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

Okay. Well, good afternoon, everybody, and welcome once again to the 46th Annual TD Cowen Healthcare Conference. I'm Yaron Werber from the biotech team, and it's a great pleasure to introduce and have with us today, Eric Crombez, who's Chief Medical Officer and EVP at Ultragenyx. Eric, good to see you. Thanks for coming.

Thank you.

So lots going on in -- maybe we'll start with Angelman syndrome. That's going to be the next, I think, one of the big catalysts in the second half, maybe even Q3, the way we're kind of calculating and trying to back into a more fine-tuned timing. The Aspire study is about 130 patients, 4- to 17-year-old with a deletion. That's about 70% of patients fall into that, Randomized 1:1 versus sham. The primary endpoint is cognition based on the Bayley IV. You obviously also have a Tandem study, the Aurora study, which we'll get into that in a second. When you're kind of thinking about powering for a benefit, what's considered clinically meaningful for cognition?

Yes. So I think, obviously, interconnected, but a little bit different. So I think the best way to think about clinically significant and for Angelman with our conversation with the FDA, we've shifted to MSD, Meaningful Score Difference. So when we're setting that threshold, and we specifically needed to do that as part of our MDRI, which is a second primary endpoint for us and set that MSD, your clinical significance at plus 5. That is a little bit different than how we're powering our Phase III study. And I think the best way to look at that is to go back to that slide of Phase I/II data that we showed at FAST. And the purpose of that slide where we're looking at Bayley cognition, both by GSV and Raw score. Our intention there wasn't just to give an update on the Phase I/II patients, how they're doing, but really show how we powered our Phase III study. We looked at GSV scores. That's fine. The FDA prefers to look at raw scores that is just what they always prefer. They don't like any kind of modifications to what you're doing there. So in that context, we saw a mean difference of 10.9 in our actively treated patients. And then the natural history data we are looking at, and I think really confirmed by any treater or anyone who understands the disease, really looking at natural history, extrapolating that into our control group and seeing a max change of one over time. With our modeling for statistical powering, we looked at 3x natural history, so a mean change of 3. So 10.9 change in actively treated up to 3 in your control group, that's how we get to our statement of greater than 90% power for success for our Phase III ASPIRE study.

So that would be almost a difference of 7 points. And what's clinically meaningful? Is it about a 3-point change or a 5-point change?

5.

5-point change. And that's from a baseline of that varies in cognition between 420 and 460, 480 depending on the patient. And that's a 1 year. How do you prevent variability or placebo effect?

Yes. So to me, the best way to prevent variability is to enroll the most homogeneous patient population you can. That is why we made the decision similar to Phase I/II to enroll only patients with full deletions in the Aspire Phase III study. So those patients full deletion, not making any UB3A. So that gives you your most homogeneous patient population. It also gives you your most severe patient population there. That's the best way to have clear signal detection and that does contrast to what Ionis is doing with putting all patients into a single Phase III with the potential there for patients with uniparental disomy, imprinting defects, missense mutations to express some UB3A and get to some language a few words just by my natural history. And then for placebo effect, the most important thing we did was to remove caregiver input. Normally, for Bayley, if you have a child here and say they refuse to walk for whatever reason, the parent says they walk consistently at home every day, it's fine. You can give them credit for that if you're doing the Bayley clinically. For our clinical trial, we do not allow caregiver input. You need to perform those tasks in front of the rater.

Each visit essentially?

Yes.

When you're looking -- so then you have the Aurora Tandem study, right? And that's a Phase II/III supporting the kind of the broader study with 4 different cohorts across all genetics and all treatment groups. What do you need to show in that study?

Yes. So we're taking a basket study approach. And again, we stayed very tight with our enrollment criteria for Phase III. So patients with other genotypes, younger and older patients in Aurora. So we took a basket study approach there. It does become a bigger trial because of that, certainly more complex, but I think the best way to round out the label for our Angelman product. And really, we're taking an extrapolation approach. So we will establish safety and efficacy in the Aspire study. We will importantly establish safety in the Aurora study, get as much efficacy data as we can and then extrapolate back to Aspire.

That study is 60 patients, right, Aurora?

Correct.

So it's about half the size of Aspire. So Aspire really is -- and together, can they drive a broad label or you file first for Aspire and then supplement?

No, yes. So 60 patients and open label. So it really does -- the basis of our approval package will be based on Aspire. And again, though, we do want to take an extrapolation approach there because obviously, all of these patients being affected by Angelman syndrome, need access to these drugs, and we do want to launch with as full label as we can.

Yes. So Oak Hill is kind of giving a regenesis to rugonersen from Roche. And they just recently published their data. They had a nice EEG delta effect, and they had some effects on the Bayley and the Vineland very similar, I believe, to what you saw and what Ionis saw. Ben Philpot did an experiment looking at data in preclinical models. And in their hands, they feel that the Roche compounds got the best mRNA expression, protein expression and reduction of the anti ATS strand. Any thoughts on that as a corollary to what we could expect in the clinic?

I think in a broad sense, yes. And it certainly is the same type of experiment, same type of data we did for our own proof of concept to enable us moving into the clinic there. And we have talked a lot about the potency of our ASO, and we started the molecular basis of that and where you're binding and continuing through expression of mRNA levels. So I think all of this data is supportive. To me, it's supportive of an ASO approach as being the right way to target something like Angelman. And certainly, I think it's great when people are looking at other modalities like gene therapy. But to me, you want to get it to as many neurons as possible. UBE3A is involved in how synapses are communicating with each other. I've been doing gene therapy for a long time. We have a lot of gene therapy. I think ASOs is the better way to do this and a better way to really hit as many neurons as possible.

Maybe just to go back to Aspire. The primary endpoint is cognition. That's got 90% of the alpha, the MDRI, the Multiple Disease Responder Index has 10% of the alpha. Is that the primary -- as you mentioned, I think you mentioned is the primary secondary endpoint. So that's hierarchically the next one after the primary?

So it's not hierarchical. And if you read our protocol, it is primary endpoint key secondary. Statistically, in our SAP that we've submitted to the FDA and discussed with them several times, because we're applying alpha to MDRI, the Multi-domain Responder Index, statistically, it becomes part of your primary endpoint family. So we will test those in parallel, not sequentially. You do not have to win on cognition to test MDRI. And traditionally, why the FDA does not like splitting alpha is because they say you're getting 2 bites at the same apple, which, in this case, it's really what we're trying to do here. We're really trying to support cognition with the idea that if we win on MDRI or cognition, we have hit a primary endpoint, we have a positive Phase III study.

So -- but they're not co-primary.

They're essentially -- so essentially, the distinction becomes important because of that hierarchy, they're both primary endpoints.

Can you hit either or? You can hit either or?

Yes. So I mean, I don't -- there's not a scenario. I mean, it just doesn't work where you wildly miss cognition and hit on MDRI. If you hit on MDRI and miss on cognition, it really means you're barely missing cognition. So positive MDRI barely missing on cognition, we would say we have hit a primary endpoint. We have a positive Phase III. Certainly, that doesn't obligate the FDA, never obligates the FDA to give you approval. They're always going to say it's based on the totality of the evidence, but we would be having a conversation with a positive Phase III study.

So on the MDRI, the MCID is 5 points as well?

So MDRI and why we've been working so long with the FDA to bring that forward is because you're bringing your key domains, behavior, gross motor, speech, sleep, language into a single tool to evaluate those individually, you have an MSD, a Meaningful Score Different for each. You negotiate with those FDA -- with the FDA ahead of time. Cognition is 5, something like receptive communication is 6. So we set an MSD for each of them individually.

And what's clinically meaningful for each subtype, for each subdomain?

Yes. So I mean, you set a separate MSD for behavior, gross motor, cognition, receptive language behavior.

Okay. So it's about -- you have to show a 1 point or 2-point change from baseline for each one and that rolls into the broader MDRI.

Yes. So like if you're just looking at cognition across the thing, if you hit 5 or greater, you get a plus 1 in that domain. If you don't, then you get a 0. If you do worse, you can get a negative 1. And then as you look across those 5 domains for an individual patient, if you hit on every domain, you're a plus 5 for the MDRI. So that's why it's a nice tool because it allows you to accept the variability that any neurologic indication happens. You're not doing sophisticated statistics where you can do funny things. And the FDA even acknowledged we can take this tool apart if we want to. So starting at end of Phase II, they really got on board with the MDRI. We had the conversation about alpha. We reinforced at our breakthrough designation meeting. So they're fully aware of our plans.

Because you're essentially validating the MDRI for the first time in this study.

Yes.

Any questions from the audience by any chance? Okay. Well, let's move to -- several programs to really discuss. Maybe just setrusumab because we do get a lot of questions on it. As you think from here on, there was a benefit and the one area that you did see more of a benefit is vertebral fractures and then obviously, in bone mineral density. Historically, vertebral fractures is not -- was less of a priority for FDA. And the bone mineral density, obviously, is a profound effect. The question is how it translates into function. The FDA is getting more comfortable with that in osteoporosis, but there's also 50 years' worth of data that doesn't exactly exist here. So as you think about a pathway forward, kind of how are you thinking about it strategically, how to approach the agency?

Yes. So strategically, we're really thinking about it in exactly the same way that we talked about it at JPMorgan. We have been talking about the changes in bone mineral density. They did hit statistical significant -- and to your point, with osteoporosis, they have 50 years of data. We're never going to have that in this case, and you're probably never going to have that in any rare disease. The mechanism of actions for these drugs are very similar. So it is very, very helpful. Yes, you need to validate any given biomarker for a specific disease, but it is very, very helpful that, that came out in December. So we are looking at that. That would obviously give you a pathway towards accelerated approval like we talked at JPM there. And then you need to support it for accelerated approval, you need a reason to believe clinically. The vertebral data was very compelling and also validated from the treating community and then really reinforcing once we've gone back to them talking about that drives a lot of disability, that drives a lot of pain. Oftentimes, that is honestly how they're treating, whether it's bisphosphonates or them thinking about the use of setrusumab there. So it's a good clinical endpoint. Obviously, we would need to talk about a confirmatory study. But then also PGIC, pain, there is supportive data that if you -- if we went down accelerated approval with bone mineral density, you would use that as the reason to support that and then you would need to do your confirmatory study in the long term.

And the confirmatory study would look at can it be a 24-week endpoint in fractures and pain? It would be a pain endpoint first or a fracture endpoint first?

I think if you gave me a complete do over and guarantee the results would be the same, I'd be fine as pain for a primary endpoint. Pain is a tough endpoint in any study. What's nice about the confirmatory study is you're on the market, you have a commercial drug, you're generating revenue that you can then reinvest into a confirmatory study. I think you're going to want to do something longer term and something we would not be able to do as an additional Phase III, but it's a very different setting there. And then crossover patient extension, I think, will be very, very helpful. I don't necessarily expect though they're going to give us a different answer than what we've already seen. I think you would -- in a confirmatory study, you'd want to do something bigger.

Okay. Got it. So maybe bigger and longer on fractures. When you say -- right, you say you're not expecting anything different from what you heard originally for the Phase III program?

Right.

And that would be presumably, I imagine with a younger patient population, the accelerated approval?

Yes. I think that's our job now, and that's where we are. And in a sense, maybe it sounds like we're taking a long time, but this is not just us doing additional analyses on everything. We need to talk to each treater. We need to understand what's happening with these patients on a patient level, like how are they feeling? What are they doing? Is this a case where people are feeling better, they're taking more risk, and that's why they're fracturing. The other side of that is there's always the possibility that your studies are telling you your drug is not as effective as you thought. So we really need to interrogate this on a patient-by-patient basis, and it takes time.

With respect to even pricing in that sense, with an accelerated approval, is the label different than you would have been otherwise and that carries some connotation to price? How are you thinking about that?

No. I mean, I think for rare disease, accelerated approval has worked very well. I don't think it's really impacted overall pricing and value of the program. I think for us, that can -- that would very much be a win. I think it's true for most diseases. It's certainly true for most rare diseases. Children do respond better. I think it's an obvious subgroup of patients to look at.

So I guess the label will be supported more on BMD and vertebral fracture. And I guess that's what I was thinking about, like the differentiation over placebo. So vertebral fractures would be sufficient.

Yes. Yes. I don't know if I should say this publicly or on record and stuff, but who looks at -- I'm not sure how many treaters actually look at labels, but that's fine. Certainly, we care about them, certainly, the regulators do and as do payers.

Okay. Maybe we'll move to -- we're all trying to read the tea leaves every morning of what's going on with FDA and then what's going on with FDA and gene therapy and it depends on the programs. For you in Sanfilippo, I think the feedback from FDA was that the data you've shown is sufficient clinically and it would not be approved, it would not be eligible. They're not going to be evaluating it for approval based on biomarkers. So that was a huge telltale sign. But in the meantime, there's obviously been kind of more challenges in terms of the manufacturing side and the IRL. What do you think is really going on behind the scenes here with FDA?

Yes. I mean, I think it hasn't been as consistent as it has been previously. And I think we're all trying to navigate around that and through that. I think it is good even with the IRL, the working team, the people working on these programs day-to-day, they have engaged with us. It's been clear with the IRL, we have said this is really a case of additional paperwork, additional written responses in contrast to the CRL, we don't need to do new work. We don't need to generate new data. So it will be a relatively quick response. I think if we understood that they wanted this as part of the CRL conversation, we could have easily have provided it. Luckily, once we resubmit, it's only a 2-week validation period compared to like your first initial with this 2 months. So it's a relatively small delay, but it's a delay, and we do understand these patients are out there, and they are continuing to accumulate damage and there's potential not to have as much benefit as they could. So we are eager to respond to that IRL and get back under review.

So it's really about SOP and protocols that they're looking for changes.

Correct. Yes. Everything is essentially paperwork or written response. We're not doing any new work.

You don't have to change -- it's a question of writing standard operating procedures of the way manufacturing just aligning the process with that, the written work. And the top point will be a 6-month review again? Or is that a 2-month review?

No. So you're kind of starting over again. So 2-week validation period up to a 6-month review.

And do you need to submit updated clinical information, clinical data each time?

No. So we did that for the CRL, and that's the big contrast here is that was a meaningful delta. And the FDA is always going to ask in that context for updated clinical data because they do not want to approve a drug and have you come back and say, I have 6 months more new data that may be changed something. So in that context, they're going to want new data. We presented that at world. It is very compelling. This is a very small delay, so no need for additional clinical data.

And the same manufacturing, the next product that is actually being made at Bedford is 301. 401, which is GSDIa, I believe, is being manufactured through a contract manufacturer.

No. So that's why we had -- that's why the CRL and Sanfilippo did result in a delay to our filing for GSDIa. We do make GSDIa at our facility here in Bedford. And that's why it was important for us to clear our validation period for GSDIa. We just press released on that. So again, that was reassuring that the FDA, it's still working, and we cleared that validation period. We got our PDUFA date for GSDIa. And to your point, anything specific to the facility in Bedford that was affecting Sanfilippo would absolutely apply to GSDIa, and that shows that we're moving through the process to answer their questions and correct anything that needs to be corrected.

So do you need to update or can you update those SOPs in parallel for 401?

Yes.

You can?

Yes. And the truth is we started our conversations with the filing for Sanfilippo when Peter Marks was in charge. So we had a conversation, and there was a much more flexibility on what could be done during a review period versus what had to be done with submission. You can call that an element of risk, but it was well discussed. When Peter left, we understood the changes coming through. So we took much less risk with GSDIa. So that package was at a different point than what we originally submitted for Sanfilippo.

In terms of the SOPs, is 301 made in the contract manufacturer?

So not yet.

Not yet. Okay.

Yes. Eventually, we will move everything into Bedford, but we need to do it. We accelerated the movement of GSDIa into Bedford just to have better control, but we're doing it one by one.

So 301 is external and then you could bring it back in. What's the timing? Just remind us seeing the data from the Enh3ance study and then when you potentially can file?

So for OTC. Yes. So OTC definitely has been quiet. I mean I think we talked about quite a while ago, and I think it probably was around really bringing Sanfilippo into the pipeline, and that really was a mature Phase III data set. So we made the prioritization with Sanfilippo and GSDIa. Honestly, good that we did given lift that's been involved there. But the Phase III for OTC has progressed. We've talked about doing an ammonia data readout at 36 weeks. That's important. Ammonia actually is a validated biomarker because of the work done with RAVICTI. So that is clinical in nature. And then full study readout is to week 64, and that's looking at responders as far as those patients who can discontinue alternate pathway medication and then liberalize their protein-restricted diet.

And enrollment has concluded or not yet?

Yes. No, enrollment was concluded quite some time ago. Again, I think, again, things were very dominated by OI and Angelman and that went back and forth. We were bringing forward Sanfilippo and GSDIa. That's been the spot like OTC has been working through. We did not stop investing or stop that Phase III trial. That trial has been ongoing in the background. And luckily, hopefully, we'll be getting both of those 2 gene therapies across the finish line and into commercial, and we'll obviously be pulling everything else up through the clinical pipeline.

Okay. So at this point for 301, you're waiting for the clinical, the 64-week data or there's other things that are going on in parallel?

Yes. No, I mean, obviously, we'll have the ammonia data internally, and we'll have that discussion. But then, yes, for full study readout to support a filing, that would be a full study readout and all patients getting to week 64.

Which presumably is this year, I imagine.

So we haven't updated guidance on that, and I'll just blame Josh.

Josh, we have another microphone. It's easy to give it to you. Okay. And then some of the questions we get a lot is for both of them. For 111, the data is compelling, huge unmet need, not a big market. For 401, how important is reducing corn starch intake? The data obviously is a lot more clinically meaningful when you look at your long-term extension, but still can it really support a price of $2 million, $3 million per patient per year commercially. Any thoughts about the level of unmet need here?

Yes. And I think to me, when you're talking about what is a successful gene therapy, what is not a successful gene therapy and look at what's happened in the commercial space today, I do think Sanfilippo is absolutely at the extreme end of unmet medical need. There's nothing you can do for these children. And by their fifth birthday, they are very damaged children and don't live past their teenage years typically. So highest unmet need, I agree, relatively smaller numbers between 3,000 and 5,000 patients in the territories we cover, but we think it will be a very successful launch and products because of that. I think GSDIa is a step down from that, but still with very high unmet medical need. And we've done a lot of educating externally with payers, regulators. The patient groups have gone and talked to the FDA, they've talked to other people because it's not corn starch. It's the fact that without that corn starch every 2 to 4 hours, they can get into a situation where they're so hypoglycemic, they can start to have seizures and it can be life-threatening, especially in pediatric age group. So it really is that honestly, daily threat that gets worse with any intercurrent illness where they can really get into serious medical trouble with that hypoglycemia.

And at that point, can you launch fairly quickly thereafter? Or is it takes time to build inventory? How does that work?

Yes. So I mean -- and that is part of the reason why we did make the prioritization around Sanfilippo and GSDIa because obviously, if you're going to be launched, you need to have that material on hand to launch. And for gene therapy, that is a big investment. So doing 3 in parallel would be a tremendous lift. We're prepared to launch and launch right away with Sanfilippo and GSDIa. Again, that allowed us with that prioritization to bring OTC and that level of investment in manufacturing up behind that.

Okay. And what can we expect this year, just minus in 701 as you continue to dose escalate?

Yes. So that -- and so I think Wilson is another example of talking about unmet medical need and being really in a situation where there are chelators and zinc available, they do a decent job of treating that disease. So yes, we clearly believe there is still high unmet medical need, but we've set the clear bar that we need the majority of patients off of chelators in order to declare success. So as we've moved through these dosing cohorts and with Cohort 4 at 4013, which is a very good dose for liver-directed gene therapies, we want to see the majority of patients off the chelators in order to declare success to move into Phase III. And doing that prematurely, I think, could you set yourself up where you maybe would not have as successful as a product as you could if you do not optimize that efficacy.

And so do you think we'll see some data this year? Yes. In the second half, most likely from the highest cohort?

Yes. So most recently, our guidance was this year in 2026, we want to give ourselves the ability because it's not -- it doesn't really help you to set an endpoint for patients coming off a chelator. So it takes -- if some are moving in the direction and they're close, fine. We'll wait for those patients to progress and then declare success. So we don't have a firm, firm, firm data on that.

Great. Maybe any final question from the audience? Maybe just to go back to Angelman, how far Aurora is still enrolling. So it sounds like Aurora is about 9 to 12 months sort of behind Aspire, 6 to 12 months behind.

Yes. I mean I think that's fair. We really did try not to overlap with our sites. We have a little bit of overlapping, but Aspire leads the way here. It is the most important of the 2 Phase III studies. So we wanted that enrolled and operationally under control before we launched the second Phase III.

Well, terrific. Eric, good to see you. Thanks for coming. We appreciate it.

Thank you.

Thank you.