean, you'll discuss that, and then it sounds like you'll just continue with COSMIC.

Yes. In COSMIC, we really did bring that along as a supportive study. So certainly, it's nice because we enrolled patients and COSMIC down to 2 years of age. So it does help with label expansion there. And then importantly, what I think COSMIC is helpful for is if people want to have a debate on the effect of bisphosphonates versus the effect of setrusumab you can now have a data-driven conversation there. And I think that's important. We don't think we necessarily need to hit on COSMIC if Orbit hits because you can still extrapolate that data back to Orbit and still get that broader label, that broader indication, but obviously, we'd have to negotiate that with the regulator. So Orbit stands alone, it would be a great label without COSMIC, but we do think there's a lot of success for them to go together into a single package.

So with COSMIC, it sounds like you did not do the first -- you didn't do a first interim, right? What was the reason for that? And to do it in Orbit and it sounds to me like -- and it's my own words, that you did it at around 6 months and kind of what makes sense -- the second is going to be 12 months and not doing a similar analysis for COSMIC. Is it because of the washout on BPs and how long they might have an effect for?

Yes. It's more because really, Orbit was designed to be the pivotal trial. And originally, that was the plan. We brought COSMIC long -- a little bit later in the clinical development plan to get that broader label so forth there. So I do think Orbit stands alone that is a great label at launch. I don't think COSMIC by itself probably is that strong of a label there. So for IA1, we only wanted to spend the alpha on COSMIC if Orbit was positive because, again, why spend the alpha unless you need to or want to.

Got it. And with COSMIC, I guess, in the very unlikely situation, which Orbit does not meet what can COSMIC get you? And I imagine it's got its own statistical plan separately?

Yes. I mean I still think it's important to remember that originally, we did plan for Orbit to go to 24 months with 200 patients. On the strength of that Phase II data, we brought that into 18 months, which would be full study readout in MVIR and brought that patient number down. We ultimately enrolled 159 patients. So I think in this scenario where Orbit did not hit in COSMIC didn't, we would just have to decide, does the strength of that data support the product and that. But again, based on the strength of the Phase II data, based on the fact that this is about fractures that people understand fractures, it's very clinically relevant. It is absolutely supported by the increase in BMD, and that is the increase in Bone Mineral Density. We think it's a very nice logical story. We were just under that, in my mind, unlikely scenario where Orbit did not hit in COSMIC, didn't when we would make that data available.

And so maybe just remind us as you just did, Orbit is 159 patients. What age groups is that? And it's obviously low, high versus placebo. COSMIC is 50, right? Head-to-head against standard of care. So maybe give us a sense of the age groups? And then how is COSMIC powered as well?

Yes. So, again, kind of as a principle, we like to really stay true to is we don't like to bring untested things into a pivotal trial that we did not evaluate in the Phase II. So entry criteria for Phase III is very similar for Orbit Phase II, 5 to 26 years of age, types 1, 3s and 4s. We did enroll a total of 59 patients, 2:1 randomization, setrusumab to placebo. COSMIC again that did come down to 2 years of age versus 5 for Orbit, 1:1 randomization, setrusumab versus bisphosphonates, and that was a total of 69 patients.

And for COSMIC, is it sufficiently powered on the primary endpoint statistically?

It is. And again, what helps there is the 1:1 randomization versus 2:1 in Orbit really helps with that lower patient number.

And what's the -- I remember with Orbit, you're looking for 50% delta sort of an absolute. What about at COSMIC?

Yes. So just to finish that out. So that was -- so Orbit Phase III is -- has greater than 80% power for a 50% treatment difference between the 2 arms, if we go to full 18 months. COSMIC is greater than 60% power to [indiscernible] the treatment difference between the 2 arms, setrusumab versus bisphosphonates.

And what time point?

So that was at full readout at 24 months.

24 months. With COSMIC in general, you would expect to see more fractures, right, than Orbit as patients are younger age. Am I thinking about this correctly?

I don't know. I mean it's 2 years of age versus 5 years of age. I don't know if there's a big enough difference there where we would really see a clear signal there. I do expect, again, because in Orbit, the great majority of patients did come in on bisphosphonates. So that placebo group came in with a background of bisphosphonates, obviously, those in COSMIC randomized to bisphosphonates and stay on the bisphosphonates treatment. So honestly, on the whole, I probably expect them to behave fairly similarly.

Okay. So the difference is, COSMIC, you stay on BPs and Orbit you wash out. And just remind us how long is the washout for?

Yes. So that's it. If you really wanted to do a true washout of the bisphosphonates, I think you're probably looking at something close to a year, it is definitely months. So we're not looking at a true washout of bisphosphonates in Orbit.

But they stopped them, is it 30 days or at least 30 days before...

Yes, before screening period, yes.

So we were in some simulations and -- statistical and it's obviously all about sort of what you know and how you set it up. But we took sort of everything we knew. It sounds like the fracture rate was 0.7 before. It sounds like the rates were higher in Orbit, at least relative to what it was in the Phase II. And so once we started kind of looking at the first interim analysis, I mean, to really stop it at that point, you had to have been sort of in the 60%, 70% plus absolute treatment effect. In the second one, we think you have ample power, based on 40% to 60% delta. And then you can probably hit on the third, if you're in the 30% to 40% delta. Are we in the ballpark?

Yes. Yes. And I think and there are some nuances there with -- and we've talked about looking at an annualized fracture rate for our modeling and how we designed our studies of 0.72 versus 1 and the difference there is with the primary endpoint, the FDA wanting us to exclude fingers, toes, face and skull and morphometric vertebral fractures. So if you exclude those, our patients coming into a trial had an annualized fracture rate closer to 0.72. If you go all in, which we do think these fractures are important, you are at an annualized fracture rate of 1. So when we were really looking -- we were looking really at that range of 0.72 to 1 for the AFR for patients not only coming into study, but what your control groups would look like across Orbit and COSMIC.

Okay. And by the way for the audience. If you have any questions, just raise your hand, we're happy to take them. on your behalf. The -- I guess the one thing that we're getting a lot of questions on, is it sounds like you'll do the second IA based on April data. let's say, if you -- you announced the last one in January, but if we're correct, and it was based on a 6-month look, that was based on an October data cut which took about 2 to 3 months sort of the clean. Is Orbit going to take just as long in the second I to clean the data? Or did you already preclean some of it, it's going to be faster?

Yes. I mean I think in a ballpark, it is going to be roughly the same timing. And the difference here is if we're successful for IA2 that means the DMC has done their job. They have looked at the data. They have said they have achieved the interim -- the goal is for interim analysis 2. Then that top line data comes to us and then we need to do our edit checks, our review of the data to make sure it's good. So there's a little bit of time there as well. But it's fair to say the time line should be very similar. Because to your point, we've obviously cleaned everything going into IA1. We continue to clean as we go. So yes, that takes time but it's also that data analysis, the interpretation of the data by the DMC and then also by us if it's successful.

And at that point, the timing which you announced that hopefully, it hit, the patients at that point, they still need to come back for a last visit -- last safety visit before you can announce or they could do that afterwards?

No. So again, and that's a little bit of a difference between IA2 and IA1. We are -- and maybe it speaks to really the probability of success here. So we are looking to have a fully cleaned locked database for those interim analyses.

Okay. And you can do that you just need all the patients to come in maybe an extra visit right before you do that second look.

Yes. So all patients will be at least 12 months. So you'll have that data visit and obviously, the earlier enrolled patients, we'll have more data there. But again, that's an important time point to me, 12 months on treatment is a good duration.

Okay. We can come back to -- yes, go ahead.

Just really quickly, some KOLs are saying that type 3 specifically is more severe than either 4 or 1. Just remind us what you saw from the Phase II if you add type 3s? And is it helpful to have more type 3s in these studies? And did you try to get more type 3 in these studies?

Yes, so...

Maybe repeat the question.

Yes. So the question was around really the spectrum of severity of 1s, 3s and 4s. 3s and 4s, absolutely being more severe than 1s, our types 3s and 4s are the ones who you're seeing in wheelchairs, really needing walk devices and truly more severe. We did have type 3s and 4s in our Phase II data set, certainly with 24 patients, it's a relatively small number of patients. But what was interesting to see was, at least in the patients who were on the study, they dually did respond as far as a reduction in atraumatic fractures, reduction in overall fracture is very similar to the type 1, which was very reassuring. Again, because we don't like to bring untested things to a pivotal trial, our entry criteria for Orbit Phase III is really does near what we did in Phase II, but it was the strength of the Phase II data -- because, again, type 3s and 4s, they can fracture turnover during sleep, just the tension of rolling over, they can fracture. They can fracture transferring from the wheelchair to the car. So getting into clinic really does have a burden and risk of fracture. They really needed to see the strength of that Phase II data to say, I'm willing to do this, I'm willing to take the risk to come into clinic. So that's when really we saw that second data release at 6 months of data, where the PIs and physicians really saw the strength of that and brought in more 3s and 4s. So we've said, if you combine 3s and 4s as one group, that's about 50% of Orbit Phase III versus 50% of type 1s.

[indiscernible].

Pretty close, yes.

I'm sorry it's 50%.

3s and 4s combine versus 50% of type 1s.

That's in the Phase III. And what was in the Phase II? I'm sorry if I missed it.

17 were types 1s and 7 were type 3s and 4s.

I'm sorry.

1-7, was type 1, 17, and 7 of the 24 were types 3s or 4s..

24. Yes?

So a greater portion of more severe patients.

Yes. Okay. We're going shift -- we're going to shift over to Angelman. And the -- we're going to do a panel actually. I think it's tomorrow afternoon. We're going to discuss the Bayley- 4 as a scale. The Bayley- 4 was really designed for adolescents, right? And the -- we're talking about a study that's going to be adolescents, obviously, and I think your competition obviously has had some data in adults already. Can you talk about maybe for communication (sic) [ cognition ], how does the Bayley-4 actually work?

Yes. So Bayley- 4 was designed for neuro...

I'm sorry, for cognition, I said communications.

Yes, we're a cognition Ionis is bringing forward communication. But just to talk about the Bayley- 4 for a second. So developed for neurotypical children is what's used developmentally for all children to make sure they're on the appropriate developmental curve. Because of that, sites are comfortable, FDA is very comfortable, they like this. So it's not -- it's restricted by age for neurotypical children for patients with Angelman because they have the skill set of much younger people, it is actually appropriate for adults as well because if you look at where they are at developmentally, they're obviously much, much more younger. And when you look at whether it's a natural history data from the LADDER study or from what we have seen, the development curve for them is very, very flat compared to those patients actively treated where those patients are now on their own developmental curve. We brought forward cognition by Bayley- 4 as our primary endpoint versus communication because cognition is fundamental to everything else you're going to do from motor, behavior, speech and sleep. So we like that foundation there and measuring it there. Ionis may have preferred communication because remember, we are bringing in patients with full deletion. Again, that's what we studied in Phase II. Those are your more severe patients, but also the great majority of patients, depending on who you talk to, between 70% and 80% of the total patient population. Ionis is bringing patients with all genetic changes in, those more milder patients actually can develop some language with our treatment there. So you have the propensity for maybe more effect there. And we all know that language is important to families and stuff, but that was the basis of our choice. And I guess what I think may be the basis of their choice.

And so how does that again, we'll discuss this in the panel a little bit more detail tomorrow. But literally, how does these -- the nice thing about Bayley- 4 it's literally fairly subjective, right? It's given by the same reviewer at the office. It's not dependent on retrospective sort of patient feedback. But how do you assess communication versus cognition.

Yes. So I mean, you have your -- I mean it's -- it's all very spelled out in the menu there and it goes up very prescribed. And a lot of these particularly in younger patients when you're looking at cognition, it is what you can do in the types of testing there. And then communication, obviously, you're breaking down receptive versus expressive. You generate receptive first following commands understanding before you develop your first words and sense. But they all have their separate set of tool questions and obviously, versus cognition and they're just -- you're looking at very different domains here for a different set of questions and tests there. And it is important because there is a caregiver part of this test where you can take caregiver feedback about what they can do at home, we've excluded that too, to eliminate any type of bias or effect there.

And so one of the challenges, these are phasic kids with developmental delay with a motor component. So even as part of communications, some of it is there's a box test, right, moving cubes around and all kinds. These are inevitably motor. So it does require some interface between motor skills as well. How does that handled?

Yes. So I mean this is the challenge and I guess, again, why I'd like to talk about [indiscernible] because it's so concrete. Neurology is hard -- this is a neurotypical test. It's not meant for Angelman children. So you can hit a ceiling effect there, and that's why you need to develop your manuals, very strict testing criteria. To your point, you need consistency there to make sure you're following this because you do want the ability if you're using Angelman patients potentially allow them to miss a certain number of tests, things that can't because of the Angelman syndrome and not because they're leveling out. Let them have a number of negative results before you say they've reached the end of their test. So you don't hint this artificial ceiling there. But again, that's why you need to have the conversation with the FDA. You need to have them buy into this and you have to have very consistent set manuals for all of these tests.

And so when you powered the Phase III, can you remind us how that was done? And what's the delta you're looking for?

Yes. So again, that's all based on the results from the Phase II and with really the strength of the results there, and we studied a total of 74 patients, 53 in the dose expansion. So a lot of experience there. And again, when you look at the data -- but importantly when you also look at some of those videos we put out, we're not looking for subtle changes. We really are looking for true development gains. We want to see these children continuing to gain new skills over time. So when we look through what we, I think, all agree are important domains, cognition, speech, gross motor behavior, sleep, honestly, you could have made an argument and brought forward any of those as a primary endpoint and still had a very well-powered study. Again, we think cognition is important. It's foundational but we wrap the rest of those tests into the MVRI, we're applying alpha towards that. So while the MVRI is a key secondary from a statistical perspective, it's in the primary family of endpoint. So it allows us to guess look at cognition, allows us to look collectively all of those other assessments because those are important. At 120 patients, you have greater than 90% power to achieve our goal here, with 120 patients, 1:1 randomization is a very highly powered study.

And you're looking specifically cognition based on raw scores from baseline, right? Can you discuss why that's important?

That's just the FDA choice. With the way the Bayley normally does it, there's just a control for age there. We released our most recent data release, both the raw scores versus GSV and they're very, very similar.

Any questions from the audience? It sounds like the study is ongoing, and it sounds like you're expecting enrollment fairly rapidly, right? And how many sites do you expect to have open and how many are open right now?

Yes so -- are we seeing exact number of sites?

I think we've said it global study.

Yes. So we are looking at site of -- I'm not sure sometimes not allowed to disclose it, not specifically. So we are looking at a relatively large number of sites across Europe and the U.S. we are bringing forward our second study Aurora, which will go down to 1 years of age up to 64 years of age and include those other genotypes. We will bring that study to regions we're not in particularly regions like South America. So really having a global footprint there. But we really wanted enough sites pulling forward Phase I/II sites and sites with experience for something like Zolgensma, who can do this do this well, but also work through 120 patients very quickly. We released that we started enrollment at the end of the year. It's important for us to get the study full enrolled as quickly as possible. certainly this year, but as quickly as possible.

And would you have sites in Australia as well and how about Japan?

So Japan, it's very important. We do have an office now in Japan, and we're really looking at Japan as part of all of our development plans really across the board. That includes Angelman. We look at Australia, we've run studies in Australia before, we did have a Phase II site in Australia. So to me, it's more -- it was really site selection based on the capacity of the site, how many patients can you bring through. For example, in some places like the U.K., it's very hard to get the nursing support in the fusion space you need there. So really where we select our sites in countries is making sure we can do this quickly and pull through this many patients in the shortest amount of time.

Right, Jaena, over to you.

Right. Some rapid-fire questions on the other assets in your pipeline, starting with UX111 for Sanfilippo, obviously, the PDUFA coming up in August 25 of this year. What do you think is going to be the uptake following approval? And can you remind us a little bit of the market sizing, especially given that there isn't really newborns screening yet?

Yes. So there's no newborn screening. And at least in the U.S., it's always a tough conversation of, can you do newborn screening for diseases? You don't have treatment? Is that ethical? And that's always been the debate there. A lot of us, particularly in rare disease would argue newborn screening is important. It can help bring forward new treatments, but that's a paradigm we're in. So we're thinking there's probably between 3,000 and 5,000 patients globally, at least in the territories we cover, 20%, 25% of that's probably in the U.S. We do think this approval will bring newborn screening available. Again, when you look at Sanfilippo there's no disease with higher unmet medical need. There's really nothing you can do for these patients besides basic support. And unlike Angelman, this is not something that you can reverse. This damage is permanent. So you really need to start treatment and stop this disease as quickly as possible. So I think given that unmet -- given that this is permanent damage, I think there is going to be a very big demand from these parents to treat their children.

Got it. Makes a lot of sense. And then for DTX401 for GSDIa, you plan to file this in mid-25. You think supply from your own manufacturing plant. How is all the tech transfer and validation taking place for this?

Yes. And again, we talked about the Phase III study we designed, initially to be 48 weeks. We released that data and the delta in time between then and filing was the amount of time it took to bring this manufacturing process in-house at our manufacturing facility in Bedford. We had a prefiling meeting, we are in alignment with the FDA and everything we need to submit as part of that package. For approval and yes we're looking forward to filing. And then that the opportunity to bring forward even more clinical data, including clinical data and crossover patients. So clinically, that will make the package even that much stronger.

How much durability data do you need to kind of satisfy that FDA requirements?

Yes. So the agreement was on the 48-week time point. We put off that data very, very compelling, statistically and clinically significant. So if we had not taken the time to bring manufacturing in-house, I think that would have been a highly approvable package. So just having that much more safety efficacy and durability will only help us.

Great. And then finally on for DTX301 for OTC deficiency, where do you kind of see the bar for success on both the primary endpoint and kind of the final percent [ CR ] 64 weeks. And do you think you can have interim data in the second half of this year, given when you finished enrollment?

Yes. So we just announced relatively recently that we fully enrolled that study at 37 patients. We are focusing on ammonia that is the basis for alternate pathway approval. So that's a clinically full approval end point there. And with ammonia, what success looks like is, because we really don't tolerate ammonia levels above the upper limit of normal, it can very quickly lead to coma and really a lot of permanent damage. So we do want to bring ammonia levels under control fully. And then for the patients who are fully able to come off of alternate pathway medications and diet we want a significant number of patients to completely come off of alternate pathway medication.

Got it. And then can you give us a brief sense of how many of these patients are in the U.S. who are above 12?

Yes. So it is -- so we designed, again, based on the Phase II, 12 years and above with late onset disease. We think there's roughly 10,000 patients in the territories we cover, 80% are late onset, again, 2025 in the U.S. But we will study neonatal onset, younger and older patients in our pediatric studies that as soon as we're well through really moving through our end of Phase III -- are moving through our Phase III study and getting going. We do want to bring the pediatric studies along as soon as we can.

Great. Are we out of time? Okay. And then finally, do you further -- do you need to transfer manufacturing in-house? And kind of what kind of scale do you anticipate?

Yes. So because, again, it does take a lot of time, and we need to build capacity as we go through. We're going to continue with the CMO for OTC at the beginning. That's our plans for launch. And then when we can, we will bring that in-house like we did for GSDIa.

Great.

Okay. Eric, thanks so much, and Josh. Really appreciate you coming.

Thank you.

Thank you so much.