I'm Joe Schwartz from the BioPharma Equity Research team at Leerink Partners. It's my pleasure to host our chat with Ultragenyx. We have Eric Crombez, CMO of the company, with us today. Thanks so much for joining us.
Great, thanks for having me.
Maybe you can start us off with a brief overview of the company's recent events and your priorities for this year.
Sure thing. I'll start with our commercial success we've had led by Crysvita. Certainly a lot of continued growth there. Great to see double-digit growth year- over- year. You know, along with Crysvita, seeing really great growth with Dojolvi and Evkeeza specifically, and also having MEPSEVII there in the background. You know, great background of commercial revenue there supporting everything we're doing on the pipeline. Then, you know, turning to the pipeline, obviously a big focus on Angelman with the data readout in the second half of this year, but also really nice to see progress with our gene therapy pipeline. We've been working on this for quite a long time, and with our PDUFA date now in hand for GSDIa, getting Sanfilippo closely done behind that, and then obviously not forgetting about OTC and Wilson there.
You know, a lot of depth in that clinical pipeline. Then again, with that maturity and getting these things across the finish line, launching into the commercial space, looking forward to the potential of bringing new things into the pipeline.
Great. Okay. Well, let's dive in. I guess we'll start with Angelman. In the Phase 3 study, we noticed that the primary endpoint will be measured without caregiver input, which seems a bit different from the Phase 1/2. Could you contextualize how caregiver input can influence Phase 1/2 study data? Is there any way that this might have inflated any of the responses that we're seeing? Or how do you think this nuance when going from Phase 1/2 to Phase 3 could influence results?
Yeah. The change was really driven by the concern around placebo effect. With Phase 1/ 2, obviously it's open label. Everyone's on drug. They know they're being treated. Using caregiver input is how the Bayley design, that's how it's standardly administered in all children. Going into a blinded study, obviously the parents don't know if their children are being randomized to active treatment or sham control. Certainly, everyone wants their child randomized to active treatment. They want their children to be doing better. In a blinded setting, you really do have to think about placebo control, and the best way to control for that, and it really does, for me, control that, is taking out that caregiver input.
That means when you're administering the Bayley in clinic with the standard Bayley with neurotypical children, if they will not do something in front of the assessor and the parent says they reliably, consistently do this at home, you can give them credit for this. Taking caregiver input out of it means they need to perform these tasks in front of the assessor in order to get credit for that. Really just trying to control that placebo effect to the best of our ability.
Okay. Interesting. In the Phase 1/2, we saw an improvement in the Bayley Cognition Score of more than 10. Is this what you're hoping to see in the pivotal? I think the minimal clinically important difference value is around five. Can you give us your thoughts on, like, how you're thinking about the magnitude of potential benefit that you need?
Yeah. I think, you know, first thinking about this statistically and what we need to hit statistical significance, I think it is best to go back to the slide we presented at the FAST meeting a couple meetings ago, where we showed change in cognition by Bayley-4, both by GSV and raw scores. We thought it was important to do this because GSV is the standard scoring for Bayley. That's what people are used to seeing, and that allows for development to continue over time in an individual child and just really tracking that over time. The FDA, they don't like any type of modeling, any type of adjustments to scores, so they always prefer raw scores and, you know, that's the conversation we had really all along heading into the Phase 3. It wasn't a surprise to us.
We understood that, but we wanted to show that data side by side to show that whether you're doing it by GSV or by raw scores, the results are similar. They're comparable. We still, you know, achieve greater than 90% power. Shifting to raw scores, because that's what our primary endpoint is based on for Phase 3, we did see in treated patients a 10.9 difference change from baseline in those treated patients. When we're thinking about performance of our placebo group, we obviously go back to natural history. Natural history tells us that developmentally, it's a very flat curve. It's not zero, but it's very flat. When we looked at the change in raw score from natural history, we actually gave that room to increase threefold when we were looking at how this could perform in a clinical trial.
Not because I expect it to be any different, but you have to allow for unexpected things to happen. When we were powering our studies and originally coming up with 120 patients for the Aspire study, we were looking at that delta of 10.9 in actively controlled patients versus that 1.2, that 3 x natural history in the placebo group. That's what drove that. Shifting to clinical significance, and I think the best thing to look at is how we build cognition into the MDRI, the Multi-Domain Responder Index. It really is the same as clinical significance, but the FDA has shifted and the field is kind of shifting to MSD 's Meaningful Score Difference. It is the same thing as clinical significance. You have to set and agree to that Meaningful Score Difference.
For cognition by Bayley, to say you have achieved clinical significance in the MDRI, it is a change of plus five that you mentioned earlier.
Okay. Super helpful. Thank you. Based on natural history data, it seems as though the standard deviation is greater for the raw scores versus the GSVs. Have you taken all of that into account in the power in the study?
Yeah. No, I mean, absolutely. Again, you know, there's kind of like what you expect if everything behaves in an ideal way, but then you have to, you know, give yourself a plus/minus accepting that things don't always go in an ideal way, and you just have to give yourself room there. You know, that's when we talk about. It was actually a question from one of the agencies in Europe, is why are you overpowering the study? Our answer was, it's neurology, and if you ever wanna overpower something, you wanna do it in neurology. Originally designed to enroll 120 patients, we over-enrolled to 129 because there was such demand, and we had a lot of patients in screening. For me, once they've committed to the study, they've signed consent, we wanna get them through dosing.
That even helps from a powering perspective.
Very interesting. I think MDRI could be considered more impactful as an endpoint 'cause it captures more about the child's development. Has the FDA actually agreed that a win on the MDRI could support approval if Bayley-4 primary cognition is not met?
Yeah. It was great to, you know, see the progress with the FDA, with the MDRI. You know, we came to end of Phase 2 expecting to have another good conversation negotiation around that. To be honest, they came to the table really kind of bought into the MDRI, what we were doing, using at a second primary endpoint, applying alpha to it. You're right, it is a powerful tool because you're looking at scoring across all of the key domains, how you're really looking at thinking about development and allowing for variability in these children because, again, it is neurology. You know, the FDA said it is a straightforward tool. We're not doing sophisticated statistics where you can manipulate anything.
It's a straightforward scoring system that you can always take apart, put back together again, and you know, they verbalize that back to us. We received breakthrough designation. We had a breakthrough meeting. They confirmed because we wanted to clarify again that they understood, you know, we're bringing forward the MDRI. We're bringing forward as a primary endpoint. We are applying alpha for this and making it a secondary endpoint, and they confirm, you know, that they're on board with that.
Okay, great. How much was the dose that you selected for the pivotal trial based on the relationship you've seen in terms of knockdown and protein restoration and how much was potentially limited by the lower extremity safety signal? I think previously the company was dosing much higher than you are in the pivotal, so I'm just trying to understand that dynamic.
Yeah. I think to me, and this is a little bit standard drug development, very true in this case, you know, you're using your animal model data, and we did have data in non-human primates, and that's, you know, obviously a good model to inform your dosing strategy for your Phase 1/2. The original dosing was done by GeneTx when they were running this program. I think, you know, really does stress that you do want to start low and build up with dosing with your dose finding because if you get ahead of yourself, it can really set you back there.
You know, once we brought that program in-house and restarted dose finding at a lower level, you know, we did that experiment, and we enrolled 74 patients in that Phase 1/2, to make sure we understood dosing and what was the right dose to bring into Phase 3. I would say that the animal model data really informed our Phase 1/2 design, and then it was the clinical data coming out of that Phase 1/2 that drove the selection for Phase 3.
Okay. Helpful. Thank you. I think the company has stated several times that GTX-102 is the most potent ASO in development for Angelman. Can you help us understand how this claim is informed? Is it based on preclinical data for ATS knockdown, mRNA or protein increases or all the above?
Yeah. Emil Kakkis, our CEO, and I are both geneticists, so we're used to thinking about it at the molecular level, and we're also at our core clinical development people. You know, yes, we are informed by what's going on in the research space, and again, it is important to understand that and inform everything. You know, we're really driven by decision-making on the clinical data. Again, with 74 patients worth of data in the Phase 2, we have a lot of clinical data. That claim is really based on what we're seeing from that Phase 1/ 2 data set, you know, and then the doses we're using and seeing a really good effect, you know, from across all domains there.
Okay. I think there was a recent peer-reviewed paper, which outlined that for locked nucleic acid ASOs, there might be a translation gap where almost 90% knockdown of the transcript was required to achieve only 50% protein restoration. Since GTX-102 leverages LNA chemistry, I'm wondering if you can remind us what level of knockdown and protein restoration you're achieving, and do you have a reason to believe that specific aspects to GTX-102's design could allow it to avoid any such translation issue?
Yeah. You know, with a lot of these recent conversations and understanding we're in a competitive space, there has been a lot of talk and conversation around this, and obviously there's a third party coming into the mix here. I actually went back to Scott Dindot. He's the scientist who started all of this work. He's been leading this work, and he continues on, and had another conversation around this. You know, really understanding the work we've done in non-human primates, understanding the relatively newer work that's been done in some mouse models. You know, we're in agreement, he is in agreement that you probably do want to have greater than 80% knockdown of that antisense molecule in order to achieve at least 50% expression from that paternal allele.
We do agree with that kind of threshold, if you will, and our data that we did in-house show that we can achieve that.
Okay. Do you have an idea of what amount of protein restoration is required to drive a clinical benefit in Angelman? How do you think about that?
Yeah. Again, you know, I'm a clinician. I'm always happy to go back to the experts and get their sense on this because, you know, this is what they've done, and they're much closer to it and know a lot more than what's published beyond in literature. You know, in talking to Scott that his thinking, our best thinking, it's, you know, it's probably around 35% to have 35% expression from that paternal allele or the maternal allele, depending on what you're doing to have a really good effect there.
Okay. How does Aurora fit into the picture for GTX-102, and how has that study been enrolling? Can you walk us through the filing plan here and what your expectations are around the label?
Yeah. Aurora is our second Phase 3 study. It's important because it does round out that label there. Yes, that's important, but to me, in all sincerity, I think what can be very hard in rare disease is when you're studying subsets, and then you potentially have a label that excludes a group of patients there. Our Aspire study, our main Phase 3 study, is still looking at patients with full deletions. That is the great majority of patients. Again, there are patients out there with missense mutations, uniparental disomy, and imprinting, and we don't wanna leave them behind. You know, everyone understands they can really benefit from this drug as well. Aurora is a true basket study, so we're looking at younger, older patients and patients with those different mutations there. It is open label.
It is going well, and the strategy there is really to do a straightforward extrapolation to Aspire. Establish comparable safety, get efficacy data that we can then bridge back to Aspire and, you know, ultimately have a full label for this drug.
Okay, great. Maybe we can switch gears to DTX401. We really like this program. We sometimes call it a sleeper to investors. It was great to see the BLA accepted last month. How is that review going?
Yeah, no, it is great, and it's great. Sanfilippo and GSDIa have kind of been jockeying to be our first gene therapy approval, with now having our PDUFA date in August in hand for GSDIa. I think it's there. It's going well. We cleared our two-month validation period. We received our PDUFA date. We're, you know, in that review period. We're receiving our IRs, our information requests, you know, we're at it, and I'm looking forward to August.
Okay. Fingers crossed. The agencies have been a bit of a stickler when it comes to CMC, especially, when it comes to gene therapy. Can you talk about how you feel about where you are in your manufacturing and ability to address questions during the review, given I think it's gonna be your first gene therapy product and your first one manufactured in-house?
Yeah, I think, you know, that definitely brings the Sanfilippo gene therapy into the conversation as well 'cause we are doing manufacturing for both of those products at our facility outside of the Boston-Cambridge area, and that's our manufacturing facility that we built from the ground up. You know, with manufacturing, with the challenges everyone's had for a very long time, we, you know, we thought it was important to bring that in-house, take control over that, and that has been a success and, you know, on all levels from manufacturing to cost of goods to everything there, and we really like having control there, and ultimately, we wanna bring all gene therapy manufacturing into that plant. That's been helpful.
You know, also with doing manufacturing for Sanfilippo and GSDIA at the same facility, that initial CRL, complete response letter, you know, did have spillover into GSDIA 'cause any findings that was specific to the facility, not to the product itself, would have applied equally. That's why we took extra time with GSDIA. We made sure we had everything right, finished, and complete for that initial filing for GSDIA. We really didn't take any risk there. We didn't wait to perform anything or finish anything and try to negotiate doing that during the review period. Again, it was great to get through that two-month validation period, and be underway now.
Very helpful. When we throw around the word curative, people often get hung up on the durability of a treatment. I think here with GSDIA, we saw the efficacy of cornstarch reduction actually get better over time. How are you thinking about DTX401's duration, and does that give you confidence in the FDA review?
It does, and I think what's nice is, and I started with our gene therapy programs as part of Dimension Therapeutics. I've been with these gene therapy programs from the beginning, and anything with DTX in front of it means that we started that at Dimension. We've been doing this for a long time, and we can always go back to those Phase 1/2 patients who have been in trial for a very long time and showing really strong durability. I think that was the question with liver-directed gene therapy is how often do your hepatocytes turn over? What is that gonna be? I think, you know, everyone had their thoughts there. I think, you know, we're seeing strong durability. We're seeing, you know, that the turnover of hepatocytes maybe is not as fast as we thought, and we really are holding onto those transgenes.
I think with the total package there, including those Phase 1/2 patients and the durability we're seeing there makes a very strong case to the FDA.
Okay. Well, let's talk about the primary endpoint a little bit. On the one hand, cornstarch reduction is a great endpoint since we can easily conceptualize how impactful changes in it are to patients. On the other hand, some people just wonder, you know, are you just swapping one thing for another, and how meaningful is it? Can you talk about some of the other endpoints, such as glycemic control or any others that come to mind that can help us envision what DTX401 could mean for patients and caregivers?
Yeah. We almost kind of shorthand the way we talk about GSDIa and reduction in cornstarch because it's always been this concept that it's been the reduction of cornstarch in the context of maintaining good glucose control. So that was always fundamental what we're doing here, and yes, you know, early on, and we've had this conversation, and I think we've made a lot of progress with the agencies, also with payers and reimbursers about understanding that, and it's not just a food product or something like that. Bringing in the patient groups to speak to the FDA was very compelling. We also did that in Europe, and that really showed the importance of this.
Prior to really understanding that you could use raw cornstarch to be slowly digested in your GI tract to give you two to four hours of glucose control took this from a universally fatal disease, that's how we really thought about GSDIa in children is they didn't make it out of the pediatric age group to something that could be managed with the use of cornstarch. Again, that's cornstarch every two to four hours, including overnight, and they still were not doing well. They still had problems with their livers, problem with their kidneys, high cholesterol, high triglycerides. They were still having a lot of metabolic problems there. They weren't thriving. They weren't doing as well as they could do.
Obviously, with gene therapy, you're getting to the heart of this disease, and you're letting them, for the first time, break down glycogen in the liver to produce glucose during times of fasting. You know, we've seen that as part of interviews that are formal part of the studies. We've seen that anecdotally where, you know, people feel like they can go back to work, they can travel, they can go away from home because they had this dependence on cornstarch and this fear if they were separated from it that this would be life-threatening for them. Yes, it's easy to think in a simplistic way about cornstarch, but it really was life-saving for them. These hypoglycemic episodes, particularly in pediatric patients, can remain life-threatening.
I get it. Interesting. Can you remind us what the standard immunosuppression regimen was in the studies and how you expect this to be implemented in the real world if it's approved, and what liver signal did you see?
Yeah. Again, you know, we started our gene therapy pipeline really focused on liver-directed gene therapy and that was really when we started with these inborn errors of metabolism programs really looking to what was done with in the hemophilia space for gene therapy, and that was the use of steroids. Using, for adults, 60 per kilogram and then doing weight-based dosing for pediatric patients. For us in our Phase 3, it's an eight-week bolus and then titration of cornstarch. If you have a rebound in LFTs, if you have prolonged LFTs, those steroids would be continued until your LFTs were in the normal range.
You know, I think with the patients we've dosed across our pipeline and what we've seen broadly with liver-directed gene therapy in the 1E13 range, we saw that typical, you know, rise in LFTs that were always clinically silent, resolved with use of steroids and, you know, resolved without sequelae. I think, you know, we understand this gene therapy, even administration needs expertise. These patients are complicated. They do need to be followed by people who understand this disease, and that means they're by and large in tertiary medical centers. You know, we've started identifying those centers where we're gonna be launching and doing this, and it. Yeah, it's something, steroids is something they can very easily handle.
Okay. How do you view the commercial opportunity then for DTX401, and are there obvious early adopters and later adopters, or is there another better way of thinking about that?
Yeah. I'm probably not always the best person to ask about uptake and penetration because being a clinician and being a geneticist and treating a lot of these patients, if you really have to take cornstarch every two to four hours. GSDIa is one of those diseases where you have a very consistent phenotype. You don't see this broad spectrum with these mild patients who are roughly fine. These are really people who need cornstarch every two to four hours. To me, if that's your lifelong regimen and having to deal with everything else that comes with that, then you will benefit from this gene therapy. You know, we talk about roughly 6,000 patients in the territory we cover, 20%-25% of them in the US, and for me, I think there will be very high penetration there.
Okay. Interesting. Switching gears again to UX111, this program's had a bit of a rocky path, but it seems like things are getting back on track. Can you give us your current sentiment and how optimistic you are that you can get it over the goal line?
Yes. You know, received our Complete Response Letter, CRL for Sanfilippo. That was obviously a surprise to us. You know, worked closely with the FDA to understand, and the FDA was, you know, very open to speaking with us to make sure we were able to satisfy that response. Unfortunately, we did receive that Incomplete Response Letter, our IRL there, and what that means is during that two-week review period, validation period, you know, in contrast to your two-month validation period for an initial submission, they came to the determination that we had not completely answered all of their questions in the CRL. What was required to satisfy the IRL was more of things like SOPs, paperwork, things we could just satisfy in writing.
Didn't have to run additional assays, didn't have to do assay work, didn't have to do any new work there. That means we can easily satisfy it. You know, if we were clear they wanted this as part of the CRL, we could have provided that. That means it's a very quick response timeline. Again, luckily, it's a two-week validation period, not a two-month validation. I'm expecting to get through this and receive a PDUFA date and really have those approvals for GSDIa and Sanfilippo very, very close to each other in the second half of this year.
Can you help us understand how well suited heparan sulfate is as a biomarker that's reasonably likely to predict clinical benefit in MPS? Are there different ways in measuring it that are important or other nuances? Any reason to question HS as an LS?
I don't think there's any reason to question heparan sulfate here, and I think, you know, the academic community, treating community, us in industry, we all understand we'd like this and find the correlation to be very strong there. You know, obviously we brought that along, you know, as originally as a basis for approval. I think what it comes down to is really what is your view on accelerated approval and the use of biomarkers generally not specific to heparan sulfate. We think heparan sulfate is a great biomarker here. You know, we think it could have been used for the basis of approval.
Unfortunately, but a little bit fortunately, with the duration of time we've had working through this filing, we are coming up to achieving the original FDA request to have all patients reach their fifth birthday, and that was gonna be their clinical outcome there. That's with the idea that by natural history, these children by five years of age are very neurodevelopmentally damaged, and if these children are still doing well, you've shown that your drug is effective. Where we stand today and with our resubmission, the majority of patients in this trial have reached their fifth birthday. We are really shifting away. The FDA said this during our last review period, that with the strength of our clinical data, we're really moving away from a biomarker accelerated approval to really looking at this as a clinical approval.
Okay. Great. Shifting to Wilson disease, we've seen a few interesting cuts of data from this program of yours. Now, for the upcoming data with the new immunosuppressant regimen, what should we be looking for?
Yeah. You know, we're taking the time to make sure we get Wilson right, and we really wanna see the efficacy there to really differentiate to chelators. You know, I think when you have a fairly effective treatment there, you know, that people argue patients are somewhat happy with, that you're really showing positive differentiation, you're doing something beyond that, and that's different than something with GSD1A. Yes, there's cornstarch, but the highest unmet medical need, you know, emphasized with Sanfilippo there. We've set a very clear bar for success. We wanna see the majority of patients off chelators doing well there.
You know, with going up in dose and using stronger immunomodulation, I think that's at least additive, if not synergistic there with making sure in this additional cohort and then looking at the data in totality that, you know, we have the right dose to bring into Phase 3 to have a successful commercialized product.
Okay. What have you seen so far, and what do you hope, or how do you hope that the next look compares?
Yeah. With our last data readout, we were very close to that bar of 50% of patients coming off of chelators. You know, we wanna obviously see if we can do better than that. What's nice about Wilson and being a copper transporting disorder is you can look you can look at copper in a lot of different ways. You really can be confident in what you're seeing, and if you're really establishing the normal trafficking of copper, yes, we are measuring clinical endpoints, we're looking at neurologic outcomes, and that's important. Having those types of biomarkers separate as a basis of approval really helps with decision-making in Phase 1 /2.
Awesome. Thanks so for all the insights, Eric. Really appreciate it.
Thank you.