Denali Therapeutics Inc. ($DNLI)

Great. Thanks, everybody, for continuing on here. It's my pleasure to be moderating a chat with Ryan Watts, Founder and CEO of Denali. I'm sure most folks know the story decently well. But maybe, Ryan, you can just give us a couple of minutes to sort of set the stage on 2026 is a big year for Denali with Hunter and the pipeline, and then we'll do Q&A. So thanks again.

Sounds great, Paul. Great to be here again. I was trying to count how many CNS days this is for me, but probably...

6 or 7. Do you know the first one was not on video. Well, no, I've done 7. The first one for you was probably 6 years ago, audio-only 5 days into COVID.

Yes. I remember that.

Crazy. So thank you. Appreciate it.

It's good to be back. And I think that might have even been a panel, if I recall, back...

Yes, that was interesting. We can talk about that another time. That was interesting.

I mean, 2026 is set up to be a very important year for Denali. I mean, obviously, we're at the very final stages of our first approval for tividenofusp alfa in Hunter syndrome. I think not only does this program have its own opportunity in a very significant area where there hasn't been an approved medicine for 20 years. And of course, in our case, engineering brain delivery has allowed us to take the next step and treat the brain for these patients. But also, we kicked off the year with new data on our Sanfilippo program, very similar to the Hunter program. We'll have new data on our PTV: Progranulin program, which is another transport vehicle-enabled medicine. We're going to read out our LRRK2 program. And then we've started three new clinical trials. Well, two are starting and one we're just about to file. So one in Alzheimer's, another in Pompe and then another in Alzheimer's. So obviously, a very exciting year. I think last year, I think as we shared in the last time we spoke a month or so ago, was the year of harvesting a lot of science for Denali. We had a couple of basic biology papers on Alzheimer's. We had a science paper on our ATV:Abeta approach, but we ended the year and began this year with our New England Journal publication on the Tivi program, on the Hunter program. And so what we've seen is a shift at Denali. Starting 2 years ago, we've harvested the fruits of our scientific efforts, but now we're applying the transport vehicle to many other programs. The goal is to accelerate and to expand the portfolio using the transport vehicle also at a time where TfR-enabled programs are highly competitive. So it's an exciting time in the field for sure and looking forward to diving into details, Paul.

Yes. Awesome. Well, I think just to start, I mean, everyone wants to know how you feel ahead of the PDUFA, right? We're in probably one of the weirdest years for the FDA that I can remember. And you're right in the center of rare disease, an area where I think people are debating to what degree have some of these negative decisions been product-specific versus a changing of the goalpost. You've expressed confidence, talked about being in labeling discussions. Like anything else you can kind of convey to us here about the 310 PDUFA?

Yes. I agree that the last year has been unusual, especially in rare disease, accelerated approval, biomarkers that are reasonably likely to predict clinical benefit. And I think we're confident. We have had a fantastic engagement with the FDA over the last 9 months or so. And I think our confidence comes from the robustness of our data, our clinical data, our biomarker data that we believe it will translate and is translating to clinical benefit for patients. And so I think having a rigorous bar has been important. Obviously, it's -- and we're going to get into, I'm sure, more details, it's definitely a complicated time with the FDA.

Yes. Yes. Okay. I mean I think you said in November, you were in labeling discussions, which was a while ago. I hear labeling discussions, and I think that's super bullish, but it's also been quite a long time. Like how should I make sense of that from the outside looking in?

I think it's important to set the context of our review cycle. So basically, what happened for us is that our original PDUFA date was January 5, as you know. And leading into this sort of late cycle meeting, we were made aware and engaged with the FDA on a miscalculation of a molecular weight of Tivi, which is so unusual, solved it very quickly. But I think by statute alone, it resulted in a major amendment and a shift of the PDUFA date to April 5. So I think what the FDA did is they did their best to stay on track to the January 5. But inevitably, there isn't the time pressure that you have of January 5. And so I think we've been really in serious label discussions in the last month. Those early days, we basically took a bit longer for the CMC review, which is now completed. And basically, the right timeline kind of moved into 2026 is the best way to describe it. Maybe not as quick a turnaround on the back and forth, but we're -- now it's -- like you said, we're here. We're at the final stage.

Right. So I mean, the value proposition of this drug, as I see it, is, hey, you can maintain the same peripheral disease control you have in Elaprase, but treat the neurological part of the disease. Do you envision a label where it sort of indicates this for all Hunter as an Elaprase alternative? Or is it going to be like more neuro-centric?

Yes, it's a good question. I think we're too close to the finish line to go into detail on anything related to the label. And I think your -- the way you articulate it is right that what this medicine does is it does essentially what Elaprase or standard of care has done historically, but crosses the blood-brain barrier. And one of the things that we've done is we've really pushed dose. We want to maximize efficacy. We want to maximize efficacy in the periphery and in the brain. And I think inevitably, these patients, even if they're like attenuated, which at sort of diagnosis represents maybe 30% of the patients, they eventually develop neurological disease like hearing deficits and other challenges. So obviously, our ultimate goal is to treat all patients, but let's stay tuned. We're at the end here. I should say we're at the beginning, hopefully.

Okay. Yes, yes. Okay. All right. And then assuming you do get approval, like how are you setting expectations and what a good outcome would be for the first year of the launch?

I think that realistically, this is a switch market. Obviously, newly diagnosed patients is going to be key. The challenge is, in general, as you know, is reimbursement in the rare disease. So the way that we're setting expectation is sort of the S-shaped curve. We expect minimal revenue this year, but our focus isn't on that. Our focus is really patient starts, like how many patients can we get on drug out of the gates. That's going to be key.

Any thoughts on pricing and premium pricing potentially to Elaprase?

Right. I mean, again, it's too close to the beginning to discuss pricing in detail. But I think in general, we see this as not just an incremental leap from Elaprase, but a very significant leap of being able to cross the blood-brain barrier and access the central nervous system while not only adequately but maybe really robustly treating the periphery.

When you talk to clinicians in this space, do they think all patients have some element of neurological disease, whether it's a lot or even just a little? Or is it really a view that this is two different diseases peripheral-only in peripheral-plus-CNS.

Yes. I think that there's a spectrum of the disease, and there's a spectrum of how physicians view this. But in general, I think no one will argue with you that even like really attenuated patients will develop, for example, hearing loss and hearing deficiency, which is neurological, I think, by definition, there's both peripheral and central roles in hearing, but in general, believed to be behind the blood-brain-barrier. And so I think it's just a matter of time. So when I say spectrum, those individuals that carry severe mutations will experience neurological symptoms early on, age 2, 3, 4. Again, that represents 70% of patients. What's interesting, if you look at those that are on drug, attenuate patients obviously live much longer, but they eventually develop some type of neurological disease.

Yes. Okay. Okay. So I guess, does that argue for switching that is like a little bit slower?

Yes, it's a good question. I mean I think there -- obviously, the patients that are eagerly awaiting are those with neurologic disease. There's no question about that. And I think beyond that, it's really educating the community and asking what else is there, what represents some type of deficit that could benefit from Tivi.

Yes. Okay. So if Tivi gets accelerated approval, how fast can you then run with Sanfilippo?

Yes. So Sanfilippo for us, we just presented data for the first 8 patients. I think what's most interesting about the Sanfilippo data is that those 8 patients, we really explored the dosing regimen, both dose level and dose frequency. And I think actually, the last time you and I met, we were a day or 2 away from presenting that data at World. But now that it's presented, I can go into some detail. And I think it's actually -- was really, really insightful. So I think one of the things we learned is that weekly dosing is much better tolerated than every-other-week dosing. And part of that is infusion-related reactions are all related to tolerance, right? And invariably, essentially almost every patient developed antidrug, antibodies to enzyme replacement therapies, some higher than others. But in general, these proteins are viewed as foreign. And what we learned in that particular study is that as we shifted even at the high dose from every other week back to weekly, weekly was much better tolerated. And so we had a dose escalation. And now we've nailed down the -- what we call key efficacy cohorts, which are fully enrolled. That data will -- that data cut will be September of this year. And the planned BLA filing will be 2027 with approval in 2027. Actually, the rate-limiting step for filing that medicine, part of it will be the data, but the real rate-limiting step is that we plan to manufacture that ourselves in our own facility. And I think as a result, we got to nail that down. We hope to be part of the FDA precheck effort where we can really engage with the FDA on the facility as it's a new facility, that could help us accelerate. But I think the timing will be driven largely by manufacturing for that medicine.

Okay. But I guess the thought is you file early in '27 then, just given the timelines you outlined?

Yes. I think probably the safest thing to say is first half of '27, second half approval, something along those lines.

I see. Okay. And we'll move on to the pipeline in a second. But with Hunter, I mean, I think it's pretty easy to size the global market just given the maturity of Elaprase. With Sanfilippo, like maybe talk about what the thought is on the likely prevalence and subsequently, the likely diagnosis rate or pool of diagnosed patients right now?

I think as you appreciate with Sanfilippo, there's no standard of care. So you're exactly right. It's much easier to size the market, understand the regionality of Hunter, where the primary cells, 1/3 in the U.S., 1/3 in Europe and 1/3 elsewhere. With Sanfilippo, there's less known. We think it's similar in size with similar distribution is probably the best way to look at it, but obviously, no standard of care.

Yes. Okay. And then you're doing Pompe, right? So what are sort of the similarities and differences between these programs in Pompe, like I guess you're targeting muscle there, so that is different. Like to what degree does your work here derisk that? To what degree is that kind of testing a new avenue for this technology?

That's a great -- that's a fantastic question. So Pompe also similar to Hunter and many of these genetic diseases comes in multiple forms based on the genetic -- the underlying genetic lesion or the mutation. In Pompe, you have IOPD and LOPD. Our ultimate goal is to compete in LOPD. And as you point out, the transferrin receptor has -- can enhance delivery to muscle just like it can to brain. What's unique about GAA in Pompe is that you're using this mannose-6-phosphate receptor for biodistribution. That appears to be limited in skeletal muscles. But by using transferrin receptor, we can pretty significantly enhance that delivery. So the ultimate goal is really head-to-head. So it's a different dynamic here with LOPD. But I'd say sort of the other consideration is IOPD, which has a central nervous system component to it, which is not treated by any of the standard of care. And so we'll pursue those, obviously, in parallel. And so it's unique. I think the way to like summarize Hunter, cross the blood-brain barrier, almost everyone has some form of neurological disease. Obviously, there are those who are eagerly awaiting. Sanfilippo is largely a neurodegeneration in fact. It does have -- we saw normalization in liver volume, which was interesting. You never think about the peripheral effects of Sanfilippo. And then Pompe, largely muscle. So it is that new dynamic of what the transferrin receptor can do, but there's still the CNS component in the more severe cases.

Yes. Okay. Great. And Pompe, first proof-of-concept timing?

Yes. So I think if we look at the three programs that we're starting clinical studies this year, the Pompe and two Alzheimer's programs, just in general, we're stating 2027. We have a lot on our plate in 2026 with the potential launch, obviously, the new data in Sanfilippo, progranulin and LRRK2. But in general, we're stating 2027 for these three programs. That being said, we got to move as quickly as we can on all three. And they all have some type of biomarker data, if it's some type of glucose metabolite, if it's Hex3 or muscle-specific biomarker, that's what we look at in Pompe. For Alzheimer's disease, if it's an Abeta molecule, it's targeting plaque. If it's tau, it's targeting tau protein and tau PET. And so in general, I'd just state that those programs are 2027 readouts, and that represents the real expansion for vehicle platform.

Yes. Yes, makes sense. Maybe for the Abeta brain shuttle, we've got the data from trontinemab that looks great. What do you see as the one to two key points of differentiation for what you're doing?

Yes. I think you're right that -- if you look at all of transferrin receptor-enabled programs, you have those that are categorically muscle and then you have those that are CNS. And the two data sets that convince me that this is going to be a new class of medicines are our Hunter data, obviously, recently published and waiting for the first approval with a transferrin receptor molecule in accessing the brain and the trontinemab data. And what you see with trontinemab, which is really fascinating is basically about 1/5 the dose, you have 3x faster plaque reduction. and it's obviously very robust. Now we don't know if speed of plaque reduction will correlate to enhanced clinical benefit. Probably the strongest driver of clinical benefit for Alzheimer's, in my opinion, is timing of intervention. So getting there as early as possible because you have two decades of plaque that cause synaptic loss and cell loss. So I don't know. I mean that's obviously one thing that the brain shuttle and trontinemab has enabled is a faster plaque progression. And number two, significantly lower ARIA rates. That being said, there is definitely room for improvement. I think tolerability, I think one of the challenges that is obviously monitorable and reversible is any hematological findings. That was part of our engineering with the -- what we call the cisLALA mutation that allows us to engage transferrin receptor, get across the blood-brain barrier, but not engage transferrin receptor and the immune system on, let's say, reticulocytes. And then I think the other component is that if you look at all of these brain shuttles, they're fusion proteins. They're essentially fusing Fabs or single chain Fvs to these full-length IgGs, and they struggle with stability. So if you look at 24 hours or 48 hours after dosing, you don't maintain the brain uptake, but also we think that also results in some risk around immunogenicity. And that's where we think the room for improvement is, is really tolerability around immunogenicity and potentially hematological outcomes.

Yes. Okay. How are you thinking about the right amyloid subtype or subtypes to target to maximize efficacy?

Yes. I think for us, we'd rather not reinvent what's already been discovered clinically. We're already inventing with the transport vehicle. So the goal is to look at all the profiles of successes and failures. And I think what you can determine is that those molecules that bound monomeric Abeta and did not bind aggregated Abeta failed in clinical studies, but those that found aggregated Abeta. And what I mean by aggregated, I mean plaque. I mean oligomers. I mean, basically, the challenge you have is if you're picking a single aggregate or single form of oligomer and you're not eliminating plaque, you're basically at very high risk because only correlation today that is plaque reduction and clinical benefit. So our Abeta arms target aggregated Abeta, both plaque and oligomeric and are less preferring to monomeric Abeta.

Yes. Okay. Okay. Makes a lot of sense. And the first study there from a proof-of-concept perspective, like you'll be able to answer some of the questions on amyloid anemia and ARIA, you think all in that first readout?

Well, it's a great question because if you think of the, let's call them like the 7th generation Abeta molecules, it's taken a long time for naked Abeta molecules to show some clinical readouts, right? And part of that is just the time it takes to see plaque reduction with standard antibodies. Now with trontinemab, that happens more rapidly. And so we hope to be able to get a quicker signal. But I think the order of assessment is essentially hematological plaque reduction and then you need larger patient numbers to look at rates of ARIA. I mean the ARIA, it's about statistics, right? It's about how many patients you treat that are APOE4 positive, which have higher risk of ARIA. And so these data will come in sort of in tranches as we first test the hypothesis around the cisLALA on tolerability and then go into plaque and then ultimately to ARIA.

Yes. Okay. You're also doing a tau conjugate as well. From a technical like feasibility perspective, I guess we haven't seen someone use transferrin to deliver an oligo into neurons yet, but we've seen it in muscle. Does the muscle space kind of derisk this along with your preclinical data? Or like how would you characterize the risk profile of this program translating from like preclinical animals to humans?

Yes. I mean this was -- to me, it was one of the most interesting technologies and surprising in the last decade is the ability to get oligonucleotides across the blood-brain barrier. Because think of your -- the path to a neuron or an astrocyte or microglia for an oligonucleotide, not only does that have to get across the blood vessel, but it's then got to get into those cells and knock down gene expression. So our preclinical data, we have a humanized mouse model that has a humanized apical domain of transferrin receptor has been pretty accurate at predicting clinical dose and clinical efficacy for the enzymes and for the antibodies. We expect the same for the oligonucleotides. I wouldn't say that the muscle delivery derisks it in any way because a lot of the muscle delivery technologies are using either bivalent or high affinity Fabs. So you obviously are testing this idea of getting the oligonucleotide into the brain, but the animal data are very convincing that this is doable.

Okay. Makes sense. Do you have a view on the BIIB080 study coming up and the probability of success of that trial?

So I think the most interesting thing about BIIB080 is it taught us that not only can you remove, you can lower tau in cerebrospinal fluid, but you also by tau PET, can see a reduction in the PET signal. What that actually means is that neurofibrillary tangles are in some equilibrium, which is not a known factor. I mean the plaque, we get it that microglia will engulf the plaque, but what we don't really know is can you remove existing neuropathology in this case, neurofibrillary tangles. And that's what the original data showed. This new data set, you -- I would read your note on what your opinion is, but 416 patients, I think it is actually 4 arms, 1 placebo and 3 dose arms. The lowest dose, I think, has enrolled fewer patients than the 2 higher doses. And I'm guessing they're going to do some type of pooled analysis. The great outcome would be directionally positive clinical benefit, some type of cognitive benefit. I think my only cautionary note is that those are not huge numbers when you're looking at things like CDR sum of boxes, and we just have to make sure that we take that with a grain of salt.

Yes. Yes. Okay. Okay. Makes sense. Maybe in the last couple of minutes here, I don't know if the stat is true, Ryan, but someone from another company told me that there are 16 brain shuttles in development in biotech. That obviously includes a lot that are probably named preclinical, early preclinical. But on the one hand, that's extraordinarily validating to what you're doing. On the other hand, right, there kind of becomes this whole question of like, is this technology going to become commoditized? Is it going to become like GalNAc in the liver? Like what would you say to that? Like why does a company like Denali deserve like a real premium for its transferrin technology over everybody else?

No, I agree that it's gratifying. I remember running the first experiments on transferrin receptor prior to founding Denali. But the idea of using transferrin receptor to get medicines in the brain was proposed in the 1980s. It's just taken a long time to industrialize it. And the question is why? Why did it take like multiple decades to get the first clinical data in Hunter syndrome or in Alzheimer's disease? And what -- and the reason for that is it's not just intuitive. It's not like you just take an antibody and it works or you take a Fab and it works. And so I think the winner will come down to those that really understand the engineering and how that engineering translates to the best outcome. I'll just give an example. I think most enzymes, you cannot deliver with just a Fab fusion. The Fc is a huge differentiator relative to a Fab for iduronate-2-sulfatase or SGSH or GAA, and it brings a whole another area of benefit in biology and around building tolerance and PK. And so that's ultimately what's going to win. I don't know if there are 60 companies. I have no doubt that there are 60 medicines using transferrin receptor, but that's what happens. In biotech, it's really simple. If something works, everyone starts doing it. And then the question is, can you get there first? Or can you be the best?

Yes. Okay. Great. Anything else you'd like to highlight in the last couple of minutes here?

I think it's -- again, it's an exciting year. We obviously are waiting for our first approval. Waiting is probably -- that's a passive term, but we're very actively and constructive with the FDA and fully prepared to launch. We have the teams hired and ready to go. Looking forward to filing the next molecule. And then we want to lead the area of blood-brain barrier engineering. And I think it's a very exciting time, and it is gratifying to see everyone latching on to these transferrin receptor approaches.

Right. Okay. And the new PDUFA is what date? The 5th?

April 5.

April 5. Okay. That's Easter.

I know it's Easter Sunday.

And Good Friday is the Friday before. Okay. We'll see when you get the notice. Okay. Best of luck. Thank you, Ryan. Appreciate it.