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

I hope everyone can hear me. Welcome to the session on challenges and opportunity of rare disease drug development. I'm Yigal Nochomovitz, I'm the biotech analyst at Citi. As always, if you have questions during the session, you're welcome to email me at yigal.nochomovitz@citi.com or showcase feature on the video to assist in answering questions that way. So it's my great pleasure to have with me this afternoon 5 English CEOs of rare disease companies. From Applied Therapeutics, Shoshana Shendelman, CEO; from Dicerna, Doug Fambrough, CEO as well; from Eiger, David Cory, President and CEO; from IMARA, Rahul Ballal, CEO; and from Ultragenyx, Emil Kakkis, President and CEO. So thank you all so much for joining this afternoon.

Maybe as a -- just as a broad high-level question. I'd love to get your thoughts on design considerations for rare disease trials. What should be avoided when designing rare disease trials? What are the pitfalls? And also what aspects of rare disease trial design should be emphasized and encouraged? Anyone happy to kick it off or I can call on people? All right. Well, Rahul, do you want to give it a shot?

Sure. First of all, thanks, Yigal, for having me. It's a pleasure to be here. I think -- as we think about clinical trial design for rare disease indications in ours or sickle cell, beta-thalassemia, a couple of considerations for us, specifically. Number one, it's the balance of open-label style design where we can actually see the types of biomarkers that we're interested in from a surrogate standpoint and potentially that could portend to clinical outcomes and balancing that with randomized control trials that actually allow us to see clinical effect and outcome specific to placebo. And so in sickle cell, where there are more patients in the U.S. that have the disease in other rare indications, I think for us, we try to balance the benefits of open-label design trials that can get us to answer as quickly to see if we actually have a signal and understand that signal with the complement of randomized controlled trials that actually prove that we have the signal and actually design those in tandem, so we can get data readouts at frequent visits. I think some of the challenges in sequential trial design, in many previous examples, is that you have to wait a long time to get answers. And one of the important things from the open-label part of our programs is that you can see data more real-time and that could inform randomized controlled trials, which, by function, had a longer life cycle.

Okay, very good. Doug, do you want to give us your thoughts?

Given the severity and rareness of rare diseases and the flexibility that the FDA often affords to development of rare disease, there are real opportunities to have strategies that cut a lot of time out of clinical development. We, for example, in our primary hyperoxaluria program, purge the FDA at a pre-IND level and proposed a development plan that went directly from a single-dose Phase I into a Phase II/III registration trial and achieved agreement with the FDA on that, proposing a single-dose in the Phase I for modeling and simulation. And then essentially having the first few months of the registration trial serve as a multi-dose safety in the evaluation and continuation with no pause in dosing. That sort of compression of time line saves a tremendous amount of time. And it's not something that is permissible in certainly less severe and more common indications.

Got it. David, do you want to offer your thoughts on how you approach the progeria development? And what you avoided and what you emphasized in the studies?

Yes, sure. Every rare disease, and everyone on this panel knows, is unique. And Hutchinson-Gilford progeria syndrome is certainly no exception. The initial studies with lonafarnib in progeria patients looked at -- these are children who die of advanced cardiovascular disease, usually by 14 years of age. So we were looking at pulse wave velocity and other endpoints, including weight gain, ability to thrive and ultimately generated good data. But I think the comments related to working with the FDA are really important. Our discussions with inborn errors division led us down a path to start analyzing an open-label study of patients to look at survival and ultimately, matching a control group of untreated patients, leading to significant survival benefit in the lonafarnib-treated patients, which has led us now to an NDA filing and a PDUFA date in November. So we're very excited about really, really good collaboration with regulators, which I think is critical given how unique each one of these diseases are and certainly progeria.

Great. Shoshana, do you want to talk a bit about how you've developed the galactosemia program and what you've avoided in terms of trial design and what you've focused on to improve chances of success?

Sure. I think galactosemia, as it's similar for a lot of rare diseases that companies are working on usage, we were the first into the disease. We're the only people working on the disease that we're aware of now and that provides a lot of opportunities for development, but there's also a lot of challenges. And I agree with everyone on the panel that there's an opportunity specifically in rare diseases to approach the agencies, FDA and EMA, early on to talk about the specific rare disease and make sure that your program is really targeting elements that are specific to that disease. In our case, our drug is an inhibitor of an enzyme that converts a substrate or product, and that product is toxic. And it's actually an enzyme that's involved in a few rare diseases. So galactosemia is the first that we're targeting, but there's 2 rare diseases that follow in close concert. And it's not the case in every disease, but when you're targeting an enzyme, it really speaks itself to a clear biomarker-based approach. So the enzyme converts galactose to galactitol, you're blocking that enzyme and so galactitol is a natural biomarker. But setting up the program and ensuring that we had everything aligned with the agency to get to that point, and how we would look at that biomarker-based endpoint, how we would tie that to outcomes in our case in the animal model, really involves a lot of close collaboration and close discussion with the agency. And that does seem to be, I think, a common trend with all rare diseases, that it's best to approach them even pre-IND and talk about the disease on how you'll approach it in the clinic? What your end points look like? So that you're not just entering into a black box of development and not knowing what's coming down the road.

Great. And Emil, you've had decades of experience designing rare disease trials in multiple companies. What are your words of wisdom to those who are trying to develop rare disease products? What should be avoided, what should be emphasized?

I think a lot of the comments people made, I certainly agree with. I think the one thing I would argue for is to develop more dynamic models adapting trials and dosing, particularly. A lot of people are still focused on cohort dosing and fixed doses and the fields depreciate the heterogeneity that exists in all rare disease patient populations. And we generally design programs which have titration -- force-dose titration dosing and allows us to capture more individual dosing strategies and allows you to get to optimal dosing more rapidly. In -- as an example, on our Crysvita pediatric study, we didn't know what the dose should be in that study. We designed a study. We started with different cohorts, but rapidly allowed them to titrate to an optimal. In this case, we had a good biomarker in phosphate. By doing that, we allowed everyone to optimize to their correct dosing within a 3-month period and allowed us to get optimal efficacy. And that design was sufficient to file and get approval for the kids as part of the indication in a single trial, where we didn't even know the dosing we started. We tried to build in this optimization of dosing because it's one of the areas where you can guess wrong and then find yourself with short either in safety and efficacy and have to figure it out on the fly. So whereas the dosing and other things, I strongly recommend people think about adaptiveness in their trial design, so they don't have to make a mistake that could have been avoided and allow themselves to evolve and optimize. And I think, particularly, and dosing is the one I recommend people realize it dosing for every patient. There's not 1 dose for any drug. There's often a range of doses that will have different safety profiles and efficacy profile for different patients.

Great. Next question I wanted to touch on was how you -- the selection process in rare disease. How you decide what rare diseases to pursue? Is there an algorithm that each of you uses to decide? What are the factors that go into the decision? Are there certain rare diseases that just don't make sense because of the biology or the geographic distribution of the patients that doesn't make sense to pursue? Just talk a bit about what are the factors that go into those really important decisions about what rare diseases makes sense to target and develop drugs for. Emil, you want to start?

Sure. Happy to. I think it's a really important part of getting -- having a good success rate is picking well. In general, I try to pick things where there are -- it's a soluble protein or factor that we're dealing with rather than structural proteins or factors. The reason is that for most soluble proteins, whether it's an enzyme or a hormone, you're blocking the potency and ability to make large changes or sufficient changes is much greater than in structural proteins or other things that are much more, say, difficult. So for a lot of enzymes, for example, a few percent of normal for an enzyme is enough to catch up the substrate and get success. Whereas in other disease, where you're got more structural proteins, you have to get 20%, 30% of normal to actually get anywhere and that threshold of effect creates a barrier to success. And so we look for soluble targets and ones that we think were small percent of improvement or increase or decrease, will have a big effect on the disease. And I think that's the best way to get a higher probability of success.

Interesting. Doug, what are your thoughts on the selection process?

I think for Dicerna, many of the criteria we apply are somewhat specific to what specifically we do, RNA interference. We have a technology that turns genes off. And for its first generation, it turns genes off specifically in hepatocytes. So that localizes you in one particular tissue where you could treat hepatocytes and provide benefit. And also limits you to a subset of rare diseases where you can actually provide benefit by turning things off. As I think everyone's aware, many of these rare diseases are loss of function. And sometimes there's a compensatory mutation silencing that can occur. But often, those are only treatable by adding something back, not by silencing. But there are specific cases where all this sort of Venn diagram of opportunities overlaps, what we can do in the tissue that we can do it in and the biology of the disease. Now within that much more refined set, we really want well-understood biology and a biomarker that we can track. Well-understood biology is probability of clinical success factor and the biomarkers practical operational factor. When all those things are going for you, really appears to be a high confidence, straightforward development plan that you can execute on. We can -- we're a small company, we can only do a few things. And so as long as we have a few of those, we have, I think, a very good selection to work with.

Got it. Shoshana, do you want to talk a bit about how you decided on galactosemia?

Sure. It's sort of a merge of, I think, what we've talked about so far. So I agree that valuable proteins where you understand the mechanism of disease really well, and you have confidence that in targeting that specific protein or enzyme or whatever your focus is that you can have an impact on the biomarker that you're looking at and that, that will also correlate to disease outcomes. And so all of those pieces fitting together really is what drove us towards galactosemia as a starting point. It's very clear that loss of a certain enzyme causes galactosemia and that this substrate is then converted to a toxic byproduct, in our case, galactitol. And we had a lot of confidence that this mechanism of disease was valid. I think that having animal models and -- in not all cases, animal models are possible. So sometimes you have to resort to things like cell-based models. But for us with galactosemia, having a really good animal model of disease that validated the mechanistic target and allowed us to get outcomes in the animal in a way that you don't always have possibility to do in clinical studies in a short concept period of time, was really important to us. And so being able to look at the animal model of galactosemia, which have the biochemical characteristics of human disease and the same clinical outcomes, things like cognitive insufficiency, developmental delay by motor cells, those are things that are really difficult to look at in humans in a short period of time and really require longer studies. And so having a good model that drives the biomarker to the outcomes and lead you to a higher probability of success when you get into the clinic, I think, is a really important gating event because you want to set yourself up for the highest probability of success when you get to the clinic. And then just the last thing I'll say, I know it's a little long-winded, is having a PD biomarker. I think, is so critical in almost all rare diseases because it allows you to select the right dose based on that PD biomarker in a lot of cases. And then know that, that biomarker is tied to the outcomes and be able to look at the percentage change that you have in humans and sort of know that, that will give you an effect on long-term outcomes over time. So for us in galactosemia, it was as tied together, all of those things: good mechanism of disease, soluble protein, good animal model that tied biochemistry to long-term outcomes. And all of those things lead you into the clinic in the sort of the best position that you can going into a rare disease.

Great. Rahul, did you want to offer some thoughts on this question regarding selection of diseases?

No, I think my colleagues summarized it well. I'll just add the concept for us in fetal hemoglobin induction is a well-validated biomarker that is approachable through small molecule, which is kind of one of our areas of expertise in finding ways to treat patients in a way that actually enables them to take the drug for a long time in sickle cell disease. Specifically, in beta-thalassemia, an oral once-a-day drug, is transformative for a number of reasons. And as we think about the calculus that gets us to the disease choice, these particular indications lend themselves to simpler format drugs that don't need long-term cold-chain transport or don't need to give refrigerated drugs that could be given easily in a number of different environments, both in the U.S. but also across the world. So if you think about the calculus beyond the biology, I would just add that we think about the actual delivery of the drug and the actual attractability that patients would need when they take it.

Great, David, did you want to add anything related to progeria, and how you came to select that indication?

Sure. For Eiger, progeria was serendipitous. We founded Eiger to develop treatments for hepatitis delta virus. And it was based on a publication in science in 1992, that showed that there was a potential druggable target on the hepatitis delta virus. Back then, delta was not well-known endemic throughout the Middle East and throughout Asia. Now from migration, delta is a rather large orphan disease in the U.S. and Europe. But back then, we didn't have a drug. Prenylation was the target on the HDV genome that we believe that we could target. Prenylation inhibitors were in development at the time for Ras prenylation in hopes of curing some forms of solid tumors. It wasn't until 2009 that we were able to actually access a drug called lonafarnib that had been well characterized, tested in multiple oncology trials. And we were able to take lonafarnib as an oral prenylation inhibitor straight into Phase II to prove concept in HDV. In that process, we dosed over 170 patients and have now, thankfully, advanced lonafarnib as the first oral agent into Phase III for hepatitis delta virus infection. The really cool twist to the story is, during this process, we were introduced to the Progeria Research Foundation, Leslie Gordon and Francis Collins at the NIH, identified the gene responsible for progeria. And identified progerin as an aberrant protein that was a prenylated protein that integrated into cell -- membranes of cells in these children and ultimately led to premature aging and advanced cardiovascular disease. We completed multiple studies. And as I mentioned earlier, demonstrated a survival benefit by blocking prenylation in children with progeria, and are now on a pathway towards our first NDA approval for this patient population that otherwise would be looking at a life expectancy of 14 years. Many of these children are now young adults. On lonafarnib over 10 years, into their late teens and early 20s. And so we're extremely thankful and excited for what the future holds.

Okay. Great. I want to shift gears a bit and talk about 2 features of FDA interaction that I think are relevant for the rare disease pipelines. And that is breakthrough therapy as well as the priority review vouchers. So to the extent that you can comment, I'd be curious to learn how significant is break -- getting breakthrough with the FDA in terms of your dialogue and progress with regulatory milestones. And then with priority review vouchers, how much does that really matter? How much does that go into your calculus in terms of the development strategy? So Emil, you want to kick that one off?

Well, we all hope we're going to get breakthrough therapy. So it's still hard to plan for it. But okay, we -- I think breakthrough therapy is I think an opportunity to accelerate the discussion with the FDA and the commitment from them to show the flexibility required to get to the end and do it promptly. In our case, with Crysvita, getting breakthrough allowed us to press ahead with filing off of Phase II peds data rather than waiting for Phase III peds, which allowed us to file a year earlier. And I think having breakthrough designation is certainly a part of that calculus that happens with the agency about accepting the importance of the therapy and moving ahead. The PRV program, I think, is something that, first, getting it reauthorized has been a challenge, but it's certainly, for us, we've had 2 PRVs, one we fully owned and one we shared. And those 2 generate a substantial amount of money for us, around $170 million, which is a significant capital infusion for simply doing the work we were going to do anyway, I think, in terms of development. But it does have an impact on some of the ultra-rare indications and one like Mepsevii, for which we got a voucher. Essentially, the voucher covered the entire development program and then some -- and does allow sometimes those ultra rares to maybe make sense if you can get them through promptly in a capital-efficient way. So I do think there is a real value there. And I think if it's not a long-term program, though, it's going to be hard to have impact on what people choose. But over time, I think it's infused a lot of additional capital in the earlier stage companies. So I think it's a benefit. So both have been a helpful program for the rare disease community.

Doug, I know, I think you have some thoughts on breakthrough, love to hear that.

Breakthrough has been really beneficial for us. The drug that we're developing, Nedosiran applies to 3 different diseases. Now they were all primary hyperoxaluria's types 1, types 2 and types 3 and they had some related biology. But there's distinct diseases. And they're kind of a decreasing order of when they were discovered and [indiscernible] natural history basis. And so while the initial indication PH type 1, it's fairly clear what the FDA wanted to see with PH type 2 and PH type 3. It's really required a dialogue with the FDA to reach common ground. We're still in process on PH3 and having the frequent communication that BTD has enabled -- has really facilitated those interactions with the FDA. So I think we're in a far better place with BTD than we would have been without the ability to have such a regular interaction with the FDA. The PRV has little bit of a different story. When we first started the program, we didn't understand we'd be eligible for PRV. There's been some adjustment to how eligibility criteria are interpreted. It'll be nice if we get one, and it will be a nice infusion of capital. But it's kind of a nice to have and wasn't part of the planning initially.

All right. Very good. Anyone else want to weigh in on those 2 topics, breakthrough and PRVs.

I'd just say I kind of concur that the BTD designation is an important aspect of development for us. And we certainly think as we pursue fetal hemoglobin as a surrogate for accelerated approval that, that would be a helpful path for us to take as we generate the data to support it.

Okay, good. And David, did you have any comments related to the PRV or breakthrough with respect to progeria?