Ultragenyx Pharmaceutical Inc. (RARE) Earnings Call Transcript
May 5, 2026
Earnings Call Speaker Segments
Thanks. A press release is out that breaks down our financial results, and it's available on our site, ultragenyx.com. On this call with me are Emil Kakkis, our Chief Executive Officer and President; Erik Harris, Chief Commercial Officer; Howard Horn, Chief Financial Officer; and Eric Crombez, Chief Medical Officer. Everyone should be aware that forward-looking statements will be part of today's call. Those statements carry certain risks and uncertainties, and it's possible our actual results could differ materially. The risk factors covered in our most recent SEC filings are worth reviewing. Emil, I'll hand the call over to you now.
Appreciate that, Josh, and hello everyone this afternoon. Sixteen years have passed since the company was started, and we expect this one to bring major change, with revenue climbing and several new drug approvals on the way. Our worldwide commercial operation should comfortably surpass $700 million in revenue, backed by a steady history of annual growth in the double digits. Two gene therapies carry a PDUFA date, and they would deliver the first-ever treatment for patients and families who currently have no other options that modify the disease. In addition, during the back half of this year, we will unroll the Phase 3 Aspire study, which is assessing GTX-102 in Angelman syndrome patients; meanwhile, we keep running global clinical trials spanning the biggest late-stage pipeline in rare diseases. Our gene therapy products are also being manufactured at the new Bedford, Massachusetts facility. Let me begin with GTX-102 for Angelman syndrome. The earliest patients who joined the Phase 3 Aspire study have now hit their day 338 visit and have moved into the open-label extension study. Over the coming months, the remaining patients will finish the blinded portion and cross over into open label extension; through the Aurora study, we are extending GTX-102 treatment to additional ages and genotypes within Angelman syndrome, and enrollment there keeps progressing nicely. Our preference is to create first-ever treatments for diseases that previously lacked meaningful breakthroughs, rather than merely operating in disease areas where rival programs sit at comparable development stages. With Angelman syndrome, we chose to make an exception because of the indication's size and the outstanding work genetics had accomplished alongside the scientist Scott Dindot to create a potent ASO that proved effective in a large animal model rather than only in mice, which frequently fail to predict human biology. The preclinical research Scott carried out in his lab at Texas A&M represents a genuine tour de force of molecular genetics. He can pinpoint and target a separate and distinct region of the antisense message transcript, which has produced more efficient transcript knockdown and stronger potency in the clinic. This superior science is what drew us to study Angelman syndrome. Later on this call, Dr. Crombez will detail the longer-term efficacy, durability and safety data from the Phase 1/2 study during his section. For now, though, I want to emphasize that we have 66 patients on therapy for an average of 3 years, with the longest nearing 5 years, showing continuing and improving benefits across multiple domains alongside a favorable safety profile. Given this longer-term data, we believe GTX-102 can produce clinically meaningful treatment effects and can be given safely in chronic dosing. From what the Phase 1/2 data show, we believe GTX-102 is making an important difference in these patients' lives and those of their families. Moving now to our global commercial efforts. We keep seeing results from our commercial operation; revenue is now coming in from over 35 nations, which stems from deliberately putting money into strong teams capable of handling the intricate approval and reimbursement systems worldwide with efficiency. That shows steady execution market by market, which our teams produce each quarter. Beyond bringing in substantial and expanding revenue, this foundational business serves as the force driving our upcoming growth stage. Right now our team is commercializing Dojolvi and Mepsevii, and they are set to take on DTX401 and UX111 too, since we anticipate approvals for both of those products later this year. By delivering first-ever therapies to patients in need, our proven worldwide commercial operation is clearing the way toward profitability in 2027. At this point I'll hand the call over to Erik Harris, our Chief Commercial Officer, to give specifics about what his team accomplished in the first quarter.
Thanks, and good afternoon to all of you. Like Emil said, the core business is still going strong, and we're in a great spot to take advantage of the possible launches coming up. Howard will walk through the specifics, but our confidence in the 2026 revenue guidance has not wavered at all. All through Q1, our commercial and field groups kept pace with rising demand for what we sell around the world. Let me begin with Crysvita, and I'd like to frame the Q1 revenue properly. The figures for North America, Latin America and Turkey lined up with what we had anticipated, given the order timing we've observed in earlier years. This is a business we understand deeply; we're seeing ongoing momentum in the fundamental demand everywhere we operate, such as markets like Brazil, where purchases in bulk by the Ministry of Health can create variability from one quarter to the next; in Latin America, roughly 30 patients started commercial therapy during Q1, which pushed the overall count of patients taking Crysvita past 950 in that region; we stay completely confident about the fundamentals and the strength of this business, together with our partner KKC in North America. Our capacity to keep identifying and treating patients who have XLH and TIO. Let me turn now to Dojolvi, where steady growth carried on, six years after approval. During the first quarter, more than 30 start forms came from our North American team, well above what they were aiming for that quarter; across North America, over 675 patients are now on reimbursed therapy. Meanwhile in Europe, about 300 are being treated via named patient or early access programs. Japan is probably where the next phase of growth will originate, since Dojolvi received conditional approval there last year, and we're eager for the product's full approval and launch in this quarter. I'll finish with Evkeeza, one more illustration of how our seasoned team can successfully commercialize rare disease products. Outside the United States, we're witnessing remarkable growth from this program as our international commercial teams work through reimbursement on a country-by-country basis and handle early access requests coming from patients and their physicians. Altogether, roughly 370 patients in 18 countries are on Evkeeza. We anticipate this will remain a significant and expanding piece of our revenue base. Past the performance of each product, I want to zoom out and point to what I consider the real strength of this business, which is the scale and reach of our commercial infrastructure worldwide. That depth of market presence across rare disease markets, which demand considerable expertise, the ability to find patients and navigate reimbursement, is a true advantage. It's what lets us bring first-ever treatments to rare-disease patients efficiently. And most importantly, this infrastructure and seasoned team are the reason I feel confident as we head into the remainder of the year. My group is getting ready to roll out two new products: DTX401, whose PDUFA date falls on August 23, 2026, and UX111, which has a PDUFA date of September 19, 2026. Across both of these programs, the preparations for launch are moving along nicely, and we're leveraging the systems and know-how that are already in place; so with that, I'll hand things over to Howard, who will walk you through our financial results and guidance in greater depth.
Appreciate that, Erik, and hello everyone this afternoon. I'll walk through the financial results from the first quarter as well as the guidance we're giving for the full year. Starting off with revenue: during Q1 of 2026, we brought in $136 million in total revenue. Crysvita contributed $93 million, which breaks down as $39 million out of North America, $46 million from Latin America and Turkey, and $8 million from Europe — every one of those figures matched the quarterly timing and patterns Erik had described as expected. Dojolvi chipped in $18 million, consistent with the steady demand growth we had anticipated. Evkeeza likewise added $18 million, which marks 64% growth compared with the first quarter of 2025, as demand keeps building after launches in our territories outside of the United States. Finally, Mepsevii added $7 million as we keep treating patients in this ultrarare indication. Operating expenses totaled $305 million for the quarter, including $30 million in cost of sales and $275 million in combined R&D and SG&A expenses. Within total operating expenses were $30 million of non-cash stock-based compensation and $30 million of expenses tied to the restructuring announced last quarter; net loss for the quarter came to $185 million, or $1.84 per share. At March 31, our cash, cash equivalents and marketable securities stood at $534 million. For the quarter, operations consumed $197 million in net cash. It's worth remembering that in Q1, our operating cash burn is usually heavier than what we see in the following three quarters, because this period carries things like yearly bonus payouts. Beyond that, Q1 of 2026 also reflected $38 million of payments related to UX143 manufacturing work, along with $5 million for severance and other costs stemming from the workforce reduction we did recently. As we remain on track to reach profitability in 2027, we anticipate that net cash used in operations will come down over the rest of this year's quarters. Now for our 2026 guidance. The revenue outlook we shared back in February still stands. We expect total revenue for 2026 to be somewhere in the range of $730 million to $760 million. That works out to growth of 8% to 13% compared with 2025, and it doesn't factor in any possible revenue from products we haven't launched yet. For Crysvita, we're looking at $500 million to $520 million, which accounts for every region and every type of Crysvita revenue that flows to Ultragenyx. That range is driven by rising underlying demand worldwide, though it's partly held back by ordering patterns in Brazil that we expect to even out in 2027. Dojolvi should land somewhere between $100 million and $110 million. On top of that, the R&D and SG&A guidance we gave on our previous call remains unchanged. To be more precise, for 2026 we're looking at combined R&D and SG&A spend coming in somewhere between unchanged and a low-single-digit decline compared with 2025. Beyond that, our outlook for 2027 hasn't changed either — we still anticipate that combined R&D and SG&A costs will fall by no less than 15% relative to 2025. On that note, I'll hand things over to Eric Crombez, our Chief Medical Officer.
Thanks, Howard, and hello everybody this afternoon. Like Emil said at the start, my remarks will center on the clinically meaningful Phase 1/2 data we've produced with GTX-102, the antisense oligonucleotide we're developing to treat Angelman syndrome; this information comes from our open-label, single-arm Phase 1/2 trials, and it helped shape how the Phase 3 double-blind, sham-controlled Aspire study was designed. Because the study designs differ, you can't necessarily expect these Phase 1/2 findings to predict what Phase 3 will show. In the press release we put out earlier today, we underscored that 74 patients in total have received GTX-102 within our Phase 1/2 program. That program ranks among the biggest Phase 1/2 efforts we've run, and its purpose was to guide the dose, the dosing schedule and the endpoints for our Phase 3 trials. At present, 66 patients are enrolled in long-term extension studies, and they are generally given the 14-milligram quarterly intrathecal maintenance dose. On average, these patients have been receiving therapy without interruption for 3 years, and a few of them have now reached their fifth year of treatment. Improvement continues to show up for these patients across several developmental domains, and there have been no new instances of transient lower extremity weakness. Back in March of this year, we took a data cut so we could showcase the long-term safety and efficacy of GTX-102 in a manuscript that's coming out soon, and today I wanted to give you a high-level summary. Beginning with the Bayley-4 Cognitive raw score: 53 patients in the Phase 1/2 study have such a score at month 12. At that point, their mean change from baseline was nearing 10 points, which is above the meaningful score difference of 6 points, and with longer follow-up, the cognitive benefits keep growing and getting better. Now for the multi-domain responder index, or MDRI, which relies on clinically meaningful score differences in line with FDA guidance rather than changes driven by statistics to classify a response as positive or negative across 5 separate developmental domains. Within the Phase 1/2 data set, we now have MDRI results at month 12, 24 and 36 assessing response in cognition, communication, behavior, sleep and gross motor. Comparing responses against baseline, the p-value across all three of those time points comes in below 0.0001. Beyond being a highly robust statistical assessment of five distinct measures, this aligns with how physicians and families view neurologic conditions such as Angelman more broadly; it's the convergence of all these developmental changes that captures how individual patients genuinely respond to treatment in a holistic manner. In the 48-week Aspire Phase 3 study, we split the primary statistical alpha between the Bayley-4 Cognitive raw score and the MDRI. There's no hierarchy here, which means both endpoints get tested simultaneously. A statistically successful Phase 3 study would be achieved if the Bayley cognition measure comes in below 0.04, or if MDRI comes in below 0.01. We also examined expressive communication within our Phase 1/2 study, as assessed by the Bayley-4. As with cognition, meaningful improvements in expressive communication were observed, and these continued and got better during longer-term follow-up. Considering the totality of data generated across our Phase 1/2 program, my confidence remains that the developmental progress and continued learning of new skills seen in these patients support the meaningful benefit of this ASO in providing UBE3A from the paternal allele. I look forward to seeing results from our Phase 3 studies and the potential to replicate these results in larger controlled studies. I'll now hand the call back to Emil for a reminder of our catalysts for 2026 and some closing remarks.
I appreciate that, Eric. To wrap up, I want to touch on several upcoming milestones for later in the year. You can find the complete list in the corporate deck that's available on our website. First, DTX401, which is for glycogen storage disease type 1a — our collaboration with the FDA continues to go smoothly, and we're anticipating our PDUFA action date of August 23. We've also been told by the FDA that no AdCom is scheduled. Moving to UX111, intended for Sanfilippo syndrome, the BLA resubmitted earlier this year is under review, with a PDUFA action date of September 19. And finally, GTX-102 for Angelman syndrome remains on schedule to report top line Phase 3 data from the Aspire study during the second half of the year. Ultragenyx has ranked among the industry's most productive rare disease companies, shepherding programs from early research all the way to approved therapies for patients with no other options; we've accomplished this across multiple modalities and multiple therapeutic areas, and we're eager to deliver the next set of first-ever treatments to the patients who need them. The surge in late-stage programs over the last few years has challenged us as it would anyone. Yet the payoff is that once we turn the corner on these programs into approved products, we have the potential for a significant acceleration in our growth that will be a special moment in the history of Ultragenyx. With that, let's move on to your questions. Operator, please provide the Q&A instructions.
This is Daniel calling in for Tazeen. I was curious whether you might give us some thoughts on the recent Angelman update — for instance, what degree of consistency are you observing among patients who are getting better on the Bayley-4 and the MDRI? And also, how ought we to think about variability across those 2 endpoints for the Phase 3? I'll say a little about that, and perhaps Eric can chime in with more. The Bayley testing we're carrying out is being handled — mostly by an outside group that visits the site and administers the test; so we're operating a very high-quality effort in regard to how we'll measure the test, which helps lower variability. As for consistency, that's not something we can speak to at this stage of a Phase 3 study, though we can describe what — how it appeared in Phase 2. It should — our expectation is that it will resemble what we've observed previously. And the MDRI, I think, has proven a very robust and consistent measure simply by its nature, since it covers multiple domains where we've seen strong results, and I believe Eric just discussed those. So Eric, what's your view regarding the Bayley-4 consistency?
Yeah, so when we look at what came out of Phase 1/2, that really shaped the way we put Phase 3 together. The key move on our part was to allow in patients who have full deletions. What that means is they make no UBE3A at all, which gives you a really uniform group of patients, and that's what produced the steady results we observed in Phase 2 and what we expect to see again. And once more, just to remind everyone, we're enrolling patients with other mutations that introduce more variability into Aurora, our second Phase 3 trial.
That's a really insightful comment, Eric. Having a uniform genetic type is certainly going to assist us in achieving consistency, especially when it comes to how the placebo or the untreated sham performs, since in the absence of any treatment, individuals with Angelman don't show meaningful improvement on the Bayley-4.
I've got a few things to ask about GTX-102. When it comes to Aspire, are you making the randomization stratified, or are you defining subgroups ahead of time based on any baseline characteristics? In rare pediatric studies, it appears that even slight differences in how severe things are at the start might matter, and there's some published work in Angelman suggesting this could affect how much things vary. I'm curious — how sure are you that Aspire is shielded from any possible imbalance at baseline?
Eric will give you the specific answer. Broadly speaking, across every one of our programs, we're quite mindful of this entire issue of skewing when it comes to randomization; so we will always — for primary endpoints, we'll stratify our primary endpoint to do the best we possibly can. That said, nothing is flawless. So here's what we're doing for Angelman.
Right, so when it comes to Aspire in particular, we're looking at two things — age and the raw cognitive score, which is our main endpoint — and we want to be confident that the balance between them holds up consistently.
Okay. And after that, we saw you still haven't given a tighter window for when the Aspire top line data will come. Any idea at what point in the back half of '26 that data might be reported? Would late in the third quarter be a reasonable guess, given that enrollment wrapped up in late July of last year? It's amusing, right? Leaving you guessing about precisely when that will occur. Yeah. I figure you can likely infer it from how things are timed. The thing is, when you shut down a trial that ran internationally, with so many endpoints, you want to handle that process carefully; and so the reason we aren't specific isn't just that we want enough time to have all of it in order before we begin the unwinding; we did say the second half, though from when the last patient went in, you can roughly tell when the last patient should come out; however, the amount of time needed to wrap up a study involving sham, involving EEG, involving plenty of other elements, that will require a bit of time. So with a Phase 3 program, we'd rather spend that time and stay somewhat vague at this point, but everything remains on schedule.
The longer-term remarks on Phase 1/2 are useful. To be clear, might we get a look at the detailed Phase 1/2 data before the Phase 3 top line comes out? And it appears you are observing a distinct static signal for MDRI in Phase 1/2. And given the point improvement you referenced for cognition, I am curious whether a static p-value for cognition is showing up relative to your — relative to the natural history study data set.
Regarding the primary endpoint, I believe we have previously discussed the natural history comparison serving as our powering analysis. I believe we stated that we would be well powered. So I think that has already functioned as the analysis. MDRI is simply an extraordinarily strong method. In my view it is the path forward for neurologic diseases. Your question was whether you will get to see the detailed data. Most likely not. I do not think we have determined which science meeting the data will be presented at yet. My assumption would be later in the year; frequently we have presented items at the FAST conference, yet no plan has been fixed at this moment for the Phase 2 data, though it is not necessarily before Phase 3 at this stage. However, we wanted to release a small amount of data at present simply to provide people confidence regarding what is happening and to offer you a sense of the magnitude and outcome and the reality that we feel confident about how the drug appears; and it keeps demonstrating favorable safety along with the sort of cognitive benefit and the appropriate benefit that leaves us confident in the design and in what we are running in Phase 3.
Understood, that clears things up. Perhaps another item then. I'm curious whether, before the stats plan's database lock, any conversations with the FDA are needed regarding how large the improvement must be on the primary endpoints? Or does merely demonstrating a static benefit count? Would that be enough?
Right. At this stage, nothing regulatory stands between unblinding and us; everyone is in agreement. Also, for the Bayley, no minimum change is specified as required. My view is that's not something the design needs, since it's a continuous variable. That said, a clinically important change of 5 points is what we'll always have, so to speak. Data will always be examined by us in that manner, yet powering for the Bayley cognition primary endpoint rests on a continuous analysis, and honestly that's the proper approach; however, looking at the figures just shared with you, 10, that's already nearly double the minimally important difference; therefore, I'd say we're sitting in a fairly good spot.
I'm Joyce, covering for Anupam. Maybe a question regarding the pair of gene therapy rollouts on the way, DTX401 along with UX111, during the year's latter half. Might you discuss the current situation with commercial manufacturing and product scale-up, plus how prepared you are overall for a launch — from a CMC launch standpoint?
Yeah. No, that's a solid question. As I see it, we've genuinely been producing since the prior year. When it comes to DTX401, production of both the drug substance and the drug product happens at our Bedford facility. As for UX111, drug substance is supplied by a contract manufacturer located in Ohio, while fill-finish is completed at our Bedford site. Throughout last year and continuing into this one, we've been operating there, and a portion of our expenses reflects funds going toward inventory buildup. That effort has progressed well — we're accumulating inventory and believe we ought to be sitting in a decent spot by launch time as things stand now. So my sense is that we're prepared, and the launch teams have certainly been putting in their work. The two PDUFA dates fall pretty close together, I'd say. That said, the physicians we're calling on happen to be one and the same group. Having the same — most sites will handle both products, while a few might take on just one — creates definite synergy. Meanwhile, the qualified treatment centers are being established and contracts are being finalized. I believe real synergies will come from having both of them, and the possibility has us excited. Obviously, approval is still something we have to secure; even so, right now we're ready and progressing on the manufacturing side as well as the commercial side, and we're enthusiastic about what launching two gene therapy products could bring.
Could you provide additional details regarding what timeline we ought to expect for receiving results from the Aurora study? Also, assuming positive findings come out of the Phase 3 Aspire trial, what do you envision as your base case concerning the age ranges for a possible Angelman label? Would that cover adults as well?
Aurora, then, serves as our path toward broadening the label into additional indications, genotypes, and age brackets; the primary Aspire trial covers ages 4 to 17, exclusively deletion. Consequently, Aurora involves a younger cohort, an older cohort, plus other genotypes. Enrollment for that study remains ongoing and is proceeding nicely, yet since it is also conducted internationally, we will keep gathering data. As a result, we expect to have certain data by then. Up to now, we have not truly disclosed anything regarding our strategy for filing and launch; our approach is to hold off and observe what the data show, then assemble our plans. Nonetheless, the Aurora study still requires further enrollment.
Steven here, filling in for Yaron. I truly value the detail you provided regarding Angelman. You said the 14 mg was typically the maintenance dose in use. Could you shed some light on whether additional doses might be in play, what would explain that, and whether the label could include more than one maintenance dose? And second, regarding profitability forecasts, since PRVs are currently fetching well above $100 million. Any insight into what you're assuming for PRV monetization, and can we count on that full year '27 profitability being sustainable?
All right. Well, I'll address the dose at a broad level. I'm not sure whether Eric has anything further to add, but then I'll have Howard handle the — PRV is a profitability question. So the overwhelming majority of patients are 14. And based on the main trial, the protocol in fact is Aspire brings them to 14 in terms of how it's done. Some patients certainly have been on drug for a longer period of time under various regimens. So I don't know if there's anything else to add it. We think it's a very high fraction of around 14.
Right, and that's the label we're aiming for. So a maintenance dose of 14 milligrams every 3 months is what we anticipate. When we move into a commercialized setting and have our DMP, we could look into potentially a different dosing approach. Perhaps certain patients might do better on an every 2 months schedule, and some might also do better with a higher dose there. But our plan is to — following talks with the FDA — go with a 14-milligram maintenance dose.
That's really the main source of what you're going to see; so Howard...
When it comes to the PRV, we're keeping a close eye on how those have been selling. In our plans, there are 2 PRVs we intend to monetize — 1 for 111 and 1 for 401. I believe we've said before that our model already assumes a bit more than $100 million for each one. So anything beyond that would be upside. Naturally, a PRV with 102 could also present an opportunity, and that too would be upside relative to our model. That's our thinking on monetization. I think you also raised the matter of whether profitability can be sustained after '27. What we intend is not simply to remain at the Raiser's Edge, but to keep growing profitability. And based on which launches we have and how well they do, we'll determine how much we can put back into the pipeline versus let fall to the bottom line. But that's a conversation we'll enjoy having when the time comes.
Before we get the Phase 3 Angelman's results in the second half, could you address what the path looks like if MDRI succeeds while Bayley-4 does not?
As Eric mentioned in today's script, we'd still view that as a positive study. Basically, we have two routes to success. Bayley Cog could work, and MDRI might offer a stronger alternative. We worked out this stance with the FDA — although they typically prefer single primary endpoints, our view is that MDRI is genuinely a smarter, superior approach for neurologic conditions; here's a chance to head that way. What we anticipate, as Eric said, is that if Bayley-4 misses but MDRI comes through positive, that still shows efficacy for the program. Naturally, that scenario assumes, say, Bayley cognition narrowly fails while MDRI succeeds. That's one possible case. We wouldn't anticipate, for instance, Bayley cognition turning negative while MDRI turns positive and having that pan out. So the real question becomes: could Bayley fail for some reason while MDRI acts as an insurance policy and backs efficacy for the product in a way that extends beyond cognition alone? So we believe there are options — two ways for the program to win, and in fact we expect both to succeed; yet as you're aware, despite our best intentions in a randomized controlled trial we can always miss, though combining the two, I think, raises our odds of a positive outcome no matter what occurs with the patients.
Regarding the most recent Angelman analysis you reviewed for the Phase 1/2 data — could you share how the 1-year plus results are accumulating for cohorts A and B compared with the various measures you observed for cohorts 4 and D? Also, has that further reinforced your confidence in the dosing regimen and techniques you advanced into Phase 3?
Alright. Yes; we have presented cohort A and B previously, and in my view cohort A and B keeps performing as expected. And I would note — keep in mind, those represent the majority of patients. C and D involve comparatively few patients. It is truly mainly A and B. So I believe what we are discussing is cohort A and B now extending across the entire year, and that is the dataset under discussion. So it is genuinely based on A and B. That A and B, in our assessment, is fairly similar to what you are observing in Phase 3, since A and B covered all the international sites. So that encompassed the 7 or 8 countries where we are also running studies, with a substantial overlap involving the sites we are actually utilizing; therefore I think the A and B cohort, which serves as the main driver of the data you heard about today, is highly reproducible with respect to our Phase 3 activities, whether in terms of sites, countries or patient type. So I consider it a solid model. I believe the data ought to align with what we anticipate in Phase 3.
This is Selena calling in for Max. Regarding Angelman, might you explain how much caregiver-provided information adds to the Bayley-4 cognition results in Phase 1/2, and how that shifts as time goes on with the extended follow-up data?
Thank you for such a technically precise and detailed question. I'm not certain everybody was aware of caregiver-provided information, but let me clarify: with the raw scores we use for Bayley-4, the FDA doesn't want caregiver-provided information included, correct? Therefore, a psychologist tester must carry out all of the analysis. As a result, per the FDA's request, we include no caregiver-provided information in the Bayley-4 primary endpoint. Now, from the — let me pull up that figure, is it 20 or 30 items, roughly? There are around a handful of items where caregiver-provided information — 2 or 3 — can make a difference. Thus, when we examine raw scores with or without caregiver-provided information, for Bayley-4 cognition we don't believe there's a meaningful problem. If you're administering the expressive Bayley, there's in fact far greater potential for caregiver-provided information. Consequently, it may have a larger effect on the — excuse me, expressive communication rather than on cognition. So it's merely a small number, 2 or 3 things, where we haven't observed any impact, yet within the Phase 3 trial we are doing it without caregiver-provided information.
Emil and Eric, my sense is that what was being asked had more to do with the Phase 1/2 data we've discussed — whether we could examine it while leaving out the caregiver input — and my understanding is that the Phase 1/2 wasn't conducted that way.
Right, that's accurate. In Phase 1/2, caregiver input was indeed included, and during our discussions with the FDA while designing Phase 3, they asked for that. In Phase 3, we in fact ran the assessment both ways, though caregiver input isn't part of the primary endpoint. Removing bias is why that matters, and this applies equally to your actively treated group and your control group, so the request makes sense.
Alright, then we can run the analysis excluding those entries, and they won't move the needle in any real way. That's exactly my earlier point: if we take them out, nothing changes materially, since only a handful among a huge set could possibly matter, correct? So here's the picture: we're talking about those 2 or 3 entries. Suppose the caregiver reports they believe development is happening, yet the physician doesn't observe it — if the physician does observe it, that could add a point or 2. Still, our claim is that removing them leaves cognition unaffected; therefore what we're seeing in Phase 1/2 would hold either way, whether or not caregiver information is included, right? Does that address your question, Josh? Are we clear? Wow, we're really going deep on COA.
Nice work on the advances. I'd like to find out what you're anticipating regarding sham performance in the Aspire study. There was a limited data set from Angelman, I realize, but have you examined any other surrogate neurodevelopmental indications with respect to sham performance? Then my second question concerns profitability. Could you comment on what you require in terms of success from the Angelman gene therapy programs that gets factored into your profitability assumptions? Are those included in the profitability guidance, or does it reflect only the current base business?
I'll hand the profitability question to Howard, and I'll take the sham performance one. Clearly, an effect from sham isn't something we'd expect. Every possible study hasn't been reviewed by us, though in our own — in Angelman and the control studies we've seen — not much change has shown up. What I'd tell you is that regarding how development performs, a placebo effect isn't really what these kids will experience, correct? The reason is they aren't thinking, "Hey, I'm getting a treatment, I should be better." That's not how they think. They aren't — changing developmentally because of an absence of caregiver input, which may be the reason the FDA prefers it, since it permits a more objective assessment of what's happening. So what we're really asking is whether they could improve on their own with nothing going on. Our position is that the deletion patients simply don't; thus we're fairly confident sham performance won't matter. Whether comparable developmental disorders exist, I'm not sure — perhaps Rett syndrome or similar types. At this stage, though, our feeling is that sham ought to be no different from a natural history, and when it comes to their Bayley cognition scoring, the deletion patients are quite stable. Not much change occurs, and we're discussing under 1 point per year. That holds whether it's a randomized trial or even a natural history. So currently, we're at ease with it. Still, I take your point right now, we seem at ease with it. And even if a higher background signal existed, we have adequate power, and we should have the power to overcome that given the treatment effect we expect.
And Jack, on the assumptions around profitability and successful launches, at this point, what we've communicated on the top line is that we're assuming ongoing double-digit growth from the products we have now, along with a contribution from launches still to come, which might include 111, 401, 102; beyond that, we haven't really said anything further, though perhaps I can add that it definitely isn't necessary for every one of them to succeed in order for us to reach profitability. And there are various levers available to us that we can pull so that we're able to deliver on our commitment of a 2027 path to profitability.
I'd like to inquire regarding the GTX-102 program, and it's wonderful to learn about the data over the long term. I believe you said 66 patients were involved in the long-term extension. Might you speak to, sort of, what led to discontinuations? And then I have a follow-up question.
Sure, that's something we can address. Honestly, it has been somewhat sporadic, though it tends to occur mostly at the start. Eric, perhaps you could weigh in on that for him.
Right. So it was very consistent — every one of those cases came down to how much of a load taking part in the study put on people. I mean, we have to keep in mind that a number of these patients are located quite a distance from the places where they get treatment, and for those families, that really does turn into a hardship. So that held true across the board — study burden was the cause behind those discontinuations.
Occasionally that occurred rather soon as well, since certain individuals came to see they simply weren't able to manage it. Still, to be clear, no safety issue was involved.
All right. All right. Then the second thing I'd like to raise concerns Crysvita. So I do value the remarks you provided during this call regarding certain fluctuations in how orders come in. But I suppose what I'm really asking is how much clarity can you offer looking ahead over the coming 2 months or so? And what exactly is it that makes you feel assured about the annual guidance figure for Crysvita?
Right. I'll have Erik weigh in a bit here, since Erik works quite closely with the caring team on what's happening. Yet I believe there's another piece to this: each year we'll have an additional variable, namely that the royalty works such that early in the year it sits somewhat lower until it passes the threshold and then rises; consequently, whatever revenue arrives, our percentage climbs as the year progresses, and that explains why quarterly revenue climbs too. So this is a matter of ordering patterns, and there will be a steady escalation of our percentage within the royalty stream. Thus it simply resets annually and climbs back up. Erik, what makes you feel confident about Crysvita moving forward across the North American territory?
Yeah. No, it's pretty straightforward; our confidence is high regarding the demand we keep observing at the base level — locating patients and patients who want treatment throughout our regions — and that's the reason we reiterated our guidance. And just like in earlier years, we anticipate that identical sawtooth shape we've witnessed before: a weaker Q1, then a bounce back in Q2, a gentler Q3, and a robust Q4. So we expect, as we did in years past, to keep fulfilling what we promised to the street.
Right. What I’d expect us to observe in Q1, though, is essentially those beginning forms. What’s the beginning? That’s the demand you mentioned. And then — as well as continuations, and so on. So demand for Crysvita is strong. It’s an excellent product. Once people begin taking it, they remain on it. And they keep expanding it, and we’re here to back them up and carry out our own efforts where we’re commercializing. But we — Crysvita has kept growing and performing well. And we simply need to recognize that there’s always this corresponding plan that will occur in nearly the same way each year. So we feel at ease about how the year will unfold.
Regarding the data from the long-term extension, I was interested—you noted positive gains across several domains, with patients still progressing in their development. How do you evaluate those measures when it comes to whether the Aspire trial has adequate power for the Bayley-4 cognition endpoint? Do the developmental improvements observed in the long-term extension study give you confidence that the statistical power for Bayley-4 is enough in Aspire? Also, what clinically meaningful delta would you be looking for on Bayley-4?
Yeah, Yigal, I appreciate the question. Well, I believe Eric recently mentioned reaching approximately 10 points, say, within the 1-year window. That covers the portion of the A and B cohort plus the others combined who have reached the 1-year milestone. Certain individuals are at the 3- to 5-year milestone. Our point is that at 1 year, they are achieving a figure quite near what we had anticipated all along, and that magnitude demonstrates we have sufficient power and it sits well above the MID at 1 year. So we feel at ease with that magnitude; incidentally, nearly everyone familiar with that particular test is astonished by how much movement occurs, since typically it barely budges for most individuals or experts, including Kim Goodspeed, our physician and an Angelman specialist. According to her, you simply never witness this measure shift; thus 10 points represents a substantial figure. What we are stating, and what Eric presented, is that as you keep tracking these children across a longer period, their Bayley cognition scores keep rising. That essentially indicates patients are experiencing lasting benefit from the ASO on how their brain functions, and they keep making further gains. To me, that matters greatly. And that is what the long-term data reveals. Naturally, Phase 3 is crucial for securing approval and filing. Yet when considering a product's commercial potential, I believe the long-term safety profile along with continued treatment and ongoing improvement demonstrate this therapy can assist children across an extended timeframe in the post-market environment. Therefore, with Phase 3 results available and the ability to reach a commercial setting, I think this bolsters our confidence that the drug could deliver long-term benefit for individuals with Angelman syndrome.
All right. Then a question about UX111. Following the resubmission, has the agency asked for any additional information after it accepted the filing in April? Or are there any further inspection obligations once that BLA acceptance came through?
Yes. What's happening is what I'd describe as standard back-and-forth with the agency; typically we don't get into specifics about those; on the BLA we've been having what I'd characterize as ordinary conversations, and they're ongoing; and right now, our sense is that things are proceeding as usual. And we won't go into much detail on that until a decision is reached. That program's potential has us enthusiastic, and we're preparing for a launch, provided approval comes through.
An unnamed analyst: I'm phoning in on Gavin's behalf. There are a couple of things we'd like to ask. First, about the Angelman MDRI endpoint — I'm hoping you can spell out the statistical approach with total precision. When you calculate the delta versus the sham control, what's the actual definition? Is it the difference between the two arms in median net response, or is it the disparity in the proportion of patients who show a net response on one or more domains? Then second, we wanted to ask about 106 in GNE myopathy, since we saw it just got IND clearance. So I'm curious what the key distinction is between this new molecule and the earlier one that appears to have gone through the same indication and failed in the Phase 3. What key takeaways from that historical study can inform how the current trial is designed?
Sure. Thanks. So regarding the MDRI, the thing we're examining is net domain improvement. Each patient, then, will have some number of domains that they hit, correct? And what we do is compare the distribution curve for treated patients against the distribution curve for control patients on net domain improvement, right? So it's net domain improvement — how many positive wins per patient are you seeing? As we've shown before, on average there were usually at least two domains affected per patient, around that figure. Those domains, though, aren't always the same, and which ones occur together can differ. In certain patients, as many as four or five domains improved. So we're going to look at those distribution curves for net domains per patient, correct? That's the statistical comparison being made. With GNE myopathy, I've been part of that program since its earliest days; it's a terrible muscle disease, we estimate perhaps 10,000 patients worldwide, and we consider it a very chronic and awful condition. The first sialic acid replacement molecule wasn't as effective, since the sialic acid simply couldn't penetrate. The novel medication is a prodrug we created ourselves, and it possesses greater hydrophobicity, which aids in directing it toward muscle tissue. Once inside, it gets liberated and carried along by a sialic acid transporter. This mechanism permits entry into the lysosome and passage through the lysosomal membrane into muscle tissue. Thanks to that enhanced hydrophobicity approach, when we administer it to animals—or to dogs—we can supply substantial quantities of sialic acid to the cells far more effectively than with alternative compounds, outperforming them by many multiples. Consequently, we believe its potency is vastly superior, which better positions us to genuinely deliver the replacement therapy that patients need. Based on earlier studies, especially the biopsy data, we know these individuals lose 85% of the sialic acid in their muscles. With this drug, we ought to achieve full restoration — possibly even exceeding normal sialic acid levels. That gives us more optimism that it could work as a treatment for GNE myopathy. Let me be clear: funding for this program comes through a venture philanthropy partnership with the patient organization, and they care deeply about what happens next with the product. Despite the constraints on our budget, we were able to keep it going because they supplied the money. So through the Phase 2 proof of concept, they are backing us, and that is what lets this program keep advancing right now.
Nice work on the advancement. Emil and Eric, perhaps going back to a few earlier questions. How convinced are you, relatively speaking, about Bayley-4 cognition compared with MDRI? My impression from this call is that it sounded progressively more favorable toward MDRI than toward Bayley-4 cognition. First, is that a reasonable read? Second, if yes, is it still possible to modify the protocol so that more alpha could be shifted toward MDRI rather than Bayley-4 cognition? Once more, any detail on that would be greatly appreciated. Then perhaps very briefly regarding OI, nothing appeared in the press release or the prepared remarks; so ought we to conclude that this program has been discontinued?
Yes, OI has been discontinued.
Ah, got it — I overlooked that. Therefore, the MDRI represents a stronger metric. Within it sit 5 endpoints, and strength is drawn from each of them. That makes it an extremely potent approach, and we have been advocating for it. It featured in our Mepsevii program. Regulators simply haven't seen much of it yet. Consequently, right now, under the arrangement we reached with the agencies, 80% of the statistical power goes to the Bayley-4, and that is genuinely where power is required since the effect size there is — well, somewhat smaller in magnitude. With the MDRI, statistical significance comes through very strongly. Thus, additional alpha allocated to the MDRI isn't truly necessary, even if your confidence in it runs higher, given the enormous power it already carries. So my view is that the proportion we've settled on is the right one, and shifting more toward it isn't needed. Regarding the OI, evaluations and conversations are ongoing. A bit about this went into the release. Once a decision is made on our path forward, the street will hear from us, but the program has not been discontinued. It is still running at present, and it will remain so until we finish, obtain answers to the critical questions, and reach a determination.
Amit here, filling in for Ram. I've got a question regarding the PDUFA dates that are approaching in the second half of 2026. Could you outline how the commercial rollout will be paced and when we should anticipate significant revenue from UX111 as well as DTX401? Then my second question concerns GTX-102 in Angelman. You brought up discontinuation rates tied to non-safety burden. Are you putting anything in place to lower discontinuation rates once in the commercial setting? Or do you not see that becoming a problem?
Sure. Launching a gene therapy is obviously complicated, including the reimbursement side and all of that. We aren't providing guidance on revenue — what revenue might look like this year — but we are definitely planning to keep moving ahead with pulling the launch together. I'm certain Erik could walk through this in great detail. I'll just give you the summary, Erik, that at launch there will obviously be various policies we'll need to manage. But ultimately, we're speaking with payers and holding the appropriate meetings right now to help lay the groundwork for the planning process around that. Reimbursement, though, will be part of the process, and then we'll need to work through it to get patients treated. So we anticipate that some time will pass between the PDUFA date and being able to get things moving. We're trying to work as hard as we can to get it going within a few weeks and to figure out how we can get it going. We expect product to be available, and it's more a matter of getting the commercial process underway. But I think the team has been working aggressively to get all the necessary pieces in place. And I've been taking part with them in some of the payer meetings that examine what the plan is, what the policies were, and how to navigate the process to get the best outcome for patients. On discontinuation rates, we actually find the discontinuation rate to be very limited. In fact, it's really a small handful; and typically, the treatment effect patients begin seeing is so meaningful that it's something patients have been seeking their — patient families have been seeking their whole lives. So we don't anticipate a big discontinuation rate. But as we move forward with it, we will clearly need to manage accessibility and convenience. And this may involve bringing in a device to help make lumbar punctures easier. But we're going to do our best to make sure this is as burden-free as possible. What I'd tell you is that taking part in a randomized study, or really any clinical study, carries a much heavier load than simply receiving something in clinical practice, given all the obligations involved. So the sheer volume of testing and that sort of thing is considerable. And for a family dealing with Angelman, I suspect that weighed in more heavily. So right now, judging by what we've observed, we don't anticipate that discontinuation will prove to be a major concern. That said, we're going to try our hardest to look after patients ahead of any problem arising and provide the finest patient experience we're able to, in terms of making convenient a therapy that requires intrathecal delivery of an ASO.
Appreciate it. That brings us to the end of today's call. Should anything else come up, feel free to reach out to us by phone or at <<email_hidden>>. Thanks for being with us.
That brings us to the end of today's call. At this point, feel free to drop off the lines. We appreciate you taking part.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Ultragenyx Pharmaceutical Inc. transcript - plus 132,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.
Get an API key View API docs →For developers and AI pipelines
Programmatic access to Ultragenyx Pharmaceutical Inc. earnings transcripts and 132,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $162 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.