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

Welcome and good afternoon. Thanks, everybody, once again for joining us at the 42nd Annual Cowen Healthcare Conference. I'm Yaron Werber, one of the biotech analyst at Cowen. And it's a great pleasure to moderate the orphan CNS therapies panel with my colleagues, Brendan Smith and Lyla Youssef. We really have a terrific lineup today and a lot to talk about. We have from Annexon, Doug Love, who's the CEO; Geoff MacKay, who's the CEO of AVROBIO; we have Bobby Gaspar, who's the CEO of Orchard; Bruce Goldsmith, who's the CEO of Passage Bio; Emil Kakkis, who's the President and CEO of Ultragenyx; and Arnon Rosenthal, who is the CEO of Alector is going to be joining momentarily. So welcome, everybody, and thanks for joining us. We appreciate it. So let's start. What we're going to do is we're going to talk first about a couple of maybe very broad drug development topics in orphan neurology, and then we'll do a round robin and dive into each company specifically. Most of you are developing drugs essentially for unmet medical needs. In some cases, some drugs are approved, more in the [ less it's almost ] storage area. Obviously, in Bobby's case, he does actually a drug approved now for MLD from his company. But in most cases, essentially, it's an unmet area with nothing approved. That presents some notable challenges about how to design endpoints, design clinical programs and obviously interface with regulatory authorities.

So maybe we'll go a little bit, Doug, maybe just starting with you at Annexon. In your experience, in your interaction with regulatory authorities and working in your area, what has been some of the challenges and solutions?

We have found the regulators to be really quite helpful in our approach. Our 2 lead indications are in Huntington's Disease and ALS. And fortunately, the approvable clinical endpoints are very well elucidated. But how do you get there, the road is a little bit, as you know, [indiscernible]. And so we've had multiple interactions with them on endpoints that in the earlier studies may be informative towards evidencing disease -- impact on the disease process. And then ultimately, what we would need to show in a definitive study. And so we found them to be really, frankly, quite creative and partnering with us on this approach. And as the field continues to progress, I think kind of the holy grail with the agency is whether or not they begin to allow biomarkers as a key measurement for outcomes in the pivotal studies, something very much like what you're seeing on the neurodegenerative side, for example, in geographic atrophy, a neurodegenerative disease of the eye.

Right. And Arnon, welcome. You're -- we're -- just starting with the first question talking about kind of broadly as you're developing novel therapies. I mean in your case, it's FTD. Obviously, you're also targeting Alzheimer's, which is obviously a much more validated sort of area. But certainly, on the FTD side, as you're interfacing with regulatory authorities, are the FDA EMA aligned [ and the endpoints in trial design ]?

Yes, they are aligned on the cognitive endpoint, like the traditional CDR. For FTD, the cognitive endpoint is the same as for Alzheimer's disease with 2 additional modality for behavioral and speech. So yes, so the FDA agreed with us that sort of these clinical endpoints will be approvable. So yes, so we have agreement on the cognitive endpoints. But as everyone knows, from Alzheimer's disease cognitive end points are still really noisy and require large patient population and longer duration. And again, sort of relying more on biomarkers or hopefully in digital health, like more frequent sort of clinical end point would be really beneficial. But yes, the FDA agree with the traditional endpoints that we are pursuing.

And maybe, Geoff, to you, what have been some of the challenges for you? I mean you're obviously looking at your one of the pioneers in lysosomal based therapies or using in vivo calls for LSDs. What have been some of the challenges and opportunities?

Yes. Well, I think, as you said, we focus on lysosomal disorders. And a common theme is that there are disease manifestations, both above and below the neck, -- and a lot of these -- well, I would say all of these diseases, the drugs that tend to get approved focused on below-the-neck manifestations, of course, because some of the standard of care today doesn't cross the blood-brain barrier. And so that, in some ways, simplifies our primary endpoints because if I think of something like Gaucher disease, for example, we'd be looking at some of the very same primary endpoints and organ size, liver spleen, bone platelet hemoglobin. But of course, what really excites and interests us is the ability to affect the neurological manifestations. And so for that, we have to pioneer some condition and cognition endpoints. And we're not at the point yet where we have alignment with regulators on what that is. We have a lot of meetings scheduled with European and U. S. regulators this year for a few of our indications. So I hope by next year, when we initiate our registration trials, we'll have that worked out. But of course, the challenge with cognition also is that the duration of follow-up will be a little bit longer. So our working assumption is that there will be secondary endpoints, not primary.

And in your case, specifically, is it -- I mean, I guess it depends on which indication when you're talking about, are there no endpoints that have been employed for cognition? Or is it just a challenge of employing that to a new indication maybe without enough previous natural history?

Yes, there is research. But of course, we're -- we kind of view ourselves as the next step beyond enzyme replacement therapy and the enzyme replacement therapy indications really did not deploy CNS measures as primary. So in terms of recognized accepted, validated endpoints, there's really a sparsity of it, which really puts the onus on us to propose them and get them accepted by regulators.

And just which indications specifically were you referring to?

Well, so what we focus on, we're in the clinic right now with Gaucher disease type 1 and 3. And so those -- there are approved products for Gaucher disease. Cystinosis, there's approved products and also both infantile and late-onset Pompe disease and Hunter disease. So the common thread there is that in each of those, there are CNS manifestations, but they're not always primary and they didn't serve as the basis for approvals.

Yes. Okay. That's a perfect segue maybe Bobby into you. I mean it's completely sort of leads right into what you're doing in MLD and obviously [ on the some ] storage front. How is your interaction been? I mean, you're in many ways a pioneer, we just had Libmeldy approved in the U.K. and in Europe, and now you've been dialogue with FDA.

Yes, it's been a very, very interesting time. So thank you, everyone, for bringing that up. I mean, just to recap, I mean we've seen some extraordinary outcomes in these patients with a very severe in a monogenic neurodegenerative disorder, [ MLD ]. The duration of effect is now over 10 years in the first patients that were treated. And remember, there's no standard of care for this disease. It's really just supportive and in many cases, just palliative care. And so the data package that resulted in the approval is, we think, is extremely strong. And so with that, we've been, over the last year since we got the approval working with the reimbursement agencies. And it's really about impressing upon them and this has been a series of interactions with the reimbursement agencies impressing upon them and getting them to understand the severity of MLD, the lack of any current treatments and the effect that Libmeldy has and the durability of that response. So I think -- I was going to say lucky, but I think we are fortunate to have a very strong clinical data package that we can present to the reimbursement agencies. And as a result of those conversations, I think we've had some very, very positive feedback. We had really a historic agreement with the NHS that was announced recently in the U.K. for pricing in the U.K. for all eligible MLD patients were very close to something in another country in Europe, and we have positive discussions with other reimbursement agencies around Europe as well. So for example, in Germany, the GBA, the HCA authority, they gave us the highest rating, 1 of only 5 or 6, I think, is of 500 medicines that have been approved, which again recognizes the tremendous clinical impact that Libmeldy can have. So I think with the appropriate data package that serves such a high unmet need, it is being recognized by the reimbursement agencies. And I think that's -- those are some very valuable learnings for us as we advance some other lysosomal storage diseases in the future, but hopefully, for the rest of the community as well.

And what about the discussions with FDA because FDA has definitely been a little bit more challenging than EMA?

No, without a doubt. And I think I've talked about this previously, but the conversations that we had with the EMA were more advanced than the conversations with the FDA. And the FDA was kind of playing a little bit of catch-up initially. And so we submit an IND, we got RMAT designation, and then we were able to talk to them again about the clinical package and what we would need to do. And I think the -- and again, in a similar vein by showing them the clinical data and the difference, the therapy makes above the natural history, I think that was the key to really turning them around and then -- and to then have a conversation with them that now gives us clarity on the path to file for MLD. So we've been able to clarify what kind of data package that they want, the information that they need to see on the natural history data. We've had a CMC conversation with them. And so I think we've had -- got much more clarity now as a result of those interactions and what we need to present to them in order to pass. So we're -- and then we've guided to a late 2022, early 2023 filed with the FDA. But it really has been a combination, I think, of having those interactions plus some very strong clinical data that we're able to present to.

Bruce, maybe over to you. I mean you're developing areas now that are, again, completely unmet medical need, GM1, Krabbe, certainly, FTD, obviously, nothing really approved, but several therapies in development. Can you discuss a little bit how the interaction has been with FDA so far and your flexibility and being pragmatic to help you design your studies. And obviously, you've had a clinical hold that was unexpected to sort through because of a device that was unexpected. So maybe give us a little bit of a view in your interactions.

Yes. Maybe I'll start with the clinical hold since that's resulted in a very easy one to a really interesting interaction because as many of you know, we're pursuing [indiscernible] delivery. And the fascinating part was that we were actually put on hold for our initial IND for GM1, because while there were needles that were approved for CNS contract, there's been a much more, I think, diligent or aggressive, depending on your stance, FDA oversight of devices that are being put forward for various delivery. And so the syringe and the tubing that connects, which is the device had never been approved for CSF contact. And so it was a little bit of an uncharted territory of retrospectively looking at DNS devices either implantable or temporary touching CSF. And going back essentially looking through the GLP studies, doing some supplemental analysis of the tubing and [ syringe ] and then showing the FDA that there was no associated cytotoxicity or neurotoxicity and led to the lifting of the hold of GM1. We then filed we got rapidly approved for the INDs for Krabbe and FTD. But I think it does go to -- and we had these discussions across a couple of different companies. It goes to the scrutiny of any drug device combination even in the clinical studies and the ramp-up there, at least that we experienced. And I think it's just a general concern of any kind of neurotoxicity that a device may impart in this case, obviously, nothing was found. That led to -- and using neuron as a proxy GM1, I think what we're facing from my perspective in the field is that certainly, traditional looking at biomarkers and looking at changes of readouts, they're in enzyme levels for secondary biochemical pathways in GM1's cases, gangliosides and putting those together to say there is likely to be a benefit could have traditionally, I think, led to accelerated approval. And certainly, there's a rationale for that in many diseases because of the extreme high unmet medical need. But in our discussions, we also felt even in the Phase I in dose escalation that it was important to include, for example, GM1, both the Vineland and the Bayley scale, which are multifaceted scales of -- had a developmental progressions and milestones. Vineland is caregivers assessment, Bayley is a trained healthcare provider assessment. And this is where I think it's important in the FDA conversations to say that if things move in the same direction of enzyme replacement in monogenic diseases we're trying to treat, then going to biochemical pathways and then clinical outcomes. So that to Bobby's point out, linking those together and really seeing clinical benefit or clinical development impact will ultimately be very critical for FDA interactions. If everything lines up in areas of high met medical need there probably is an opportunity for biomarkers, but I think the FDA is moving towards, in general, looking at through clinical outcomes and clinical benefit. In GM1 and Krabbe in a monogenic ultra-rare places almost storage diseases that many of us are trying to address. There's a certain advantage to this because the clinical milestones are so rapidly deteriorating, that showing clinical benefit, not just due the stabilization, but actual benefit is very critical. When you get into a more complex disease and reference this for [ CD or for ] FTD, it's just a more complex endpoint that takes longer to develop. So it's -- whether it's cognitive endpoint, whether it's [ tinted ] endpoints, those are just materially different, I think, in terms of looking at the magnitude of effect and longevity of following the patients.

Yes. And Bruce, for things like Krabbe, which is extremely rare, or GM1, which is still fairly rare, do you need a placebo? Do you think you're going to need a placebo control or natural history is good enough?

So it's a great...

[indiscernible] opinion, I know [ I'm also going to have ] an opinion.

I think we're hopefully aligned. I don't think it's either feasible or ethical to do a controlled study. Then -- but then it does raise the question of, do you have sufficient natural history to show that there is an effect, which does become an -- so probably there's published literature, et cetera. As you get into -- which -- and I think probably it might be a little bit larger in terms of known population because there's newborn screening. And there is a de facto care that's used, which is sensible transplant. So there is much more visibility and awareness of tracking. GM1, I think the scarcity of that is not just necessarily inherent epidemiology, but also newborn screening. And it just becomes an issue, but we and others are [ just ] being that natural history now along the way. So I'll leave it to [ Bobby and Emil ] to comment if they agree with me, but hopefully, we're all aligned.

Emil, let me ask you, but it sounds like you guys had an agreement. Let me ask you, I mean, you're moving now into Angelman. Ovid obviously has pioneered the CGI, but there is no validated endpoint as of now in Angelman. And you've had a very interesting interaction because they really started with genetics, which is really the foundation, which is a very unique situation going to the FDA to start a study. And then obviously, Ultragenyx came into the picture. How is your FDA interactions been so far in terms of designing the endpoint in the study?

Well, I think it's been an interesting pattern. I think they were very supportive in the early days and getting things going. And then we had a safety issue, which created a problem in a whole, which we've overcome eventually. The endpoint discussion in Phase I/II is not nearly as difficult as it is when you're [ targeting straight ]. So it's actually pretty easy at the IND stage. So you can assess a bunch of things and look at them, and there was really not a lot of debate about the endpoints. The truth is with Angelman, there's a lot to assess. There's really 6 domains that you can evaluate. There are tools that have been used for all of them in language, expressive and receptor communication, only the Bayley is commonly the [indiscernible]. You can evaluate sleep as well as growth motor and fine leaner scale. So there's always individual scale, whether done in a format, which is looking for a relative improvement or in an absolute [ rosner-type ] approach. So there's a lot of things that you can use. And I think the CGI score, the one that Ovid has. There's other ones that we've -- that genetics has, Roche has another one. Those types of scores, they've accepted them for Phase III, although whether validated or not, they've accepted them as a tool in Phase III. And usually, a global pressure scale score in order to be useful as the primary endpoint has to have some criteria or underpinnings of what's being determined to make the score. It can't be just doctors feeling about stuff. It's got to have some basis for where those -- that plus 1, plus 2 type decisions are being made. So we have CGI score as a global score and then we have one for each domain as well as the Roche score. And we haven't had that discussion yet with them. Right now, we're really getting to figure out our dose, but getting that optimized. But I think the more complex problem here is not which one to pick, but if you pick one, are you really accurately measuring the disease, right? Because we have this kind of almost a little bit of mine attention to the idea that because of statistical things, you have to pick one endpoint measure one thing. And I ask people, if you go to the doctor and he's going to say, "I'm going to measure this one thing and figure out how you're doing, you say what about the rest of me. You haven't asked anything else. We would never accept a single endpoint of the determiner of how we get healthcare. Is it? Or how doctors should treat you? Yes for some reason in clinical development, we've created this artificial structure for statistical reasons, which I call attorneys and statisticians to force us into this box and single endpoint, it's really wrong. And I'm trying to break through that. Now CGI that's a global score is a little bit breaking out of that box. That's what it had, right? And it was multiple domains kind of coming together in a global assessment that is kind of integrating more domains and overall impact. But we are focusing -- we believe that a better strategy is a multi-domain responder index type of design, where each -- you measure 5 endpoints and you basically add up wins or declines in each endpoint. And each patient, if they don't have a problem, it doesn't matter, they don't score it to 0, it disappears. And what we're not hearing about the CNS thing, the biggest problem is that when you're using a score that some of your patients don't have that problem. And so either you select only the ones with the problem, which then gives you a slice that's narrow and hard to enroll or you have a bunch of people who are creating noise because they can't have the problem, therefore, their score is being to dilute. The [ MRI ] eliminates that because if you don't change, you're not moving, it falls away, all we look for big movements up or down, which will only happen in problems that they have the problem, right? If they have a problem, it can get better, if they don't have it, it's not going to change. So we wrote a paper on the MRI. I think it's for Angelman would be a very preferred setting for it. And what I would ask [ Telone ] in CNS is, I think a preferred way to go at CNS, where you have very diverse clinical manifestations in a heterogenic population. And all of those disease scores that everyone gets sites and do those combination synthetic disease scores, they ruin power because they mix things where you might change with noise from other things where you won't change. And that mixture basically destroys. If you look at any kind of disease scoring endpoints, those have been horrible in terms of power and sensitivity, and we need to kind of change our way. But -- so I want to leave a little revolution on the change on endpoints. And that means changing the FDA, which, of course, I'm sure it sounds like a [ bronchiotic ] type of -- goal. But I do think there's some really big improvements if you're going to do clinical for any of these endpoints.

Yes. That's a great point. So I'll come back and I'll ask you in a couple of minutes about the multi-domain responder index and how you might weave that. Arnon, let me put you on the spot, still on a thematic question, you're in Alzheimer's, the FDA just went through a little bit of a public Alzheimer's experiment with Aduhelm. And I then want to open it up for everybody essentially. It was a great idea, right, use biomarkers, move toward an accelerated pathway, help the patients out, unfortunately, didn't quite work the way they wanted. In general, even in FTD, are you seeing any differences with how FDA is handling things? Or you already had your path, you already know your endpoints, it's really for you, it's a little bit of noise. It's not a huge amount of change.

I mean, there is an incremental change. I think that the FDA is not giving up on the traditional clinical endpoints sort of it's not going to substitute clinical endpoint with biomarkers. So I think that we still have to show that the noise the traditional cognitive effects of the drug and maybe biomarkers will help a little bit if the cognitive effect is on the margin, but we don't see substitution of the main clinical endpoint with surrogate endpoints. So what we don't see a big change in this regard. And in a way, the issue is that the biomarkers that we try to use, including the [indiscernible] probably are not great predictor of the clinical decline. So it is, in a way, justified to really ask -- still ask for clinical endpoints.

If I might add, Arnon. The problem though is -- I'm not sure are in your particular diseases, but [ many of the analog ] diseases we're talking about have irreversible components. So if you look for a linear relationship between a biomarker and that disease, it depends on the stage of disease, irreversibility. And so that failure predict, I think is a false conclusion. It's a false conclusion. And what I would say to you that in CNS disease, we need to get smarter about understanding primary disease activity biomarkers if we want to really treat disease, particularly in brain, where there's a lot of irreversible disease, which creates noise and hurts us. And there are many of those diseases were, you have to treat before they're symptomatic actually, to save their brains. In that setting, clinical endpoints can never work unless you are following for very long periods of time and have enough confidence in the decline in natural history because you're not going to run a placebo-controlled trial that's 5 years long. So I really think that we need to change this mindset that exists that biomarkers are compromised. If they're high-quality prior disease, they should be more predictive and get you closer to the right answer, and we just need to change that view. It's just that I would say to you, Arnon, do you think you could develop an HIV drug using hospitalizations for opportunistic infection? Did you develop a good drug? You could never develop a good drug for HIV that way, never. And so why are we so accepting of using a downstream after death and destruction of happen kind of marker when the disease process is happening way earlier way before that. We need to get our mind out of that. I agree, though, Arnon they have been very difficult. I just think it's a fundamental failing that we are so focused on field functions [indiscernible] we got to think, right.

Right. But who's failing is it? And I agree with you completely that yes, if we had sort of predictive that these biomarkers, we would use it. And for our FTD, we are using -- we are recruiting at risk patients that have the genetic mutation, but even this like when they are going to convert is an issue and...

Yes. Well, there's 2 failings. One is some of the disease are hard. We don't have an answer. There's no scientific answer. It's not a regulatory feeling. It's not -- I don't know who the marker is. I work on disease where we know the marker is because it's a metabolic block and you know what's accumulating and you know that you can -- it's right there in front of you, it's not rocket science to figure out. But even then, that's not been accepted. And for a disease like Sanfilippo or Niemann-Pick C or you were talking about Krabbe, all those neurology disease by the time you're seeing symptoms, those patients are already gone. There -- and particularly in [ late in fatal battens ], it's pretty clear, once they have symptoms, you're so far past the decline, they're going to decline in rareness what you do. Even in gene therapy experiment, they show that affecting them at a certain number of weeks doesn't matter the process already going past the threshold point. So I would just make a pitch your own...

And I'm not sure I agree with that in childhood neurologic diseases. So I think it's an open question to say -- and Bobby shaking his head too, and Geoff looks he's looking forward. So I think there's an open question whether or not doesn't want to take a shot at this. Whether or not intervention at disease manifestation will change the trajectory of childhood neurological disease is because we don't understand fully the plasticity that might occur in really young children. And we had 2 children in our first -- and it's only [ none of 2 ], but one of the children tile -- one of the children had a 24-month developmental delay based on both Bayley and Vineland and they started to regain milestones. And that's pretty stunning. I mean we talked to the investigators as...

Yes. I don't disagree. What you're saying, yes, there is some reversible disease for sure. What I'm saying to you is there's a lot, but every patient I'm sure that has to be different.

Fair enough. Fair enough.

And that is the problem, the noise. And so you'll have a few great examples and your other ones were somehow it didn't happen. [indiscernible] So I'm not disagreeing you. I think there's reversible disease. I'm just saying if there's enough irreversible and variable amounts of it that makes it hard to figure out what's going on.

Yes. And that's the whole goal, I think, of [ entry criteria ] of making sure that we have the appropriate level of disease to have the -- I mean it's like oncology. You want the disease to have a chance to respond. You want the drug to have a chance to also have an effect. And that balance is tricky. I totally agree with that, and I'll stop.

Yes. I mean I would just say just -- as a general note, I would take the view it's the earlier the better. I mean I think once -- in MLD, for example, our criteria are very strict as to who is eligible for therapy. It's -- they've got to be at a certain stage as far as cognitive function is concerned as far as motor function is concerned because what we've seen is beyond that time point, then is very, very difficult to stop the progression. But -- and other diseases may be different. The biology of other diseases is different the effect that it has on neuronal degeneration may be very different. And so I'm not saying that's a general thing, but that's absolutely the case in every disease. But I think as a general point, the earlier that we entered, the better the outcomes will be for these individuals. The one comment was made earlier by Bruce, and that is newborn screening. And as much as we're developing medicines, we should also be campaigning for newborn screening for these different conditions because the sooner we identify these diseases, the better all our medicines will have a chance of having a better effect and a more long-lasting effect as well. And just on MLD, we've obviously kind of started to invest in that. We have pilots now going in Europe. There's a pilot that started in the U.S. as well. But if you can get there, if you can identify these babies at birth, intervene with what are potentially curative therapies, you change the complete landscape of that disease, that disease doesn't exist anymore as a result of early diagnosis and as a result of a once-off intervention. And that is a price we're fighting for, I think. And I think it's got to be a combination of these developments together with better screening for these conditions.

Great. I think it's very worth. The problem is that if they won't screen for things until drugs approved, then you have this Catch-22 problem. You can't diagnose the patients early enough, you can run the trial to get approved and you take patients like with Sanfilippo or 3, 4, 5 years old, they already have damaged their brains. That is the problem.

I mean I would say that as soon as we get any kind of clinical readout, then that's really trying to put that in place. I mean, for example, for...

[ No, the rest ] won't let that happen.

Not the rest, but you can get kind of state-wide pilot starting to once you have a therapy that's at least in clinical trial and showing some early benefit. And that in itself generates the data that you can then take to the rest. So you can start early. You don't have to wait for the rest.

Yes, I agree.

Let me -- let's move to the round robin and Arnon with you. I'll just pick up on your last point, and then I'll move it to Lyla to move onward. So it looks like the FDA [ on the hills of ] Aduhelm is not willing to move to imaging alone as an accelerated pathway? Or what is there? And I'm talking about, I guess, more on the Alzheimer's side, but I guess also on your case on the FTD side, would imaging be acceptable with robust secondaries on hard cognition and hard functional endpoints? Or is that still very squishy?

Yes, I don't think that the FDA would commit just to say that biomarkers as approvable endpoint, I think they want to see still, yes, the clinical and sort of effect on the clinical decline means I think that anything else on its own is not going to stand. But that's something that, yes, we could.

So what would be an accelerated pathway then for Alzheimer's or FTD?

Yes, maybe sort of clinical improvement that's on the edge of statistical significant plus a lot of the other biomarkers and imaging tool could be acceptable, but there is -- I think the FDA will wait to see actual data before they commit. I'm not going to commit ahead of traffic.

Yes. Okay. It makes sense. Okay, Lyla, over to you.

All right. Thanks, Yaron. All right, Doug, let's maybe focus on 005, which is in development for a variety of indications, including [ Krabbe ] Huntington's disease, warm autoimmune hemolytic anemia and ALS. Let's maybe first focus on the recent update in Huntington's disease. So initial Phase II data came out earlier in the quarter, full data expected later in Q2. Could you maybe briefly just discuss the key learnings from the initial data? And what important additional data you're expecting from the full data set? What might you need to see in order to advance that into Phase III development?

Yes, sure. Good question. I think it leads a bit into the discussion we were just having on endpoints, biomarkers and the like. So our approach is to preserve functioning synapses in advance of neuronal death. So upstream approach. It's a process that occurs with the classical pathway and C1q targeting synapsis early in the disease process. In fact, before signs of symptoms, you have this wholesale renewable functioning synapsis and then ultimately, [ neuron own death ]. So we do like where we play in the disease process. We're really, frankly, quite encouraged by our initial data set in Huntington's disease. So importantly, we showed the drug was safe and well tolerated. [ Full tar ] engagement in periphery and across the blood-brain barrier [ metadozes ] we went in with. And importantly, we showed improvement on the key clinical measure, the UHDRS, in greater than 50% of the population. And for those patients who have access classical comp activity at baseline, 3/4 of those patients showed improvement. And so we're really encouraged by that. Importantly, we saw the improvement starting as early as week 6, and it makes sense mechanistically that if you are preserving the loss of functioning synapsis, which by the way, the loss of functional synapsis correlates most highly with cognitive decline and motor deficit. If you're preserving that really effectively right on the outset, you should see early improvement. And so we saw that. That was on the on-treatment portion of the study. The full data [indiscernible] will include both the on-treatment data as well as the off-treatment data. So we'll look and see what we see on the off-treatment data due to patients. Is there a durable response do the patients begin to decline, and you can go in different directions depending upon what you see there. But based on the early improvement in the patient population as well as what is potentially a precision medicine approach in this chronic neurodegenerative disease, we're encouraged with potential next studies, which would allow us to confirm these results in a larger placebo-controlled study and move forward from there.

Got you. That's very helpful. Maybe just a follow on also with the key point you made with the key biomarkers that can be used in these indications. So one of the key bar markets you're using is NFL, the neurofilament light chain. Could you maybe just briefly touch on how well established that is in ALS and Huntington's disease and how you've leveraged your learnings from the NFL trends you've seen in the Phase II update of Huntington's disease with that indication and then maybe also ongoing studies because we have seen some modifications, but I just want to make sure we have all the updates.

Yes, certainly. Well, NFL is certainly an interesting biomarker. And we and the industry is learning a lot more about it. I think it's very well established in articulated disease stage of disease. So I think it felt very useful in understanding stage of disease and progression of the disease. I think the open question is, whether it's a leading indicator for disease efficacy. So impacting NFL as a surrogate for disease efficacy. And I think in more aggressive diseases, chronic neurodegenerative diseases that are aggressive, that may have a peripheral component like ALS or SMA or even autoimmune conditions that are aggressive like [ MS ] relapse retaining data set and GBS are very aggressive autoimmune condition. You can inhibit NFL or you can show an impact on NFL and it translates to an efficacy improvement. I think in the slower progressing chronic neurodegenerative diseases, NFL may be the lagging indicator. And particularly in our instance, where, again, we're preserving functioning synapsis before you have neuronal death, and it fail measures depth of neurons. And so we think you are ultimately going to move NFL, but it may take years to do so. So we've taken those learnings from the initial data set in HD and done a couple of things. One, we extended our ALS study to give ourselves more time to look at potential outcomes, both on the clinical measures as well as on NFL. And secondly, and more importantly, as it relates to HD, we really have put our focus in on the clinical outcomes, the approvable measure, the UHDRS, which again, aligns very nicely with our mechanism of action. And so that really is where we've put our focus at this stage.

Got you. And then maybe I'll just have one quick follow-up before passing it over. In terms of touch continuing on NFL, you extended the treatment duration in ALS, is there any read-through to what you've seen so far on the trends for [indiscernible], for example, which is in an ongoing study of [ Phase II/III]? Is this a leading lagging indicator? How are you thinking about it in the context of that program?

Well, so [indiscernible] very aggressive disease. It is acute disease, the more neuromuscular paralysis really sits at the intersection of immunology and neurology and in our proof-of-concept study that's able to statistically reduce in rapidly within 8 weeks. And so we're still encouraged by that. We are in a Phase II/III study, as you alluded to. It's a placebo-controlled study, it's blinded. So we have not seen the data, but we are encouraged by a couple of things. One, as it relates to ALS. We're encouraged by what we saw in our proof-of-concept study in GBS [indiscernible] as well as on disability encouraged in what we saw in the Huntington's program on the -- not just [ on matters ], but the motor measures, particularly TMI and how that may translate into our ALS study. And so it's one of the reasons we gave ourselves a little additional time and less see an outcome on both clinical measure as well as in [ FA ].

Got you. And I'll kick it over to Brendan.

All right. Great. I actually wanted to kick it back to Geoff just for a little bit of a follow-up on [indiscernible] here. Obviously, a lot of overlap in terms of clinical approaches and challenges to some of the points that have been raised throughout the conversation here. But if we can dive a little deeper into your program specifically, given that there are a few commercially available ERTs in Gas Shape, can you maybe first calls what you think the real market opportunity here is for both Type 1 and Type 3? And then why you think maybe like borrow approach specifically makes sense here?

Yes, sure. So the -- I mean, just some simple numbers, there's about 14,000 Gaucher type 1 patients and about 1,500, 2,000 Gaucher type 3 patients. So starting -- and to your point, there is a safe, effective standard of care available enzyme replacement therapy. So one reasonable question would be, what is the unmet need and what is the commercial opportunity. To help answer that, the field actually track 750 patients on ERT prospectively for 10 years to sort of, I think, unequivocally answer the question, how is today's standard of care performing? And the result is that 60% of them are uncontrolled on one or more key clinical manifestations by year 4. So 100% of these 14,000 patients could benefit by being untethered from ERT. 100% of them could save the healthcare system millions by getting off ERT, but at least half of them would significantly benefit from better control. So I think that the commercial opportunity for Gaucher will be dictated by the results of our Phase I/II trial, but we think at a minimum, targeting the recalcitrant population, which is still a very, very big number. And of course, to turn that into revenue even today, standard of care generates about $1.5 billion, $2 billion of revenue. So these are large markets where ERT only does a partial job. But to bring it back to what the focus today is, which is CNS, type 1 is largely defined as not having neurological features, perhaps with the exception of GBA Parkinson's because, of course, the most common genetic risk factor for Parkinson's disease is the GBA gene. But if we focus on the Type 3 neuronal pathic Gaucher disease, where by definition, there's a CNS complication, the market opportunity is much more clear cut because these patients have a very serious CNS aspect to their disease and their treatment is they have to go to a hospital once every 2 weeks for a 1- to 3-hour infusion of enzyme, that doesn't cross the blood grain barrier. So today's approaches don't touch what is the key clinical manifestation. And of course, the whole argument for lentiviral gene therapy is, first of all, around pharmacokinetics is today's enzyme replacement therapy, you have to have an infusion every 2 weeks, and yet, it's gone from the blood within 12 hours. So you have the stop-start delivery of enzyme, which, of course, is suboptimal. But importantly, the lentiviral approach not only leverages the blood lineage to manufacture and distribute protein throughout the body, [ millions ] of leukocytes manufacturing and distributing protein. But commonly, people think of this approach as having gene-modified cells migrate to the bone marrow and engraft and then produce progenitor cells. But these cells also engraft in the microglia compartment of the brain. And that's really where the benefit of developing a way to address the CNS manifestations can occur. And I have to say we're very inspired by Bobby's MLD program by Bluebird's ALD program by 500 Hurler's patients in stem cell transplant. So we really don't think that we're pioneering a concept when we say that these gene-modified cells can migrate to the microglia compartment and in graft, what we're really trying to do is to apply them to a novel set of disease indications. And we think that the Gaucher disease type 3 application is great. What we also like from a business perspective is that a preexisting Gaucher program. So of course, -- we -- from day 1, we have CMC analytics worked out preclinical complete and a good safety database already on the back of Gaucher type 1. And to link this back to the earlier discussion, I do think that Gaucher type 1 -- excuse me, Gaucher-type 3 has a heterogeneity of disease manifestations, which aren't necessarily linked to the genotype. So this is one of those diseases where the multi-domain responder approach is actually a very good fit.

All right. Great. And I guess one last point here for you in Gaucher. I guess really tying into our earlier conversation from the data you guys have seen so far in Phase I/II. At this point, what do you think a potential registrational study could look like?

Yes. So in Gaucher type 1, we're -- these are early days. We're running our Phase I/II trial. We've released some data a year ago. We have not given a clinical update we plan to in Q2 of this year. So we have aggregated quite a fair amount of data that we've yet to disclose. But we can say that the program is very much on track. The Phase I/II trial in Gaucher type 1 targets patients that are well controlled on ERT. And in the case [indiscernible] for example, the Shire Takeda product, they discontinue ERT and then we track them over time. So the types of things that we're looking at is of ERT, can we normalize platelets and hemoglobin, which is an accepted FDA approvable endpoint, can we minimize degradation of bone and organ size, liver, spleen. And what's nice about these indications is that they all have a track record of being approvable endpoints, and they all tend to impact in short order. So I think we have a very clear clinical regulatory path in front of us with the exception of what I mentioned earlier is, are we going to target first line across the board? Or are we going to minimize this to target the patients that clearly have poor outcomes with ERT? And we just can't lock into that until we see more data.

Okay. Makes sense. And then I want to shift over to you, Bobby. I know we touched a little bit on the MLD program with [indiscernible]. I did want to ask really quickly about how the -- some of many of your key learnings from the launch in Europe so far and how that maybe shapes the way you're thinking about a potential U.S. launch, not just in MLB, but obviously for some of your other late-stage indications. And then I do want to make sure I ask you about the MPS programs after?

Yes, sure. So I think I touched earlier upon the reimbursement aspects of this, and I don't want to go through that again, except for the fact that the clinical data, I think, has been very, very helpful in shaping the response from the HCA agencies. So the second thing is really what does our commercial infrastructure look like for a condition like this. So what we've done is really go to -- because there's such a high unmet need, physicians will be willing to refer their patients to expert treatment centers. So we have these qualified treatment centers. We have 5 at present in Germany, in France, in Italy, in the Netherlands and in the U.K. as well. So they will receive patients, not just from their own countries, but from other EU countries as well in order to be treated. And for the time being, we think that is sufficient for the referrals as we get better diagnosis, better newborn screening, then we'll go to other areas as well. And we're already thinking of the Nordics and the [ Benelux ] as well to have other treatment centers there. They also, as a result, they also will serve as centers from which for patients from the Middle East and Turkey because we recognize that consanguinity has a part to play here, and there may be a higher incidence of patients in Turkey and the Middle East. And in fact, the third patient that we will treat on a commercial basis is coming from Saudi Arabia to Italy for reimbursed treatment. So that's the commercial infrastructure that we are putting in place. And because these are centers of both transplant and neurometabolic expertise, they will also be the same centers that we will go to for MPS I or MPS IIIA in the future as well. So that commercial infrastructure can hold not for us for MLD but for multiple neurometabolic diseases down the line. So I think that's one important learning. And we've had very, very positive responses from the treatment centers, very willing to act as treatment centers for this unmet population with such a high unmet need. I think the other important learning is about diagnosis, the need to get there early, they need to engage the community in disease awareness, diagnostics and in newborn screening as well. And I think we'll be repeating the same message of MLD to MPS-I and for other MPS's down the line. So learnings, I think in summary, I think the clinical data has excited not just patients but clinicians, KOLs and the reimbursement community. We have a plan as far as our commercial infrastructure is concerned that can be leveraged not for just from one disease but for multiple diseases. I hope that answers your question.

Yes. That's great. Super helpful. And obviously, it transitions right into the MPS 1. So obviously, you have Phase 1/2 and MPS-1 really wrapping up, plans to kind of start the Phase 3 later this year ideally. And then we're ahead, we thought update in the MPS-3A study world-like [indiscernible] last month. And it seems like we're still on track to get some cognitive measures from that study sometime in the first half of this year. So I guess as you kind of look at the different endpoints, obviously, different disorders, different indication completely, right? But you're kind of takeaway from these really multifaceted and genetic disorders here. What really is kind of guiding the way you're thinking about designing a pivotal play based on that. Because a lot of different ways it could go a lot of really important functional endpoints here. So what -- where are you kind of -- not just in your conversation with FDA, but really when you're looking at the data, what's kind of guiding some of these decisions?

Yes. So I do think the diseases are slightly different when you take MPS-I compared to MPS-III because MPS-I not only has the neurocognitive problems, but it has some systemic problems as well. And importantly, remember, in MPS-I, there is a standard of care in allo transplant, which you have to compare against. So the study design for MPS-I is being put together as we speak. I mean, we went through last year and had interactions with the FDA and the EMA. So we held a parallel scientific advice meeting and got input from both of them. And as you might imagine, the inputs are not necessarily the same and at times are slightly contradictory even on the same subject. So there's -- you have to kind of balance all of that and come up with a global study design that is compatible both for the FDA and the EMA. So that is what we're putting together now. And our milestones this year are to get that cleared midyear and then start the study by the end of the year. But what I can say is -- and we'll give clarity on exactly what that study design is, how many patients, how long the follow-up for once we've got IND clearance. But I think the important messages are what Emil alluded to before, that it's not just about one specific outcome. It is about multiple different outcomes to look at and also against the standard of care, which is a transplant. So as you might imagine, we want to show that there is better survival that we don't get the same transplant morbidities such as graft versus host disease, graft failure, et cetera, but also important clinical benefits such as cognitive outcome. And the important -- and the detail really is about how you construct and put all of that together into an approvable endpoint. And that's where we've been kind of putting our heads together to think about how best to do that and statistically how we show that it is better than an allogeneic transplant. For MPS-IIIA, it is different because you don't get the same systemic manifestations. There isn't a standard of care. There's no ERT, there's no transplant option. And so this now is about showing benefit over the natural history. And we've talked about this before that you need to have a natural history that is very carefully put together and that is -- that again, the agencies are comfortable with and then how you demonstrate an improvement above the standard of care.

Okay. All right. Great. Super helpful. All right. Bruce, I do want to ask you a little bit about the GM1 program. I know we touched on some of it earlier, but obviously, you guys had some pretty exciting data from the first 2 patients treated came out in December and then updated again world. So I guess one question I do want to ask you, obviously, you're looking at a bunch of different biomarker responses across the board here in GM1. But how do you think we should interpret the biomarker response typically beta-gal activity and ganglioside levels relative to a patient's cognitive response. Obviously, there's variability just like we touched on. But based on your data so far and natural history data, is there a clear correlation here that we should be able to expect?

Really brief and say it's too early to say. I think that -- I think this is a challenge across the field, which is there are very few total number of GM patients with 1 even treated with gene therapy and much less so in this patient population of early infantile. So we've seen one patient which had a 2-point fold increase in beta-gal, see a decrease in gangliosides within 30 days. That's fantastic. The second patient saw a 1.6-fold increase in beta-gal and I should say even though it's just ICM, we saw it in the serum as well as in CNS, which was -- which we replicated the animal studies we've seen. And in that second patient, we saw an increase in ganglioside -- GM1 ganglioside and then stability for 6 months. The issue with gangliosides is we have no context to put that in. What's the natural history. If this patient had normally developed with the disease, maybe would have skyrocketed. We just do not know. So we're trying to pull in information for the natural history study we're running of looking at both beta-gal and gangliosides over time to compare it. And I think this is, by the way, not specific to GM1, across the field, we're looking for the first time at CNS measurement of some of these enzymes, some of these outcomes and trying to put it in an interpretive set and we're the first ones to measure CSF beta-gal because it hasn't been -- we've had to modify the assay, then we have to put it into natural history context, et cetera. So I think we're all struggling with the collective -- in the field, we're all struggling with the collection of biomarkers. And if you don't necessarily have clinical outcomes such as these validated scales, then how do you interpret those changes.? So I think, Brendan, it's -- what we've been saying is it's early to say whether GM1 ganglioside in this particular setting can be interpreted as predictive of clinical outcomes. And we'll also just -- I'll just make the last point that we're measuring CSF serum, but we're not measuring tissue like the brain or peripheral tissues. We don't know what's going on in the specific neurons that are taking up, hopefully, beta-gal and maybe locally, there are areas where gangliosides are going down. it's impossible to really know these things. So that's why ultimately, the clinical outcomes were so gratifying, and we have to keep looking at this. To Bobby's point, we need longer follow-up to make sure that this is stable. We need to see if we can correlate all of these biomarkers together and put together the totality of the story. So we're right on the cutting edge for some of these diseases.

And I think that ties in really well to what we were talking earlier about. Okay. So Emil, I did want to take a chance really quick before we have to wrap up here to ask you about the Angelman data coming up. Obviously, a lot of really exciting news coming out of the program so far. We're expecting 128 data around midyear of this year. One question that we do get obviously the stake has been pretty good at this point after the updated protocol. So if you get to that point, first, I guess, first of all, where do you think the bar is for efficacy in that Phase 1/2 study? And I say that just I asked that just said do you think you need to maximize before moving to a Phase 3 and kind of in that same vein, if you don't see any safety problems at that highest dose by then, do you think that you would expand additional higher doses?

Thank you. So one thing I would say is very surprising and helpful is that we're actually seeing clinical changes. We did see clinical changes in a short period of time. Because honestly, when we started the [indiscernible] program, I thought it might take a year after you knock it down, it might take a year to see something. So it's a little bit surprising you can actually use clinical endpoints to assess what's going on in a matter of 2, 3 months. So that is a plus -- that was probably the biggest win of the early data is that actually a patient that doesn't speak words can speak, go from no words to 12 words in a 3-month period of time. I think that's just unheard of that kind of change. Now our view on what we're trying to do is we're trying to knock down the [ anascript ] to get the expression occur, but we're looking for a trajectory, Brendan. So we talk about having plus 2 changes that has very much improved at least 2 domains in a period of a couple of doses, we're kind of what we're trying to gauge are do we have a trajectory of improvement. If we see a trajectory improvement that's steep enough to suggest it's in line with what we were seeing before, we're going to hold at that dose and watch how they do. And as long as the patients are getting to gain ground from a developmental function standpoint that is [indiscernible] improvements in different parameters that we're looking at, and then we'd have confidence that the knockdown must be enough to release enough [ UV3 ] to improve their function. Whether it's good or great or how much it may be hard to know, but we think that the amount of change that you can see that within just a couple of doses that would give you 2-plus changes, much improved to 2 domains is a big amount of change. And therefore, we feel pretty comfortable that threshold should take us to a good place on efficacy. And since we know what that looks like for the patients we did before, we kind of have a sense that this is possible and tractable. So far, we've seen good clinic activity. We haven't gotten to 2 domains or plus 2. We have a lot of plus 1 domains and multiple changes. But the encouraging part is we're seeing the same type of changes that we saw before, each patient having somewhat different pattern as you would expect, because each patient with Angelman has slightly different severities and differences in how they express the disease. But so far, it's been good. By midyear, we would hope to have these 2 cohorts treated through board doses and have their output day 120 to 8-day output. Our expectation is though, as we get through that with the safety that we will add an expansion cohort where we start the dosing at the newly cleared doses and allow ourselves to ramp up the efficacy further as we move through the therapeutic window. So we're encouraged where we're going to be in midyear, but I think we will want to increase the dose to the point where we're getting the kind of trajectory of development, I think, would indicate a meaningful change for patients. And I think we have still a lot of room because we started relatively low in the window because we're trying to avoid the lower extremity weakness safety event. But so far, the safety has been clean. So we're pretty encouraged by the changes in how we're administering.

Right... Okay. I think one last thing for everybody in 5, 10 words or less, just you'd like to end these sessions by what you think the biggest surprise in the development of orphan phenotherapy is going to be over the next 10 years. So you want to start, Arnon, what do you think?

Yes. I think there will be therapies for major neurodegenerative diseases, both rare and prevalent. I think that we know a lot more about the genetics and mechanism of diseases. And so I'm pretty hopeful that there will be transformative medications.

Doug, what do you think?

Yes, I completely agree with Arnon. What I'd add to it is I think we're going to see the advent of precision medicine more in some of these larger neurodegenerative indications. These patients have a lot of heterogeneity. Metrics always been approached from that perspective. I mean, I think you'll see some breakthrough advancement in that area as well.

Jeff, what would you say?

To keep with one of the things today, I think that the regulatory pathways will continue to evolve to adjust to the unique nature and needs of rare disease patients.

Bobby?

Yes, I think it's about understanding the genetics -- if we know what the genetic susceptibility is that we can understand the mechanism, and therefore, we can intervene to that pathway and also that mechanism. And we now have tools that will deliver genes and proteins to the CNS, and that's going to be extremely exciting.

All right. Bruce, do you answer?

I was going to go with all of those, but I'll come up with a new one. So I think one of the things that we really do need is -- and what will be important is better diagnostics from a CNS and CSF perspective and really understanding what is -- going back to the biomarkers like what is actually happening on a cellular level, if we can, but going beyond the CSF and really understanding and that will help development, I think, and accelerate these programs.

All right. Emil, last line of the question.

Sure. I think that the surprising thing is going to be seeing how many diseases we thought were untreatable, unmanageable that will actually have figure out a way to treat them. And I'm looking at Angel we're doing now, to me, that's just -- I thought as a geneticist, that would never be treated, right? And now all of a sudden, there's tools and ways to do it. And I think -- the advent of so many multifactorial approaches to treatment, you're going to open the door to treating things we haven't achieved before us going to be very exciting. And the only thing is, can we get the regulatory process to catch up with biomarkers and endpoints. So we start doing smarter, more efficient development because if it gets harder and harder, we're not going to be able to deliver what I think we could deliver with the science we have.

All right. Great. Well, thanks for everyone for joining us today. It's been a great conversation. Appreciate everybody sitting around for actually a few minutes and enjoy the rest of your day.

Thanks everybody.