Taysha Gene Therapies, Inc. (TSHA) Earnings Call Transcript & Summary

Great. Good afternoon, everybody, and thanks for joining us for our last session here today. Very pleased to have Taysha with us. Before we get started, I just need to read a quick disclosure statement. Please note that all important disclosures, including personal holdings disclosures and Morgan Stanley disclosures appear on the Morgan Stanley public website at morganstanley.com/research disclosures. So pleased to have a broad range of the management team here. We have RA Session, the CEO; Suyash Prasad, who's the CMO. I think that's your title, Suyash, sorry; and Kamran Alam.

I thought maybe a good place to start because it's so recent is just the FDA obviously had their 2-day hearing on AAV. So I thought maybe you guys could just comment on what you thought are the key takeaways from that and any implications you see from that.

No, a good question, Matt. And first and foremost, thanks for having us today. We really appreciate the opportunity to be able to participate. I'll start and maybe Suyash can give some thoughts. I think there was a lot of good discussion coming out of the meeting. I don't know if there's a lot of takeaways coming out of the meeting. I do think what was really interesting and what it actually did was reinforce some of the things that we were already doing. So it was good to know that this is where the FDA was thinking. I think we've been fortunate. In the last 3 months, we've had about 7 interactions with regulatory authorities around the world, including the FDA, regulatory authorities in Europe as well as in Health Canada. And I think what we are hearing is there is an alignment around some of these key issues that had been coming up in gene therapy, particularly the focus around toxicity. I think probably the one issue that did come up that we -- that was a little bit more relevant to what we were thinking about was this whole notion around DRG inflammation and just the agency thoughts around there, but it's probably where they spend the least amount of time, which was pretty interesting. But what I would say is it really centered around kind of practical issues of how to manage and how to look for potential areas around DRG tox, whether you're doing nerve conduction studies in the clinic, how to look for these in preclinical studies, in animal models. We're pretty fortunate to have is we've looked pretty extensively across a large range of animal models, both rodents and NHPs, and we haven't seen this. And Suyash can over some of the recent data we have, particularly for our GAN program, but it is important to understand what's top of mind for the FDA. I'd probably say the majority of the day was really dealing with liver tox and kind of systemic delivery of large doses of gene therapy. And some of the key takeaways there were really relevant to us that the panel really talked about was more targeted routes of delivery. That's exactly what we're doing. If you're trying to treat the CNS, intrathecal delivery, direct to CSF delivery, that was a key focus. Having fully characterized material and a focus on CMC early, not moving from, let's say, adherent to suspension into [ macula ] making sure that you have fully characterized material and consistent material, high-quality material from the time of your pharmacology studies all the way through tox, all the way through the clinic. That's something that we take pride into as well as this whole notion around empty-to-full capsid ratio, that they've kind of woven to some of the tox issues. And we take pride in having a -- we take pride in striving for, at the very least, 10% empty capsids, so 90% full capsids and even in our GM2 program kind of achieving above that. And so we -- these are things that we thought about and things that we've woven into our development program. And to be quite honest, they cost a lot more to do it right on the front end, but it obviously saves you time. It buys down risk and improve probability of success the further you get along the development pathway. I'll stop there. Suyash, I know you have some thoughts on this as well.

Yes. Thanks, RA. I think it covers quite a lot of it. But in general, I think there's a good -- 2 days of good discussion, and I think we all needed to kind of come together and have that discussion. Having said that, much of it was not really new to us. There are a few gems here and there, a few pieces of information that may just perk up our ears but in general, most of what we already knew. And as RA mentioned, over the past 3 months, we've had 7 regulatory agency discussions across 4 countries and safety, always, it's a large part of that discussion. So a lot of the plans we put into place across our programs, a lot we've discussed at the FDA meeting were known to us already. At a high level, I think the fact that CMC characterization and looking at the attributes of CMC product, quality, purity, how do you downstream processes work, et cetera, and weaving that in with safety in general, I think was an important discussion to have. And what I hope actually coming out of this meeting is maybe some discussion on standardizing CMC characteristics across companies. No one's really doing that, but that will be a nice thing, I think for the [ team ]. I think there's a little bit of a focus on maybe think about using slightly lower doses, maybe about using slightly weaker promoters just in general to have an eye -- keep an eye on safety matters because a couple of years ago, people were thinking, in the field, we can dose pretty high -- dose high. In that way, we're going to have a nice durability of effect. But I think the FDA pulled back a little bit from that and thinking, okay, you [ can ] get by with a modest strength promoter [ on ] a strong promoter, you're better off doing that. You can get there with a low dose and a high dose, do that. And it actually fits very nicely with our narrative because when we're giving drug intrathecally, you, by definition, are giving a lower dosing for perhaps systemic drug. And then the other thing that came out was about targeting organs more specifically, i.e., trying to avoid the liver if it's not a liver-directed gene therapy, trying to avoid the heart perhaps. And once again, that fits nicely with our narrative. We are targeting the brain and the spinal cord intrathecal delivery. At a high intrathecal dose, it is actually relatively low systemic exposure. So a lot of what we're doing already fix that thinking. And there was also mention appropriately of being very clear about NEDs and being very clear about [ no Ls ] and selecting the right dose for the clinical trial within that range. And I think it was good that they have some discussion on empty capsid, but I don't think it's quite enough. I think there's still a lot more to be delved into around empty capsids. I've always felt empty capsids are a contaminant. They add unnecessary, needless viral load. And as RA mentioned, our targets get less than 10% empty capsid with an [ 8 capsid drug ]. It is very hard to do 2 things: first of all, reduce the amount of empty capsid without reducing yields; and also it's very hard to make sure that process is consistent from lot to lot. And I think it's something that companies need to address and focus on because it has a clear impact on safety. I'll stop there. There's more we can talk about, but I'll stop there for now.

Yes. That's great. I mean -- and I think one of the other things that I want to touch on is, obviously -- and you've highlighted this, you've had a lot of regulatory interactions. And maybe not asking for sort of the overall pitch but just how those regulatory interactions impacted? How you guys are thinking about what you need to do over the next months and year in terms of the key programs?

Yes. Maybe I'll start and Suyash, please chime in. And Suyash has kind of been leading those interactions on our behalf. What I think the value of having multiple regulatory interactions in multiple jurisdictions, you tend to actually see the same people over and over again, particularly outside the U.S. Obviously, at the MHRA, it's kind of the same team. In Europe, it's the same team. Health Canada, it's the same team. You typically don't get the turnover or just kind of the novel team structures that you would normally get with the FDA outside the U.S. So what happens is you're able to go in and have more of a platform type of conversation, to be quite honest, because they know the next meeting is coming up with them in a month's time, and you're talking about another program. So it becomes much more of a collaborative discussion, and you're able to take learnings from one program and apply it to the next, again, kind of central to our thesis around economies of scale and risk reduction. And you really are able to broaden the discussion more versus just kind of going into a single agency, particularly where the U.S. and understanding kind of where OTAT is. They're just overworked obviously with COVID, the number of programs going through that. They're down about 100 employees. And so what we're getting is a lot of people that may not necessarily be as experienced with gene therapy, maybe certainly experience from a regulatory perspective but maybe new to the gene therapy division. And you tend to get a little bit more of a check-the-box type of discussion versus a more kind of in-depth collaborative discussion because you're having this -- because you're having the conversation for the third and -- for the third time with the same people. You're just talking about a different indication. I'll stop there, but Suyash, maybe you -- obviously, you've been in all those discussions.

Sure. And I think we approach the regulatory meetings -- but RA is absolutely right. Philosophically, everything we do, we do as a platform, and we approach the regulatory meetings in that way. We walk into these meetings saying, okay, we're going to have several conversations with the FDA over the next year, several conversations with the MHRA and with Health Canada and with PHI in Germany. And we really walked into those meetings with that approach and tried to link a second meeting, what happened in the first meeting [ approach with ] our future meetings. And that's -- at one point, one of the agencies actually said at the end of the meeting, "Great discussion. We look forward seeing you next time." So they're approaching this as a collaborative endeavor as well. What I will say, the tone of the meeting every single moment, extremely collegial. We've learned from the regulators. The regulators have learned from us. And we -- there are certain topics of recurring items. For example, potency assays and other CMC characterization and monitoring for DRG inflammation, every regulator wants to talk about those 2 topics. And then each regulator has their own little bits and pieces they want to know about as well. And so we're getting very good at talking about our approach not just for the particular indication we're talking about with that agency but also across our platform. For example, with DRG, it's very clear now how we're meant to monitor in a clinical trial. And we do this in 2 general ways: There's clinical ways of monitoring [ where ] inflammation, and there's neurophysiological ways of monitoring that inflammation. The clinical ways are we perform deep tendon reflex examination at baseline and on an ongoing basis. And if the reflexes become depressed at all, that's a sign something maybe happened with the DRG. We also look at sensory capability through 5 specific modalities, pain, temperature, vibration, et cetera. And that coupled with nerve conduction studies have baseline every 3 months of the first year and then 6 months thereafter. That satisfied every agency in terms of monitoring the DRG inflammation. That is despite the fact we've never actually seen it in any of our animal studies. So these discussions are going very, very well. They're very nice. I actually enjoy them because it's really a good way of learning how the regulators are thinking, and we've got an opportunity to shape that discussion. They are always excited talk to us because we're bringing some really interesting, potentially life-saving highly innovative treatments to them. I'll stop there. But in general, they've been great meetings, and I'm looking forward to the several more we have coming up this year.

Suyash, maybe you want to comment just quickly on -- just since we were on the whole notion of DRG inflammation, maybe you want to comment on some of the recent data we got from our NHP study in giant axonal neuropathy.

Sure. Yes, there's 2 key pieces of data that we have on DRGs that, I think, really helped shape the discussion. The first is -- and this was discussed at the meeting last week. The first is the fact that not everybody sees it. Jim Wilson has seen it. He's published extensively on it. A couple of other centers have seen it. But actually, many centers have not seen it, even though we're all looking for it. We have not seen any DRG inflammation in any of our NHP studies across our range of programs. Most recently, about 3 weeks ago, I received the study report from our GAN redosing NHP study, where we took 8 NHPs. Those -- 4 of them on 2 occasions, so they received 1 dose of the GAN construct followed by second dose, and they would actually dose very successfully. And 4 just received 1 dose. And there were new proxy 2 months later, and there's absolutely no inflammation of DRG. The DRG were pristine. That's the first piece of information, I think, that's important on DRGs, the fact that not everybody sees it. The second piece is that, in some diseases, we actually have to get drug in for DRG. So avoiding the DRG is not the right thing to do. GM2 gangliosidosis, CLN1, giant axonal neuropathy, all these neurological diseases have DRG pathology as part of the disease process. We know more toxic data [ for ] preclinical data. And in our GAN study, we've actually -- in our preclinical package, we actually took GAN knockout rodents, treated some with vehicle and treated some with drug and looked at the DRGs. The rodents treated with vehicle showed significant degeneration of the DRGs and neuronal inclusions, which are the pathological aggregate that collects; and the rodents that were treated with drug actually have much better looking DRG. So actually, with GAN, the CLN1 or GM2, we actually need to get drug into the DRGs to actually help correct the underlying pathology. And I think there's a better -- a nuance there that people sometimes forget, people thinking, avoid DRGs, avoid DRGs, avoid DRGs. That's not the case for every disease.

Okay. Okay. Good. I think that's an important point, and hopefully, we can watch how that evolves because it makes sense. It makes sense. I want to make sure we touch on some of the programs. So maybe we could talk about GAN first. And obviously, at the R&D Day, you had some natural history data. You have some visual acuity data. Maybe just put that in context for people and how you think about that data influencing that program.

No, absolutely. Suyash, do you want to start? And I could just give some voice over.

Sure. So at the R&D Day, we represented a lot of our motor function data with the MFM32; but actually, the new data we shared, you mentioned it. That was the eye data. And this was really interesting to us. And I think it's important for the neurological diseases to see the eye and the optic nerve as extensions of the central nervous system. So it's kind of the outward manifestation of the central nervous system. And so across our program, we take a really close look at the eye and have a neurophysiologic look around the eye functions. It's also critical because it's one of the major features that affects patients and families in a really emotional way. You start to lose your vision. It's devastating to patients and families, especially in children who are losing their ability to move, losing their ability to speak, losing their ability to hear. Often, the vision is the one area of communication that still works for a while. And when they start losing their vision, it's really devastating. So from those perspectives, it's really important to look at that particular capability. Now we presented data using our visual acuity from GAN, is in something that is the LogMar, which is a logarithmic test. It's basically a test of visual acuity, a bit like a Snellen chart, which you do when you go to the optician. There's more research-focused way of doing that. And the patients dosed with drug, all but one of them, their LogMar scales are deteriorating over time to the point at which you can predict when they're going to go blind. But with treatment -- and the slide's in our corporate deck. It's a nice slide to look at. But with treatment, you actually halt that decline in visual acuity. And once again, it speaks to the need to treat early and prevent that visual loss over time. And that mirrors and reflects what we saw in the MFM32, which is the motor function scale, where these patients are dropping [ about ] 8 points a year off treatment. As soon as they get treated with our medium low dose or our medium high dose, that flattens and so they stop deteriorating. And we have the high dose data yet to come but very compelling on motor function and eye outcomes. And we have a whole host of other end points yet to share with you. I'll stop there. There's more I can talk about but I'll...

Yes. No, but I think that's great. And I wanted to -- RA, I don't know if you want to make comments, but I also want to make sure we just give people a taste of what's to come here at the end of the year in terms of regulatory update and the high dose data. And just make sure people have the right expectations [ on what we're ] going to get.

Yes. No, absolutely. I think I'll start, Suyash, and please chime in. I think what we've been able to demonstrate and what we've informed The Street on are 2 doses that Suyash just mentioned, the 1.2E to the 14 total Bg dose and the 1.8E to the 14 total Bg dose, where we basically see a total arrest of disease progression. And I think if you looked across those 2 doses and that data set and how it just differs, basically is an 8-point difference from the natural history in the MFM32. You couple that with the totality of data, the visual acuity data and some of the other end points that we've had, plus the durability, safety -- long-term durability, long-term safety, you would basically say this would form the package of something that should be approvable at a regulator. Obviously, Europe has a very clear pathway around conditional approval, and that's really going to be our going in discussion with the European regulators. And obviously, at the U.S., it will be around accelerated approval. And we've seen some flexibility at the FDA, particularly in the Department of Neurology there with the recent approval of Alzheimer's drug. And even if you go back -- and it seems like a long time ago. But even if you go back to some of the guidance that was actually issued in January of this year by the FDA around the development of gene therapy for neurodegenerative diseases, this study checks all the boxes, right? Plus, we still have the high dose data yet to come. And I think when you look at the data set, one of the clear things around development of gene therapy for neurodegenerative diseases, one of the most important variables are going to be its early treatment because once you've lost the neuron, you can't regenerate a neuron. Once you've lost that neuron, that -- your neuron is gone. But if you're able to identify patients early, you're able to stop kind of disease progression in its tracks. So as we look at the data set that will be coming out later this year, what we see as a clear win is halting disease progression, kind of performing in line with the other efficacy data that we generated earlier this year. You may see an improvement. But at some point, if you treat a patient later in disease, the most you're going to be able to get is a halt in disease progression because you're not able to regenerate a neuron. So for me, anything that establishes kind of consistency in the data set where you're either halting disease progression, significantly slowing disease progression, that's going to be a clear win. It supports, again, and validates the data set that we've generated to date. Keep in mind, patients that have been treated at the middle high dose and the middle low dose, that's 8 patients, and we'll have 3 patients treated at the high dose. What we're fortunate of as well is, by the time that we had the data for the last patients and the 1-year follow-up, we'll have a few patients that are actually approaching 18 months. So we'll have a little bit more data, longer-term data and durability data that we'll be able to elucidate and provide a little bit more robustness to the data set as well. I'll stop there. Suyash, maybe you want to give your clinical assessment.

Sure. Yes. From a clinical perspective, this checks so many boxes, the data package. So it makes me very comfortable going in and talking with the regulators. We've got this wonderful natural history study with years of prospectively collected data in a very disciplined way. We've got clear stabilization of disease at the median line and even high dose -- and high dose yet to come. And that stabilization is clinically meaningful. We've got long-term safety, long-term efficacy and long-term durability. And I find myself in a somewhat unusual position because most of the time I go into a regulator with a discussion on pathway to approval with maybe a year's worth of long-term data. And I'm having to make the argument that, that will -- is reflective of 3 years to 5 years. We've got -- the first patient was dosed in 2015, so we've got at least 6, 7 patients with 4 years' worth of long-term safety and efficacy, durability data, so it really puts us in a very nice spot. As RA says, this checks all the boxes from a conditional approval perspective in Europe. It does check all the boxes from the FDA guidance issued earlier this year, although the FDA was a little unpredictable. So I mean [ to be honest, it puts us ] either in a good way or a bad way, but we're a little more cautious about the FDA, a little more confident when we think about our operation in Europe. But I'm looking forward to having these discussions.

Yes. I think just to contextualize it, we think, from our own building out our commercial infrastructure and for our own modeling purposes, we basically slice the world in half. We have ex U.S. and we have U.S. For ex U.S., we're going to go into MHRA, go into EMA, and we're going to talk about the conditional approval pathway because we check all the boxes for that pathway. We're fortunate -- I've worked on programs that have gotten conditionally approved. Suyash has worked on a number of programs that have gotten conditionally approved, so we're pretty confident about that. And keep in mind, the conditional approval supports reimbursement and named patient reimbursement in multiple jurisdictions in Europe and outside of Europe. So it takes care of the whole EU, but it -- also, you're able to reference that approval in Turkey, which is a great rare disease market; the GCC region of the Middle East, which is a great rare disease market; parts of South America, particularly Brazil, Colombia, which are very good rare disease markets; Israel, which is a great rare disease markets; and parts of Southeast Asia. I've actually been fortunate to actually commercialize the drug that was conditionally approved in those markets doing exactly what we plan to do here with giant axonal neuropathy. So that's how the ex-U.S. strategy is working. And we model in a 2023 approval. In the U.S., we think there's 3 scenarios. Two of them are high probability. One is a little bit less probability. The 2 scenarios that we're modeling out and working on is the first is, basically, as Suyash said, the study checks all the boxes. We go in and we make the argument to the FDA that, look, we're using a like-for-like commercial process. We're manufacturing the drug at the same CDMO partner that manufactures the clinical trial material. We're manufacturing the drug using the same cell line that we did with the clinical trial material. We're using the same bioreactor. We're just using the scaled-up version, same media, same excipients, and we're doing it. And the only change is that we're taking it out of the clinical facility and moving to the Phase III commercial scale facility and that we have a nice analytical comparability package. We could use that to support Module 3 and file your BLA. This is a lot easier than what we were doing in our previous life when we were moving from a HYPERStack to an iCELLis when we were trying to get Zolgensma approved. This is a like-for-like process. There's not any difference in that. So it should support analytical comparability. So that's going to be our going-in strategy. That's scenario 1. Scenario 2 could be the FDA comes back and says you guys have made some compelling arguments. You're absolutely right. The study checks all the boxes. But what we'd like for you guys to do, instead of analytical comparability, do some clinical comparability, i.e. treat 2 patients with your commercial-grade material and then use that as a basis for your Module 3 and file your BLA. We think there's about a 6-month gap between scenario 1 and scenario 2. Scenario 1 would be a first half 2023 approval. Scenario 2 would be a end of 2023 approval, right? So we think there's about a 6-month gap. Scenario 3, we think is unlikely, but obviously, I'm not the regulator, right? But that would be where they basically say, look, guys, this is a great study. Go do it again, do a pivotal study. This would go against the guidance, but obviously, the FDA has their own mind, and then they can make that decision. But what we're basically going to go do. And when I say we, Suyash and the regulatory facing team is going to go in and do is make the argument with the guidance document in hand to basically say this checks all the boxes from the guidance that you guys issued and have more of a collaborative discussion.

Okay. Okay. Good. No, that's a perfect outline. Look, I mean I know there's -- we didn't get to GM2 and a lot of other stuff, but unfortunately, we're at the end of time. So I'll look forward to the next discussion and all the additional updates we're going to get from the other programs in the second half of this year. So -- but thanks for being here this afternoon.

No, Matt, thanks for having us. We really appreciate it. Great discussion.

Okay. Bye-bye.

Bye.