Voyager Therapeutics Inc

NASDAQ:VYGR

Good afternoon, and welcome to Voyager Therapeutics Fourth Quarter and Year End 2024 Financial Results Conference Call. [Operator Instructions] Please note that today's call is being recorded. A replay of today's call will be available in the Investors section of the company website approximately 2 hours after the completion of this call.



I would now like to turn the call over to Trista Morrison, Chief Corporate Affairs Officer at Voyager.

Good afternoon. We issued our fourth quarter and year end 2024 financial results press release this afternoon. The press release and 10-K are available on our website. On today's call, Dr. Al Sandrock, our Chief Executive Officer, will briefly review key recent and upcoming milestones, and we will reserve most of our time for your Q&A. Joining us for Q&A are Dr. Toby Ferguson, our Chief Medical Officer; Dr. Todd Carter, our Chief Scientific Officer; and Dr. Nathan Jorgensen, our Chief Financial Officer.



Before we get started, I would like to remind everyone that during this call, Voyager representatives may make forward-looking statements as noted in Slide 2 of today's deck. These statements are based on our current expectations and beliefs. They are subject to risks and uncertainties, and our actual results may differ materially. I encourage you to consult the risk factors discussed in our SEC filings, which are available on our website for additional detail.



Now I will turn the call over to Al.

Good afternoon, everyone, and thank you for joining us. As Trista said, we plan to keep our remarks brief and prioritize your questions. On Slide 3, I want to remind you of why we're so excited about Voyager. Our pipeline includes 4 wholly-owned and 13 partnered programs. We have already begun to generate clinical data, and we have multiple opportunities to generate more in the coming years. We are particularly excited about our 2 wholly-owned programs targeting tau, which we view as the most important target in Alzheimer's disease. We also have 2 platforms to enable CNS delivery. I think most of you are familiar with our TRACER capsid platform for IV-delivered CNS-targeted gene therapies. We are also generating data with our ALPL-based nonviral shuttle. I am hopeful we will be able to share some of that data with you later this year.



Finally, our partnerships have been a significant source of non-dilutive revenue for us. That's a big reason we are able to report $332 million in cash as of the end of 2024. And with $8.2 billion in potential future milestone payments, we believe partnerships will continue to contribute significantly to our bottom line. As I always say, we are open for additional business. We are always discussing new partnership opportunities. While we are building a multimodality neurotherapeutics company here, I want to make a comment about gene therapy, which comprises much of our current pipeline.



Despite continued setbacks in the field, it is possible to create a gene therapy that drives value for patients and investors. Zolgensma proves this. I want to emphasize that many of the foundational principles behind Zolgensma's technical and commercial success are principles Voyager also adheres to. This includes focusing on genetically validated targets in severe diseases with high unmet need. It also includes IV- delivery, which we view as critical to commercial viability. We believe IV- delivered AAV capsids will be required to enable gene therapy in most CNS diseases given the limitations of localized delivery. The potential of our IV capsids to efficiently deliver across the blood-brain barrier, not only in infants, is presumably why Novartis came to us for an SMA gene therapy partnership. I'm not going to belabor this point, but I do think it is important to differentiate Voyager's approach from the broader gene therapy field.



On Slide 4, you can see our pipeline. I won't go into a lot of detail here other than to point out that our SOD1 silencing gene therapy program did move back into the research stage as we announced last month. The payload did not meet our target profile, and we are going to need to identify a new payload to advance that program. At the same time, I will note that VY-1706, our tau silencing gene therapy, has moved forward into IND-enabling studies and is advancing toward IND in 2026.



On Slide 5, I will note a few more quick highlights from the quarter and upcoming milestones to watch. I mentioned that VY-1706, which was selected as a development candidate in Q4 2024 has now advanced into IND-enabling studies. We are really excited about the data from our 3-month non-human primate studies, where we are seeing 50% to 73% knockdown of tau messenger RNA quite broadly across the brain. We have previewed a little of this data in our corporate deck on our website, and we will share more at the AD/PD conference in April.



Our anti-tau antibody, VY-7523, performed well in a recently completed single ascending dose study. There were no serious adverse events, and we saw dose proportional pharmacokinetics as well as a CSF to serum ratio of 0.3%, consistent with other monoclonal antibodies approved for the treatment of Alzheimer's disease. We initiated a multiple ascending dose study in Alzheimer's patients, and we expect initial tau PET data in the second half of 2026.



Finally, I want to point out that in Q4 2024, UCB's bepranemab demonstrated for the first time that an anti-tau antibody can impact tau accumulation in the human brain and that this may correlate with clinical benefit. It's important to acknowledge the study didn't meet its primary endpoint of the CDR sum of boxes, but our team walked out of the CTAD meeting feeling better about our anti-tau antibody than when we walked in. Looking forward, I think there are several opportunities this year for third-party data to continue to build excitement for tau. Merck has antibody data expected in mid-2025, and I look forward to seeing what we learn at AD/PD in April, AAIC in July and CTAD in the fall.



Okay. I promised I would keep it short. I just want to thank all of our employees for their hard work, especially pushing to achieve those end of year goals like getting the development candidate for the tau silencing program, working on the VY-7523 single ascending dose analyses and initiating the multiple ascending dose study. With that, we will open the call for questions. Operator?

[Operator Instructions] Our first question comes from Jack Allen with Baird.

Congratulations to the team on the progress. I guess, Al, maybe I'll start where you left off your opening remarks there. You mentioned some external readouts that could be interesting in the tau space. Any additional color you'd like to provide ahead of AD/PD as it relates to things people should be looking out for? And then I have a quick follow-up as well on your partnered programs.

Hi, Jack. So, at AD/PD, I hope to see data from bepranemab as they -- I believe they may be sharing their data on exposure PD relationships, so PK/PD, which they didn't have a chance to share at the CTAD meeting last year. And also, more data on what -- how much decrease in tau spreading do you need to see in order to see a clinically relevant effect? So that's one thing I'm going to be hoping to see.



The other thing might be more information about subgroups where greater efficacy can be seen. They started to talk about that a little bit at CTAD. For example, the effect of APOE4 carrier status as well as initial tau burden, and there may be more information on that as we learn more about which are the ideal patients to be treated with an anti-tau approach. We also note that other companies have started to share data with their anti-tau program. So, I know that, for example, Eisai has been sharing data on fluid-based biomarkers with their anti-tau that targets the MTBR region. And so -- and I don't know whether other companies may also be starting to share data as well. But there's a lot of interest in tau. There's also the tau silencing approaches that we know, for example, Biogen has. And I look forward to seeing any updates there might be on that. Toby, did you want to add anything to that?

Al, I think I certainly agree with your comments and sort of echo the comments on exposure response, both initial PK/PD, but also, in particular, the tau PET to clinical relationships. I do think on the subpopulations, for example, I'd like to see details on the tau -- low tau or APOE group they described previously. Curious what they think the pull-through of APOE is. Is it just that those individuals without an APOE allele have low tau? Or is there some other defining characteristic of that population.

That's very helpful color. And then more on the finance side, but I just wanted to ask, it seems like the 2 programs with Neurocrine on the gene therapy front are expected to enter the clinic or at least had INDs filed this year. Any additional color you can provide as it relates to thoughts on upcoming milestones from either Neurocrine or additional external partnerships that you've forged as well over the years?

Okay. Thanks for the question, Jack. This is Nate Jorgensen, the CFO. And so, what we have said is that there's $2.9 billion of developmental milestones. So, these are milestones that I think are not bio bucks like some companies report. But if you add all the bio bucks together, it's over $8 billion, as Al mentioned. So, there are some, I think, pretty meaningful milestones over the next few years that could help extend our cash runway past the mid-2027 guidance that we talk about externally.

Got it. And are those milestones at all accounted for in your guidance or are they additional upsides?

No, they're not. That's all upside to the mid-2027 cash runway guidance.

Our next question comes from Phil Nadeau with TD Cowen.

A couple from us. First, on the tau gene silencing IND. Can you give us some details about what needs to be completed before that IND can be filed next year?

Well, I mean, the main thing is that we have to complete the GLP tox study that we just started, and we need to be sure that we have a therapeutic window. As we said, our initial non-human primate study shows 50% to 73% knockdown, which is exactly in the range that we want. We just have to be sure that now we don't have any safety issues that allow for that dosing -- for the right dose to produce that level of silencing.

Perfect. And then a second question on the ALPL shuttle. Can you discuss where that could be most applicable? What indications are you thinking that would be most useful for? And any sense on when one of those candidates could advance?

Well, we're looking broadly across various diseases, various targets. This is a receptor that allows for BBB penetrant of one of our leading classes of capsids. And that capsid gets in across the CNS pretty broadly. So that leaves open a lot of different diseases, both spinal cord as well as cerebral cortex and even subcortical diseases. And then the kinds of drugs that we would -- we are thinking about transporting across the BBB would include proteins such as enzymes or antibodies. And also, we're starting to assess whether or not oligonucleotides can be transported as well. That leaves open a wide array of possibilities still. And so, beyond that, you might imagine what the best antibodies or enzymes for oligonucleotides we might be thinking about.



And look, as we noted with the antibody, that our anti-tau antibody, we get a CSF to serum ratio of 0.3%. So, literally 99.7% gets thrown away. It'd be great if we could get more of the antibody into the brain. Also, all ASOs currently are intrathecally administered for CNS diseases and that provides a burden to patients. It also produces a severe gradient in the CNS, which I think is limiting. So, there's a lot of opportunities. Todd or Toby, do you want to add anything to that?

I think you've captured most of it, Al. The other part of your question was moving things forward. And while we haven't shared any data, we're hoping to share more of our early work later this year.

Our next question comes from Pete Stavropoulos with Cantor Fitzgerald.

Congratulations on all the progress. First question I have, as we look forward for the tau silencing gene therapy, I'm wondering if you can talk about the key differences that we should expect from tau silencing versus targeting antibodies. And thinking about the Biogen data for 080, what's your view on the combination approach, meaning a tau silencing or knockdown with a monoclonal approach?

Thanks, Pete. This is Toby. Good to hear from you. So, I think fundamentally, our belief on the knockdown approach with 1706 is a couple of key points. One, as we've highlighted, we use our second-generation TRACER capsids. And so, it will be injected IV once, and that gives us a couple of clear advantages. I think one is that it allows for better biodistribution accessing the vasculature via ALPL. And so that's quite distinct from the grade you get with an intrathecal injection in the lumbar space with an ASO. So, we think that presents an opportunity for broader tau knockdown. And so that is, I think, quite an important point. And then in addition, the fact that it's IV in one time, we think prevent is clearly some benefits both for the patient and potentially for the health care system in terms of ease of uptake.



I think the other point I'd make is that, of course, the Biogen data, which comes out in, we think, in mid-Q3 of '26 will be an important inflection point. And of course, we have an IND for our program in 2026 as well. So, we think that's an important pairing. In terms of the antibody, I think fundamentally, the premise of the antibody is slightly different that you're trying to impede spread. And so, the idea there is that you would want to target that to a population in which tau is not yet spread. On the other hand, for the knockdown approach, if you look at the data that Biogen has shared, in that case, you can remove preexisting tau. And that was shown by tau PET, which is quite a remarkable observation. So, there may be some broader latitude. There may be space for sequencing of a beta amyloid therapy with an antibody and then a tau knockdown approach as well.

All right. And just one more question. The way that we viewed the SOD1 gene silencing program is that it would provide proof of concept for the capsid and its ability to cross the blood-brain barrier efficiently by looking at changes in NfL, which we would hope to have seen similar to tofersen. How are you thinking about establishing the human proof of concept? Which program will likely help you do so? Is there any agreement or expectation of a partner sharing some data or allowing you to once it's generated?

So Pete, this is Toby again. So, I think you've rightly keyed on the fact that the next opportunities to generate capsid POC really sit with the Friedreich's ataxia and/or the GBA program, which are partnered with Neurocrine. I'll just remind that the INDs for those programs are coming up this year. And I think what I'd say holistically is Neurocrine is running those programs, but we have a strong collaboration with them across the development teams. And in both cases, in Friedreich's, there's an opportunity for biomarker measurement in terms of frataxin levels. And then in GBA, there's an opportunity for biomarker measurement in terms of both enzyme GCase levels and substrate levels. So, in both cases, both programs offer the opportunity to understand that the capsids are working. Todd, anything to add?

I think you've got it, Toby.

Our next question comes from Lili Nsongo with Leerink.

Two questions from my side. So, the first one being on maybe comparing and contrasting the 2 approaches for tau. So preclinically so far, can you give us a little bit of perspective in terms of potential differences you've seen between the 2 approaches in preclinical studies?

Todd, do you want to touch that? In the preclinical study.

Sure. On the preclinical side, we've seen positive results from both. One aspect that we have discussed with the tau antibody program is that we do see differences targeting different epitopes. So, for example, our C-terminal targeted epitope works quite well. In terminal targeted antibodies that have failed in the clinic failed in our seeding models. This is a model where we inject Alzheimer's patients derived pathological tau into the brain of a mouse expressing human tau and look at the ability of a treatment to stop spreading. So, the in terminal failed antibodies didn't work. Our antibody does. Many other antibodies do not. I will say that the tau knockdown also shows efficacy in similar type models, and we aren't driven by specific epitope with a tau knockdown approach.



So, for the tau knockdown, we think we might be hitting a mechanism in 2 ways. One, we're reducing the amount of tau and subsequent pathological tau to spread cell to cell. And then we're also reducing the amount of tau on the recipient cell to then receive that pathological prion like material. So, we think that there are interesting similarities, but also some key differences in the fundamental mechanisms of those 2 approaches.

And as a follow-up, maybe could you provide a little bit of color in terms of the MAC study design. So, I know you haven't shared the whole design at this point. But maybe could you give us a little more in terms of how the study design has been impacted or informed by the results that we've seen with bepranemab?

Thank you. This is Toby. So, it's an excellent question. Fundamentally, maybe a couple of points. So, I think what we've learned about our antibody, including the preclinical work and up through the SAD is that the antibody to remind, it binds pathologic tau, which differentiates it from some other antibodies, including UCBs that we have appropriate PK. We've got good CSF penetration. So, in that context, we really think in the MAC study, we can drive to an effective determination of if we have a signal on tau PET, particularly in the context of what we've seen thus far. And that's important. I think in terms of the population, what we've learned from bepranemab is that the lower tau may be important in this context. And/or APOE status. And so, what we've shared so far is that we've adjusted the thinking based on these data to focus on the earlier MCI and AD populations, which are necessarily lower than tau.



I do think we'll need to be ready to respond to learnings as they continue to evolve. We've already highlighted that we're hoping to hear some discussions with AD/PD on the deeper side of these 2 subpopulations and/or exposure response. In addition, we've highlighted some other potential readouts in the field as well. And so, we'll certainly continue to monitor the field and adjust as we're able. I think that's what I'll end.

And I might want to add that in the case of Alzheimer's, the history of Alzheimer's clinical trials, sometimes multiple ascending dose studies, for example, with the anti-amyloid antibodies have provided some really meaningful information in terms of the effect on PET imaging. And here, in this case, wouldn't it be fair to say, Toby, that given the data with bepranemab, and particularly the safety and also our single ascending dose results, we're planning to push the dose pretty high, right? And at the highest dose, we're going to really take a look at how much we could impede the spread of tau by PET imaging and also look at fluid-based biomarkers. And so that's the plan. And as Toby says, we'll be monitoring the situation with all the other data coming out.

Our next question comes from Samantha Semenkow with Citi.

Just sort of a forward-looking question for me. I'm wondering if you can share anything about how plug-and-play you think your ALPL shuttle nonviral shuttle could be? And based on what you've learned so far in your discovery work, is it feasible that you could see shorter delivery times once you've worked through development, say, for each type of molecule you're looking to transport? Or is it more expected that, say, every enzyme or every oligonucleotide you're looking to transport would have its own set of unique challenges that you would need to optimize for? Any color you can share there would be very helpful.

Well, that's an interesting question. I mean, in some ways, the TFR-based shuttles have been sort of plug and play in the sense that it's worked for multiple different kinds of modalities, and we're starting to see data emerging that even beyond proteins that oligonucleotides may also be transported. Look, but even Denali, who are the leaders in TFRs are also looking at CD98, right, as another shuttle -- another shuttle vehicle. So why would that be? And I guess it's because every receptor is going to have its own safety, distribution and kinetics. And so it could be that for certain targets and certain diseases, it's more optimal to use one shuttle over another. I think time will tell. Right now, I think the field really only has 1 or really 2 shuttles that they can turn to.



So, everybody is using those. But as time goes on, we may be able to be more selective. So, it could be plug and play. But as we learn more about these various shuttles, we may start to tailor them to the right disease and the right target. That's how I see it now. Of course, it will be great to get more data to see whether we're right about that.

That's very helpful. And just as a follow-up, is there anything you can share about what that preclinical data set could look like sometime later this year for the nonviral shuttle?

Well, obviously, we're going to start by showing data in animals. So, it's obviously going to be in vivo data that's going to count the most. And we expect to be comparing to some of the TFR-based shuttles, comparing ALPL to start to get an idea of how different it is and whether there are certain advantages to ALPL. And we hope to be able to show more than one payload as well, see how plug and play it actually could be. So that's what we're hoping to show, and we're working hard on that, and we'll see whether we can get there.

I think, Al, you captured it. I mean, we remain very excited by that and hopefully look forward to sharing that later this year.

Our next question comes from Joon Lee with Truist Securities.

This is Mahdi on for Joon, and congrats on the progress. A couple on capsids for us as well. So given that for SOD1 ALS program, the recap Gen 2 is going to be the same for VY-1706. What are like the data points that gives you the confidence that the neurotox that you have seen is not related to the capsid?

Well, there's 2 reasons. One is the timing of the adverse events. So, in the case of the SOD1 program, in the initial time points in the several days after we inject, we do see the expected slight bump in liver function tests and in neurofilament levels. And that's exactly where other programs have shown those kinds of adverse events with IV-delivered AAV, and that's to be expected. But the capsid clears from the bloodstream within days. And then what we saw was with a delay of about 3 months, we saw then the neurofilament start to go back up and that was coincidence with some neurological adverse events observed at cage side. And then when you look histologically, you see evidence of neurodegeneration.



So, it's that delay, which is when expression has really come into play that we see the adverse events. And so, the timing is not right for capsids, much more consistent with expression of the payload. The other thing is that we use the exact same capsid with 4 other constructs with various promoters and various payloads, and we see no such adverse events in nonhuman primates, even at doses that are comparable to what we tested with SOD1 and even at the 2- to 3-month time point. So, it's both of those pieces of evidence that gave us a lot of confidence that it was the payload and not the capsid.

And maybe I'll just add -- sorry, when Al talks about what we're looking at, we're looking at histopathology. We're looking at NfL as a biomarker, and we're looking at sort of clinical signs. So, we see none of that with the capsid with these other payloads.

Our next question comes from Ry Forseth with Guggenheim Securities.

This is Ry from Debjit's team. Back to the third-party readouts, I'm sure the decision matrix is very large, but we wanted to get a grasp on the particular outcomes and how those outcomes would be actionable for your either preclinical gene therapy or 7523 MAD efforts. Maybe outline how adjustments could be made to these programs in response to the data?

Are you talking about the third-party data with antibodies?

Either silencing efforts or antibodies.

I see. Yes. So, I think one of the biggest pieces of data, again, is going to come in terms of antibodies from the J&J study, which will read out next year. And I say that mainly because it's a large study, well controlled and large and long enough to get a pretty good idea of whether or not, first of all, do we see an effect on the spread of tau with a different antibody. In other words, reproduce what UCB has shown, but then also get a much better handle on whether or not there's a clinically significant consequence to that impeding of tau spread, which we hope to see. So, I think that we'll have to wait for the J&J data, which I think is next year. In terms of the knockdown approaches, what's remarkable about the BIIB080 data, for me anyway, is the fact that you actually reduce the tau PET signal, which I didn't think would be possible because I always thought that the pathological tau was in neurofibrillary tangles, which is pretty much not going to be reduced by simply reducing the synthesis of new tau. But that's exactly what Biogen has shown. And what's intriguing is that they seem to see a pretty big clinical benefit of that.



Now the flaw there is that there was no control group in that study, but they've done a great job comparing it to natural history and to the control groups of other trials. And it does look like a large effect. I would think that with the passage of time, we will know even better whether or not that clinical effect is real and durable. And so those are the kinds of things I'm going to be looking at. Of course, we always want to know, answer the question, does epitope matter? In the case of the anti-amyloid antibodies, epitope did really matter. He's now speaking about the antibody programs, of course. And then the only other question is, is it important to be specific for pathological forms of tau versus all forms of tau. UCB was not specific for the pathological forms of tau. Ours is, and I believe many of the other antibodies are as well that are coming down the pipe. In the case of anti-amyloid antibodies, it was important to be specific for pathological forms of amyloid. So, we'll see if that also applies to the tau. Toby, Todd, what did I forget?

Maybe I'll highlight the Merck readout. It's a C-terminal epitope pathological antibody. It's a short study focused on biomarkers, but gives you an initial chance to answer the question, can a C-terminal epitope reduce at least total tau? And so that's an important concept as well.

And maybe just one more question from us. With the April AD/PD NHP gene therapy data, will you provide distributional data in terms of the percent of vector that goes to the brain relative to the liver? And if not, maybe you could frame for us Voyager's thinking on the utility of those kinds of measures?

So we -- let's see, at AD/PD, we will present on the tau knockdown program. We will describe data showing delivery, both in terms of the amount of vector and knockdown of pharmacology that we achieve in the brain and vector in peripheral tissues as well. So, it may not be in a percentage format, but we'll be talking about delivery to the important locations, and those include on-target and what we think of off-target tissues.

Our next question comes from Jay Olson with Oppenheimer.

We have a question about the new preclinical data for the tau silencing program to be presented at AD/PD. Can you talk about the target level of tau mRNA knockdown that you plan to achieve? And are there any particular brain regions that are more important for tau mRNA and protein reduction?

Well, I'll start with the last question and maybe, Todd, you can answer the first question. So, when I think of Alzheimer's disease, it's a cortical disease, the cerebral cortex. And it starts in the temporal lobe, but then it spreads to all the other major cortical regions. So, to me, the key region in the CNS that we need to look at for tau silencing is the cerebral cortex pretty broadly.

So Al mentioned earlier, the general range that we're targeting. We're seeing 50% to 73%. That's the general range of knockdown that we're looking at. We can get -- and we're showing that we get broad knockdown in some of these key regions, the cortex in particular, in the nonhuman primate that we think we need to achieve those kinds of knockdown and have an impact on the disease.

Great. And then can you comment on how you're thinking about indications for the tau silencing program? Would you start with Alzheimer's disease? Or are there other tauopathies that you would start with?

Well, assuming we don't partner it, we would -- look, our intention is to partner it. And so, it may be our partner decides which indications to go after. But you make a really good point here, which is that there are a variety of tauopathies that we could go after. The most well-known ones, of course, are TSP progressive supranuclear palsy. There's also frontotemporal dementia due to tau mutations in tau. And then there are a lot of other diseases as well, including potentially chronic traumatic encephalopathy. But I do think that if it stays in our hands, we will probably look very hard at Alzheimer's disease first, where it's -- a lot of us do believe that tau is a really important target for Alzheimer's disease and that -- and so much is known about the natural history of Alzheimer's disease and how to make measurements, both fluid-based as well as imaging measurements. So, we're going to take full advantage of all that knowledge and look hard at Alzheimer's disease first. That's our approach. What do you think, Toby?

I agree, Al. And maybe I'd sort of amplify that fundamentally, I mean, we've seen with the BIIB080 data how critical tau PET can be. And this is best worked out in Alzheimer's disease. Looking at some of the other tauopathies, use of tau PET and in some cases, flu biomarkers is less well settled. And so, one of our core tenets is that you go into diseases where you can get proof of concept around the biomarker tools relatively quickly. And in this case, for the tau, we think that really sits with Alzheimer's. I certainly think the other indications are interesting once you've shown that.

And maybe if I could ask one follow-up on ALPL. Have you done any experiments to compare the ALPL shuttle to other blood-brain barrier shuttles like transferrin receptor and see what the differences are?

We haven't shared that data, but we are in the process of comparing to TFR.

Our next question comes from Patrick Trucchio with H.C. Wainwright & Company.

Just a couple of follow-up questions on the VY-7523. Just the first, I'm wondering how the SAD data influenced selection of dose levels and frequency for the MAD study? And then separately, just given the competitive readouts in the anti-tau space and those that are upcoming, I'm wondering how the MAD study design may position VY-7523 for differentiation. And then separately, I'm wondering the cash runway extending into mid-2027, how will you balance investment in internal programs versus potential licensing or business development opportunities?

Maybe I'll start and then turn it over to Nate. So, I think in terms of the SAD data, just to reiterate, we saw dose proportionate PK an acceptable safety profile for a molecule in the single ascending dose study and we saw a CSF serum ratio of 0.3, which is quite consistent with approved molecules. And then I think I'd layer on top of that the sort of the observed safety in the broader bepranemab study didn't highlight any risk of ARIA. So really, what the SAD data gave us was the confidence in the MAD study to move forward with our planned doses and frankly, as Al highlighted earlier, to try to push those doses to levels where we can clearly test hypothesis to whether or not this antibody would impede the spread of tau pathologic tau.



In terms of differentiation from the competitors and what are we looking for. I think fundamentally, the study will be designed to look for tau PET signals. And so really, what we're trying to do, I think, frankly, is see where we compare to the other antibodies that are modulating tau PET signals, particularly UCB. And I think our simple aspiration here would be that we are at least as good as, if not better than that, given our specificity for pathologic tau.

Yes. May I add that, look, you hit the nail on the head in the sense that we do -- our aim is to try to differentiate and put the best foot forward with our antibody. And I would also say that we know the doses or we know the levels in the brain that were needed to impede the spread of tau in that animal model that Todd mentioned earlier, where we inject human pathological tau into P301S transgenic mice. And so, we want to be sure we exceed those -- the EC50, if you will, of that effect of impeding spread. And we're pretty confident we can get there with the doses that we have -- we will be choosing based on the single ascending dose studies that we did. And Nate, do you want to answer the second half of the question?

Yes. So, what I will say is that we understand the financing market out there and the cash runway to mid-2027 is important to us. We would be very thoughtful if we do any type of transaction or then which would shorten that. We would have to bring in some really interesting assets that potentially with the clinical -- near-term clinical catalysts. In terms of the other types of deals, those are something that Al always talks about that he's open to. Like there's, I think, opportunities out there for us to do additional BD deals to potentially bring in money or potentially take some of our pipeline assets and bring it to another partner, of course, for the right price.

Yes. I mean, look, if you look at our history, we've done anything from simple capsid licenses where the other company takes the capsid and runs with it versus partnerships on actual programs that we may have started here and then we may continue to do the work here reimbursed by the partner. So that just goes to show we're open to any kind of partnership that makes sense for the partner and for us. And I hope we continue to do those kinds of things.

Our next question comes from Yanan Zhu with Wedbush Securities.

The first question is on the MAD study. Was the initiation earlier than you had expected because I believe the original guidance was very broadly in 2025. And the second question is, I know that you compared the antibody approach versus the knockdown approach and compared the mechanism of action. Given the timing that do you think it's possible that the knockdown program could potentially catch up to be a more promising approach as compared to the antibody approach?

This is Toby. So, we did start the study a bit earlier than our guidance. And I think this just reflects the clinical team is up and going and executing. And then maybe I can help -- can you clarify your second question for me, please?

I know that the antibody approach is ahead of the knockdown approach. But given I think Al talked about this mechanism and also the Biogen and the data, do you think eventually the knockdown approach could potentially be more promising than the antibody approach for AD?

I mean, what I'd say here is we have 2 different approaches. And I think one, we have some early data from Biogen suggesting that knockdown could be quite clinically efficacious. Al highlighted that this was a natural history comparator and a comparison cohort in another study with an effect on CDR sum of boxes that was maybe 2 to 3x greater than sort of what is observed for beta amyloid therapy. So potentially a very large effect of the knockdown approach. So that's very exciting within the constraints of an intrathecally administered ASO. Antibody-based approaches, I think the data we have there is the UCB data, which is the main clinical data set. And that effect was slightly different. And so, I think the potential there is potentially different.



But fundamentally, to me, the question is more going to be of the appropriate sequencing and the appropriate risk benefit and what is appropriate to the disease stage of individuals as we've previously discussed. So really, there may be a point in time when you want to stop spread in a low tau population early in disease or there may be a population in time when someone has a broader spread of tau when you want to take a knockdown approach. Or you may learn the differential responses across different patient populations. So, I think fundamentally, Alzheimer's is complex and physicians and patients are going to need options in terms to fully treat this disease.

I would add that I believe we're still in the very early stages of learning about tau and how to best approach this target. If you look back at the anti-amyloid field, it took us decades to learn from each other and to get to the point where we finally had some drugs that were approved. We're just in the very beginning stages. We're only just now starting to see biological effects that are starting to be seen in humans. And I think there's going to be a lot to learn. And I think for right now, the prudent thing to do is to pursue both programs, and we will make data-driven decisions, internal data versus -- as well as external data. We will learn from those, and we will make data-driven decisions.

Our next question comes from Yanan Zhu with Wells Fargo Securities.

First, I wanted to ask about maybe a follow-up on the brain shuttle approach. I think, Al, you touched on this. What is the limitations of the TFR-based brain shuttles that you see at your vantage point and therefore, areas that you might hope to differentiate it in? And then I have follow-ups on the tau program.

Well, from what I see, there is evidence of hematologic adverse events, which is not surprising given the function of transferrin receptor. And even when you use ligands for different epitopes on the transferrin receptor you still see some adverse events because even if you don't block the transferrin receptor, you probably lead to its internalization, and therefore, you end up with loss of function type adverse events. If you look at the human genetics database, humans are not very well -- very tolerant of loss of function mutations in transferrin receptor. It's a pretty critical pathway. So now in contrast, humans seem to be pretty tolerant of loss of function mutations, or I should say, more tolerant of loss of function mutations of ALPL. So that could be an advantage. Of course, time will tell. But I would say that the -- that's one of the potential advantages, but there could be more as we learn more.



Let me add one more piece, Yanan, is that we sometimes forget about the fundamental properties of drugs such as PK and half-life. And so, I mentioned kinetics and distribution previously. And so, we'll be looking at those things, too. That's not to say that the TFR-based approaches are not promising. They are promising. In fact, so promising that that's why we're so excited about the idea of using a shuttle-based approach to improve delivery of other modalities beyond gene therapy.

On the tau antibody program, I was wondering, is it fair to say that the Merck antibody is even more similar to your antibody than the UCB antibody is? Have you compared in your tau animal model directly your antibody with -- against Merck's antibody and what might be the findings? And lastly, I was wondering about your thoughts on effector function for the tau antibody -- the general tau antibody approach. If you can comment on whether yours is effector or not? And why does that matter if it does?

I can take this. This is Todd. I do think you're right. In a lot of ways, the Merck antibody is more similar to C-terminal antibody. We have not shared any data comparing the C-terminal Merck antibody to ours. So, we don't have that to share. On the effector function, it's an interesting question. And we've gone with effectively an IgG4, human IgG4 for our antibody. Evidence to date suggests that to have an effect in these spreading models, and it looks like perhaps with bepranemab in the clinic that you don't need effector function and eliminating it eliminates or reduces some risk around neuroinflammation, we think. So, we've gone with an hIgG4 for our antibody.

Our next question comes from Sumant Kulkarni with Canaccord Genuity.

Apologies if this was asked before. And Al, I know you gave a quick history lesson there on anti-amyloid products for Alzheimer's. But what are the key results you need to see or learnings from external programs that might lead you to making a firm go-no-go decision on your anti-tau antibody program? Or would you need to see internal results in order to make that decision? And what might those internal results look like?

Yes. I mean that's a tough question to answer. What I'm looking for, Sumant, is really how much of an effect on tau spreading do you have to see to see a minimally clinically significant effect. So, in the case of the anti-amyloid antibodies. And by the way, you can learn from different antibodies against different epitopes. But the key question is how much of an effect on amyloid PET imaging did you need to see to see an effect on CDR sum of boxes, which is the required endpoint for approval. And you can actually draw a graph with all the different antibodies and see that if you don't get to a certain level of amyloid lowering, you don't get a big enough effect on CDR sum of boxes. So that's precisely the kind of thing I think I would like to see with the anti-tau antibodies. And I would say that, anything less than a roughly 30% effect on CDR sum of boxes, I would say probably not clinically meaningful. So that's what I'm looking for is what is the effect on tau PET imaging that gives us a 30% effect on CDR sum of boxes.



Once we know that for sure, then we can look at our antibody and say, okay, it met that hurdle or not. I think it's going to take a little bit more time to get that. It took more than a decade to generate that data for the anti-tau -- I mean, sorry, the anti-amyloid antibodies and various companies sharing their data. So -- and like I said, I think we're in the early stages still of the tau-directed antibodies.

I would now like to turn the call back over to Dr. Al Sandrock for any closing remarks.

Thank you, everyone, for joining us today, and we look forward to speaking with you again soon.

Thank you. This concludes the conference. Thank you for your participation. You may now disconnect.