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

Good afternoon, everyone. My name is Gena Wang. I'm SMID cap biotech analyst at the Barclays. It is my great pleasure to introduce our next presenting company of Ultragenyx Pharmaceuticals. With me on the stage, we have Dr. Eric Crombez, Executive VP and Chief Medical Officer. Thank you, Eric, for coming to our conference.

So maybe before we dive into the specific questions, maybe very high level overview of Ultragenyx.

Yes. Great. Thank you. So Ultragenyx, we are founded and have stayed very much true to our mission of focusing on rare diseases with high unmet medical need. We do have 4 commercialized products led by our product, Crysvita and a very robust pipeline, which I'm sure we'll dive into today, led by our drug setrusumab for the treatment of osteogenesis imperfecta. We also have some late-stage gene therapy programs looking forward to our first gene therapy approval for Sanfilippo quickly followed on with our filing of GSDIa. We do also have earlier-stage programs for OTC and Wilson and then very importantly, our ASO for Angelman syndrome.

Great. So I will dive into the setrusumab in osteogenesis imperfecta, that's I think the investor key focus in the past months or years. So now after first interim, now everyone is waiting for the second interim so if I did my math correct, I think based on the clinicaltrials.gov, the last patient completed enrollment is May 1, to update at least in the clinicaltrials.gov. So I assume sometime in the late April or early May, the complete dosing. So 12 months follow-up that will be the second interim. So if I carefully correct, sometime around May, that should have at least the DSMB should know the second interim, whether it hits or not. So maybe walk us through the steps, how would they -- how soon would take DSMB to communicate with you and if it's negative or positive, how -- what will be the next steps you would take?

Yes, great. So just for a little bit of context. So we are heading into our second area analysis for our lead Phase III program for setrusumab for the treatment of osteogenesis imperfecta and that is the Orbit study originally designed to be a Phase II/III study. We also have our second study, COSMIC. COSMIC is important because it goes down to patients down to 2 years of age and is a head-to-head study to bisphosphonates. So with IA2, we are looking to do interim analysis on both Orbit and COSMIC in parallel. We will stagger those DMC meetings. So we will have that data readout for Orbit with 159 patients enrolled with all patients to at least the 12-month time point. So that will go to the DMC first. Certainly, it takes time to bring all of this data into the database. We need all of this data cleaned and finalized to prepare that data package for the DMC, they need to do their job. If positive, the top line data then comes to us, we have a very, very small group of people who review that data. And then we need to do our review, our quality checks and make sure we're confident before we make a public disclosure there. So really, that is your period of time between last patient at 12 months versus data readout. Importantly, that's led by Orbit, again, we will bring that interim analysis for Cosmic along in parallel. In the event Orbit is not successful, and we are not notified by the DMC that we have hit success criteria. We actually will hold that data readout for Cosmic. So we don't end up spending that alpha there. But that is the process for IA2.

Okay. So maybe the several questions here. First, like I know the senior leadership do not see the blinded events, right? Does any clinical team within the company see the blinded events.

No. So it's really been an important priority for both of these interim analysis that we protect the blind for this study. We protect data integrity really so that ultimately, if any, regulatory agency body, the FDA comes and does an inspection, there's no concern for a bias. They introduced bias into the study. So we have maintained a very hands-off approach there. The DMC really is our independent body there, and we have another external vendor doing all of the data preparation. They prepare that data package for a DMC review. And it's only when and if IA2 is successful that we would receive those top line TLS. So we have remained blinded to that data. It is important to remind ourselves that the DMC does have a responsibility to monitor safety. We have a responsibility. So certainly, if there's any change in the safety profile or a new unexpected safety event that would also be communicated to us.

Okay. And then is it fair to say if the second interim did not hit, you would know the results quicker since you do not need to do additional analysis. And will you share with like how long it would take for you to share with the investors.

Yes. So I think it's somewhat of a similar process. Yes, if it's not successful, we would need to do -- we would not be doing our own validation and quality checks, if you will. But I think fair to say a very similar time line in that case to what we had with IA1.

Okay. So basically, a few weeks. Right. Okay, good. And then if hit, of course, you will not announce anything, but will you share both Cosmic and Orbit together with us?

Yes. So I think, again, we will be looking at top line data first. And I think, yes, there would be some level of disclosure there. Our attention with a positive readout for IA2 to quickly shift and really immediately shift to the data package for filing there. So we will really have the team head down working on that, and then we would need to decide when would be the appropriate time to have really much more of a more thorough data release there. But certainly, we'd be initially focused on top line data, which for OI, it is fracture data supported by DMD. So I think with top line data, you have a very good sense of what's going on with this drug.

Okay. So I just want to make sure I understand. So that if it's positive, you only share initially top line Orbit data and then do you need to do some -- like usually, how long does it take to at least clean up some of the data and you'll be able to share the top line, like you hit the stats or a p-value or the magnitude...

Yes. I think we're talking weeks. It's certainly not months plural. I think it's more on the scale of weeks, certainly not months. But it's not -- at the same time, it's not days.

Okay. Okay. So basically a similar time line. .

Yes.

Okay. So now going back to your stats assumption. I think if I look at the earlier statements, the baseline AFR is about 0.72 and you power 80% to detect 50% benefit. And by the way, I did have some investor pushback or why you're powering like compared to your Angelman program, your powering assumption is much higher 90% and the magnitude benefit also is higher here why the powering was a little bit weaker compared to the Angelman. So there was some pushback. Would that be -- for me, the first question is, so far, patient, do you see, at least from the baseline perspective, is that consistent with your powering assumption? And then what makes you feel confident this powering assumption will be sufficient. You'll be able to -- with current study that will be sufficiently powered.

Yes. So I certainly wouldn't describe the power as weaker compared to Angelman. I think there's a big contrast with something like osteogenesis imperfecta and what we're seeing with setrusumab because fractures is a very straightforward clinical endpoint. We all understand fractures, and that's really supported by bone mineral density. So we don't have the type of variability with something like osteogenesis imperfecta like you do with any neurologic indication. So yes, with contrast, with Angelman, you could argue we overpowered that study, but we wanted to allow for some natural variability you will see with patients there. So to me, that's the real difference there. It really is a principle we have going back to osteogenesis imperfecta, where we do not bring untested things into a Phase III trial. So the Phase III part of Orbit really does mirror what we did with the Phase II part of Orbit same entry criteria, same overall study design, same endpoint. So there's nothing that we did with the Phase III design and conduct that would tell me there's something different compared to what we did with Phase II. So really strong and powerful data set from what we saw in Phase II, I think will hold true in pull through. Again, you do want to allow yourself some room for some level of variability for any reason, again, we saw that 67% reduction in fracture rate in the Phase II part of this were powered to see a 50% treatment effect. So that does give you some room to accommodate variability there. So that greater than 80% power at 18 months to detect a 50% treatment effect based on the Phase II data is a very, very well-powered study. And again, we did enroll 159 patients with a 2:1 randomization. So we do have quite a large number of patients on setrusumab, especially in the context of a rare disease trial.

Regarding the 67% reduction, right, that's the median and then now the study is a mean. What was the mean reduction for that Phase II study?

So we haven't disclosed that. We certainly know that we have that, and I can reassure everyone here there's no meaningful difference between those 2 values. And looking at the 2, it would not have changed anything for the Phase III powering.

Okay. Good. And I do have investors basically proposing why giving yourself unnecessary volatility putting first interim, second interim, why not just go through the final, given you have a very high probability hitting the stats for the final. So maybe thoughts there.

Yes, that's really twofold. There's a clinical reason and then there's really a business reason there. And clinically, these children and young adults in the control groups are continuing to fracture. And we know that based on the annualized fracture rate, these patients had coming in at baseline. So we thought with, again, the strength of the Phase II data, if we could have a positive interim analysis, we would unblind these -- we would unblind Orbit, we would get all patients on to setrusumab and hopefully stop this accumulation of unnecessary fractures. On the business side, with 60,000 patients in the territories we cover, there's a lot of value to bringing these launches in and getting into the commercialized setting as quickly as possible. There's a lot of value there.

Okay. And then another question is that if you miss second interim and what will make you feel confident and you will hit a final and then the magnitude could be more than 50%.

Yes. So again, I mean, to me, and again, it's very clear and very powerful signals we saw out of the Phase II. And again, with the design of these studies, I don't expect to see anything different. We didn't introduce anything new, different or variable there. So again, if the data from Phase II holds up, we will have a successful Phase III.

Do you know the patient on baseline bisphosphonate, any rough idea what percentage of patients were on baseline bisphosphonate they had to win off.

Yes. So it is the great, great majority of patients in Phase II and in both Phase IIIs came on to study with previous treatment with bisphosphonates and really for quite a long duration, we weren't really seeing patients who had just had a few doses of bisphosphonates and came into trial. So really, that annualized fracture rate we saw coming into study that baseline rate really was on the background of treatment of bisphosphonates.

So was that the majority of the patients were you talking about like 70% patient or...

Greater than that.

Okay. So like 80%, 90%.

Yes, to be honest with you, I don't have the exact number top of head, but it was the great majority of...

I see. Okay. Okay. I think that's helpful to minimize the placebo noise there and also the fracture rate could increase for the placebo, right? Okay. Great. And Cosmic study, what is the assumption for the bisphosphonate arm the fracture rate?

Yes. So again, when we were powering the study and really looking to model out that treatment effect, with these patients coming in already on bisphosphonate, really the 0.72:1 annualized fracture rate we were modeling on really was on the background of bisphosphonate. So that's really the range we were looking at for both of our control groups for both Orbit and Cosmic.

Okay. Very helpful. So maybe we switch gear on the Angelman program. So maybe I started with the primary endpoint. In the past, we saw different endpoint, right? I think for the first time, we saw the Bayley-4 full cognition raw score in the past always the GSV score. So normalized the scores. So maybe tell us the differences between these 2? And then what are the total raw score for the cognitive measurement?

Yes. So again, the Bayley was designed for neurotypical children. It's how clinically all children are just measured to make sure there's no development delays, no concerns for issues with development. And it was really the FDA who brought forward the idea of using raw score. It's their preference just to use total raw scale as accumulated through this assessment versus using the GSV, which is controlled for age and other factors there. Just looking at cognitive raw scores by Bayley-4, which is our primary endpoint for our Aspire study. There's 81 questions there. Each one is scored 0 to 2, 0 means the scale is not present. One means it's emerging and 2 means it's been mastered and that's how that raw score is measured.

So if I do a quick calculation, the total store could be 162, correct. Okay. And then when we look at the differences, you did show the natural history study. I think it was natural history 2. When I look at the patient population, actually relatively small so maybe like I think in natural history to only 16 patients, we did see the raw scores of 1.2 and a standard deviation 9.5. So how confident you are with now the enrollment with the Phase III studies, you think that, that could be replicated?

Yes. No. And first of all, it's really great to have a natural history data set to pull from here because we always talk about the importance of natural history with rare disease. But oftentimes, these natural history data sets aren't well collected. They're not monitored. It's not as strong as you want it to be. The data we're pulling from is from the LADDER study is a very well designed -- there's a lot of integrity to that data. So it is nice to have that resource available to us. and they have both data on Bayley-3 and Bayley-4. But really -- and we've shown this in previous presentations, we think of neurotypical children on their developmental curve, which means they're growing learning and gaining new skills. The Angelman natural history data is it's really a flat line. These children at a certain part really are not growing and developing and learning new skills. And when we most recently presented data at the patient organization, meaning the FAST meeting we actually tripled that effect size of the natural history study, and we're still by doing that time 3 with the data we saw in Phase II, we still have a very well powered, and that's the data we use to power our study. And again, with 120 patients in that Phase III, we have a very well-powered study, not just for cognition, but also for the other important domains, language, behavior, sleep and gross motor. And we are bringing forward a key secondary endpoint of MDRI bringing forward all of those assessments into a single test with a very straightforward scoring. It's a key secondary endpoint, but part -- statistically part of the primary analysis that we are applying [ alpha ] to that 0.05. So we have cognition raw score supported by the MDRI, both with Alpha applied.

I'll be good. So when we look back the GSV scores, when you're including Cohort 4 to 7 and A and B, we did see like there's a meaningful decline regarding the pool data, 6.7. We went back Cohort 4 to 7, that was 9.5 score roughly. And so which means A and B may be only 5-point something scores maybe what could be the reason leading to that differences regarding the cohort A and B.

And when you talk about decline, you're talking about a difference between those groups not patients are declining. .

Right.

So our numbered cohorts were -- our original cohorts, those were dose finding and then our lettered cohorts. We thought it was important to do our dose expansion cohorts. That was another 53 patients that we brought forward to make sure we have the correct dose to take into Phase III. But while again, for a rare disease Phase II trial, that's really a number of patients, in order for you to make decisions to go into Phase III, they're still relatively small numbers so to see a difference between those 2 groups doesn't concern me. There is variability with Angelman patients with any neurologic indication, Again, that is why we do like tools like the MDRI to bring all of these domains into a single assessment. So yes, there was a difference there. But to me, that's just due to small numbers and just really some natural variability with Angelman patients. Again, that's why it was important for us to bring this Phase III study forward with 120 patients to allow for any variability that may be out there.

Okay. Very good. And then for the enrollment, you already dosed the first patient last December. Any thoughts how is the enrollment ongoing? And do you try to involve as many as possible before Ionis start the Angelman.

Yes. No, certainly, it's an important program to us, and we mean that we really have staff to the team with the right people and the amount of people to do this as quickly as possible. So yes, we did announce that we started enrollment at the end of last year. we've said we'll complete enrollment in the second half of this year, but we really are prioritizing this program, and we do want to get enrollment on it as quickly as possible. While Ionis is working through study start-up, and before they get into the clinic in dosing, yes, we'll take advantage of that period and we will enroll as quickly as we can.

Okay. Good. What is your internal plan, like estimate regarding the timing to complete enrollment?

Yes. So again, we're trying to go as quickly as possible. It is 120 patients is intrathecal dosing under sedation. So to me, this is not a patient fine question. Really, there is a lot of demand. I mean it really for a neurologic indication, most of the time, you're really looking to stop further loss and just sustain what patients have. It's very rare when you actually have a type of treatment that allows patients to continue to grow and develop and learn new skills. So there's a lot of demand out there from these families. But we need to make sure we have the right sites who are able to bring these patients in with enough staff and enough room in their dosing suites to do this and do this on an ongoing basis with that many patients in a relatively short period of time.

Any quick update regarding the oral Phase II/III study?

Yes. So Aurora is our second Phase III study. So again, with the idea of not bringing on tested things into Phase III, our Phase III will continue to enroll patients with full deletions. 4 to 17 years of age, our Aurora study will dose younger patients, older patients and patients with other genetic changes such as Uniparental disomy or missense mutation. So to me, Aurora is important because that will bring this treatment available to all patients and not needing to do a stage effect of having this available first with patients with full deletions. And then and then later to the rest of that community. The greatest majority of patients is with full dilution, but we do think it's important to bring this forward for as many patients as possible.

Okay. Good. We have less than 2 minutes maybe quickly high-level update regarding the MPS IIIA, GSDIa and Wilson.

Yes. So MPS III Sanfilippo, we've announced that we received a PDUFA date of August 18. So we've set up our launch readiness team. We are preparing for approval in August. We're excited about that. I think that's a great way to really start off the kind of the evolution of a developmentally-focused gene therapy pipeline to a commercially available gene therapy pipeline. To me, gene therapy launches and programs that have been successful are those of the 2 unmet medical need. And with a disease like Sanfilippo where these children are born, grow and develop for a number of years, plateau and then start to lose ground with absolutely nothing available to them now, there's a very high unmet medical need. So we do expect quite a bit of demand at launch and then that will hopefully support our next launch for GSDIa, glycogen storage disease type 1a, we've said we will file with the FDA midyear and then we'll look forward to that approval as well.

Wilson.

Yes. So Wilsons, also in the mix as long as well as OTC, which is in a traditional Phase III program. Wilson, we went through the data set from our original plan cohorts. We did make the decision to go up to a higher dose, still within the E13 range, which we think for a liver-directed gene therapy is still a safe range. And then bringing forward better immune modulation. Certainly, steroids does help but we think with the amount of B-cell and T-cell depleting drugs out there, really trying to get that immune response under control, Will, in combination with a higher dose, have the greatest impact because we do want to get the great majority of patients off of standard of care, off a chelation before we go into the Phase III part of that study.

Great. Well, thank you very much. We look forward to the data update later this year.

Great. Thank you.