Denali Therapeutics Inc. (DNLI) November 11, 2026 Conference Transcript

Great. Thanks. Sorry, man. I got my drink station. Yeah, thanks very much, everybody. Happy to be moderating this chat with Joe Lukak, Chief Scientific Officer of Denali Therapeutics. I think we'll kick off and maybe hit a couple questions on the EVLA launch and how that's all going, and then I think we'll use this as an opportunity to maybe do a deeper dive into each pipeline program and preview some of the data readouts next year. But yeah, thank you, Joe. So maybe to kick it off, like, do you want to just level set for people and review the updates you gave on Avalaya in the past quarter and, like, overall, how's that launch gone relative to your guys' expectations?

Yeah. Thanks, Paul. Thanks for having us, by the way. Very excited about the Avalaya launch so far. Seeing a lot of strong interest in the product and really what we're hearing from both patient families and KOLs is that the drug is really making an impact for patients, which is fantastic. And it really takes patients and physicians and payers working together in order to have a strong launch like this, and that's exactly what we're seeing. So we have about 70% of commercial lives covered with 20 states covering on Medicaid coverage at this point. So, very encouraging start to date, and it's led to $7.8 million in commercial revenue in just the second quarter after launch, which is fantastic.

Yeah, that's great. You know, I think with Hunter, there's always been this kind of question around what percent of patients have meaningful CNS symptoms? Like, how do you define that? Like, with now being in the market for a bit, like, what are you seeing in terms of, like, you know, the way physicians are segmenting patients and, like, sense of urgency to treat the L-Aprase-treated population with this medicine?

Yeah, we're seeing a strong level of interest. And I think, you know, our feeling is the majority of patients have some sort of neuropathic symptoms. And, you know, of course, we're targeting patients within our label. But really, what we're hearing from KOLs is that it's making an impact across a broad patient population.

Okay. Yeah, understanding that your label doesn't cover everybody on Eloprace, but for the overlapping population there, at this point, what do you think is a realistic goal in terms of penetrating that population? Do you think it's a goal to switch most patients eventually over time? What's the thought process?

Yeah, I mean, our goal is to really, you know, to switch 100% of eligible patients. So in the U.S., there's about 500 patients with Hunter's disease, but 375 of those are eligible under our label. And our eventual goal is to really pivot all of those patients over to Ablea.

Okay. And so you guys have been guiding each quarter for the next quarter. Maybe just talk about how you guys have set expectations for the next quarter and just more broadly, what metrics or things might you share or might you not share, I guess, going into 2027? Do you have a sense of that?

Yeah, so in our first quarter, we had a pretty big leap from our first quarter, our second quarter, from 3.6 million to 7.8 million. That was well ahead of what we had predicted, so we had guided around 10 to 12 and significantly beat that. And the drivers of that were really faster expansion of payer coverage, and that was fantastic. But then also a number of patients from our clinical trials switching over more quickly than we anticipated, which was great to see. So it was really rapid growth in the past quarter. But what we've described many times is that we expect somewhat of an S-shaped curve here, where we'll see a steep ramp up at some point. And I think what we saw was based on those factors I mentioned, that that S-shaped curve was pushed in a bit faster than we had thought. So we're still expecting growth, but in a more measured way moving forward. So we've guided for the fourth quarter to be between $30 and $32 million in commercial revenue. We think that that's going to be a more predictable number.

Okay. Okay. Great. You're also set to, I think, complete the confirmatory trial next year. You know, I think from our perspective, like, I think it's pretty clear of Leah works, given the heparin sulfate data and the neurofilament data, But there's always this risk of going from an open-label study to placebo-controlled on clinical and functional endpoints. Maybe talk about, like, how that study is designed and, like, your level of confidence that in the absence of placebo-controlled data, you've powered that, like, somewhat conservatively.

Yeah, great question. You know, we have a lot of confidence in the outcome of Compass, which we'll get data next year and then hopefully file on it in 2028. And the reason why is we've already had 47 patients of data as part of our phase one and phase two programs, so a really high number of patients. And what we're able to do, you know, in that first study is not only show a biomarker readout, which was the basis for approval, but a lot of other behavioral and cognitive endpoints as well. And so we were able to use that data to power Compass in a way that we felt like we had a lot of confidence in, and when we do go and file, we'll file based on the entirety of the data. So I think we have a high level of confidence we're going to be able to reproduce those results in a placebo-controlled fashion.

Maybe, like, more specifically on the actual, like, primary outcome measure for Compass, like, what is your, like, expectation for the placebo arm and, like, how that might all play out?

Yeah, I mean, I think that we have a good idea, we feel, of how that placebo is going to behave based on natural history data that's been out there, both from what we've used in our previous powering study. So, I think that, again, I think we can make some good predictions. Now, with a rare disease, you know, powering, getting enough patients to run that trial, you know, in a convincing fashion is always a challenge. But we've tried to really, you know, run a robust study, not just to understand the cognitive readouts, but then we also have a second, you know, part B of that study to look at, you know, to look at other aspects, somatic aspects of disease. And so the hope is really that Compass will open us up to really the full patient population in Hector's disease, including adults.

Okay. And so that's reading out in the first half of 2028?

Yeah, it's reading out in 2027, and we'll file based on it in 2028.

Okay, so the data we'll actually get next year.

28. Data will read out in 28.

Okay. Okay. Okay. Sounds good. And then for...

Complete in 27. Sorry.

That's what I thought. Okay. But no worries. Thanks for clarifying. As it relates to San Filippo, maybe talk about some of the initial data you generated there and what do you think is the necessary package there to garner accelerated approval?

Yeah, so we've had a lot of interactions with the FDA around our San Filippo program and spent a lot of work to align on heparin sulfate reduction as a basis of accelerated approval for that program. And so, you know, that's really exciting, I think, leveraging some of what we learned with Avala. And so what we've seen thus far with that program, very encouraging results. So even in our dose-finding cohort, we're seeing greater than 80% heparin sulfate reduction with many of those patients normalizing on treatment. So very encouraging data so far.

Do you expect to share neurofilament data at some point? And how important is neurofilament or not to sort of a corroborating, as a corroborating component to like a filing there?

Yeah, so I mean, I think that the basis of the filing will, of course, be heparin sulfate. We do feel like neurofilament as a biomarker can be very useful, right? And I think we showed this as part of our AVLEA program, where really, you know, an impact on some of the behavioral endpoints correlated well with a reduction in neurofilament. And so I think what neurofilament tells you is not just that you're, you know, impacting the lysosomal biology, but you're having downstream impacts on disease as well. And so that's something we'll plan on looking at as part of our San Filippo program as well.

Okay. Do you know the timing of that next data update there?

Uh, we, so the, in terms of our, so we're, we're planning on a filing based on a cohort of 20 patients. Those patients have completed the study already. Um, and we plan to present a detailed data in the world conference, uh, early next year. Um, and based on that, we'll plan to have a filing at some point during 2027.

Okay. Okay. Great. Um, you know, as you expand the portfolio, uh, in lifestyle So in storage disorders, you have this Pompeii program that I think is going to have some initial data there. Maybe talk about the thesis there, and I guess the secondary question to that is we have ERTs for Pompeii. What's the evidence out there that those are inadequately controlling the disease or not getting sufficient target engagement in tissues like muscle and diaphragm?

Yeah. So, yes, existing therapies, a number of existing therapies for Pompe disease, all of them are, you know, the first generation was a standard enzyme. The second generation molecules were an enzyme with additional M6P, so that's the receptor that usually enables enzymes to be taken up. If you talk to the community, I think there's a broad sense that the impact on disease is inadequate, and there's still unmet medical need there in muscle. And I think part of the challenge in muscle is that it has very low M6P receptors, so the uptake of those enzymes is quite difficult. Actually, Pompei drugs are some of the highest-dosed ERTs out there, and we think it's partially for that reason. And what we've been able to show, you know, and even with the second generation molecules, they were not able to show superiority over the first generation molecules. So I think it's kind of consistent with that biology. What we've been able to show, at least pre-clinically, is that transferrin, you know, we focused on its ability to deliver to the CNS, but it also very effectively delivers to muscle tissue. And in preclinical models, we've shown pretty clear superiority of a TFR, an RTFR-enabled product, DNL952, pardon me, over some of these first and second generation enzymes. So very excited about the potential to replicate that and really better deliver not just to muscle as a tissue, but to really the lysosomes and the right components of the cell within those muscle cells in order to have an improved beneficial clinical effect.

In the absence of doing muscle biopsies, which I think you're not doing, how might you prove out differentiation in the proof of concept readout here?

Yeah, good question. So our first study, our phase one, ongoing phase one study in Pompei patients, I think it's important to note that these patients are already on treatment. So what we're doing is somewhat of a switching strategy to start, and then looking at impact on biomarkers for patients who are already on one of the current enzyme treatments. So we'll be able to see a little bit of a differential impact there, and we're looking at two specific biomarkers to start with. Urine Glyc4, and that's really a readout of glycogen, which is the substrate of the GA enzyme for Pompe disease.

And that will be, like, significantly elevated in these patients?

So, you know, there is a challenge with these patients that for some patients who are on standard of care, their GLIK4 is normalized. Some patients are not, right? And so I think it's, again, evidence that you're not having complete impact on the pathway with those existing medicines. But it's something that we're going to have to monitor as course of the study. The second thing we're going to look at is serum creatine kinase, and so that will be a direct readout of muscle health.

Yeah. Okay. Okay. Makes sense. And so we'll get some data on this next year. Is that right? Okay. Great. Maybe we'll try to hit on everything else. So I think you're also going to have some data next year for your A-Beta brain shuttle. Maybe talk about how this program is differentiated from some of the other shuttles in the clinic.

Yeah, so we're very excited about our Alzheimer's programs, including our anti-A-Beta program. Of course, we're not the only player in that field, and there are a number of competitors, including Trontimab from Roche. I think the positive side of that is what Tronti has shown is very encouraging data in terms of plaque clearance, and that's an exciting molecule as well, but not without its liabilities. And so maybe to start, I'll talk about how our platform differs from those like Roche's Brain Shuttle. And I think a key element there is that we don't have any appended sequences on the molecule. So Roche's Brain Shuttle is characterized by an appended domain that binds TFR, where our transport vehicle platform is based on protein engineering directly in the FC domain of an antibody in order to introduce low-affinity TFR binding. We think that's going to be critical. We think that appended sequence increases the risk of immunogenicity, and this is an issue that Tronti has had a bit of in their clinical studies, at least what they've reported, as well as it can potentially be clipped off, which would limit then the ability for brain uptake. And the last differentiator is we've carefully engineered our anti-beta molecule to have effector function. So antibody effector function seems to be critical for efficacy of these molecules. For the plaque clearance part, you need that antibody to be active. But combining effector function with TFR binding causes risk of anemia. This is something that I think has also had potential with other molecules. We have shown preclinically, at least we haven't talked about or shared any of the clinical data yet, but that we do not preclinically see anemia with this molecule because our molecule is designed to bind either transferrin receptor or FC gamma receptor for effector function, but not both at the same time. So we feel like there's going to be some important distinctions in terms of safety there in particular where we're going to be able to differentiate from some of the competition in the A-beta space.

Yeah. Okay. That makes sense. And do you feel like this is the kind of thing where, like, in a short initial data readout, you can answer the questions related to, like, robustness of plaque lowering, ARIA, things like that? Like, when might we get those insights?

Yeah, so we're starting our phase one study with DNL921, which is our anti-beta molecule in healthy volunteers. And just in healthy volunteers, you'll be able to get to some of these questions around potential for immunogenicity as well as anemia. Will you be looking at CNS exposure? We will be looking at exposure broadly. I think we have an understanding enough of how our platform behaves in human from Avalet and other products. It's also worth noting, and I think this applies to all TFR-enabled molecules, is measurement of CSF antibody is actually not a very good proxy of brain levels of antibody. It has to do with TFR being, it's really its capillary-mediated distribution, right? So it turns the brain into any other organ in terms of the way that drug is delivered. And we've shown in a range of preclinical studies over years that the CSF levels of drug are not reflective. What matters is target engagement. And so to get to your point, the healthy volunteers will be helpful for looking at anemia, looking at some of these other things. The next step will be to go to patients and look at our ability to clear plaque. I think that will be critical in terms of the best measure of brain uptake and activity of the molecule. That's something that we think, based on available biomarkers, that we can do pretty early on within our patient study. We'll, of course, also look at ARIA-based side effects as part of that study. But in order to get really hard numbers on ARIA, of course, the incident rate is relatively low, so you're going to need larger numbers to really refine that. But I think we'll get an initial read from the first patient studies.

How would you characterize your confidence right now that you can get to a relatively infrequent sub-Q administration?

So, you know, we are looking at various routes of administration. I think we're, you know, we're not guiding it yet on exact dosing paradigm. But what we do feel is that we are exploring sub-Q administration as potential for this molecule. And that's another point of, you know, area that we're very excited about.

Okay. And do you guys envision yourselves partnering this after proof of concept or taking it all the way?

Yeah. So, I mean, I think it's important now to take a step back and mention, I think Denali, you know, we're well capitalized at present. So we ended last quarter with $1.01 billion in cash. So we do have the capital to progress these molecules right now through early stage studies. I think once we generate some of these initial data points that we're talking about, our Alzheimer's disease programs are great fits for potential partnering, but I think we'll decide at that time based on how the data looks, whether we're going to entertain partnerships or continue developing the molecules on our own.

Okay. To that point, you're also working on Tau. I think the Biogen data were promising but not without controversy. What do you make of that data, and what's the read-through to your Tau approach more broadly?

Okay, yeah. So, not to comment too much on others' data, but I think that that Biogen data was an important data point for the Tau field. The first thing I'll say is there is a lot to be encouraged about from the Biogen data. And, you know, the first thing to be encouraged about is that you, you know, they were able to reproduce their tau lowering in a larger clinical study and correlate that with tau PET, so looking at readouts of tau pathology, and really validating for the first time that you could lower tau expression and this very reproducibly results in a reduction of tau pathology is a huge leap forward for the field. Having worked on tau for some time, this is something that people have been interested in for a long time, and really seeing that come to fruition is fantastic. I think where it gets a little bit more tricky is looking at the correlation between tau pathology in that study and some of the clinical endpoints, right? And they, of course, saw somewhat of an inverse dose response. Again, I think the encouraging part is they did see an impact on cognitive endpoints in that study. I think the concern is, you know, the reasons behind why there's that inverse dose response and what it means for other tau therapies like Denali's.

Yep. Okay. So, you know, it seems like the two ways people are looking at tau in the brain are CSF and PET. What's your view on, like, the relevance of each of those? Like, I guess from my perspective, like, it feels like PET is really what's telling us ultimately what's happening with, like, toxic intercellular tau. Like, do you agree? Like, because I would think for your approach, it's going to take, or for any approach, it's going to take some time to really be able to, like, read out PET, right?

That's right. Yeah, and if you look at the Biogen data, you know, they're measuring PET at 18 months after dosing, right? And so, you know, the way that I would think about the different biomarkers in the tau's pet space, and again, like A-beta, we've really been enabled in this space because there's good biomarkers now to help us understand what these drugs are doing. And so, looking at tau levels in CSF, from our perspective, that's essentially a target engagement. You're asking, are you knocking down tau expression? Are you getting it to the right, you know, the right parts of the brain? You're a drug to the right parts of the brain. I think the real key is that tau PET. There's a lot of data out there showing that tau pathology is the best correlate with cognitive outcomes in Alzheimer's disease. So as tau pathology increases, your cognitive outcomes decline. And I think that's why people are so excited about tau as a potential therapeutic strategy. And so I think that this is where, you know, in expanding upon how our program might differ from, you know, an interthecal-based program like Biogen's, you know, again, is that much like any of our TFR programs, it's delivered through capillary-mediate delivery in the brain. So when we think about the Biogen data and say, is there a narrow therapeutic window for tau? Is it that if you knock down tau too much, you're going to have issues? I think that we can't rule out that possibility, but I think what's more likely, potentially, is that they're dosing very high doses of this ASO therapeutic. And when you look at biodistribution, and Biogen has shared some of this data publicly as well around the CNS biodistribution of that therapeutic, It's very uneven. You get very high concentrations of drug in certain brain regions, not high concentrations of drug in others. And so that also, you know, and, you know, in that study, they also saw increase in safety signals at higher doses that may be related to drug levels, right? So the other possibility is that there's just, you know, the drug levels are causing effects that would impact that clinical output. That's something that we don't think that we would see to the same extent with our TFR-based approach. And I think for a couple of reasons. One is we're, you know, we're delivering, you know, systemically, and we're, again, capillary-mediated delivery. The second is we've shown, you know, a number of times that our TFR-mediated delivery has very even biodistribution throughout the CNS. So we think that there's going to be distinctions there, and it's going to put us in a better position to get to the right level of tau knockdown safely.

Yep. Okay. Makes sense. Any more color on what we can expect with the initial data readout, like number of patients, duration of follow-up, things like that?

Yeah. So with our tau program, the first readout that we're looking for is tau knockdown in the clinic.

That's CSF.

That's CSF tau knockdown, yes. Thank you for clarifying. And the reason why, again, is back to what I said, your tau PET readout, which again would be the next step after that, that's an 18-month readout. So that takes quite a long time to hit a point where you're impacting pathology meaningfully. Even tau knockdown in the clinic, it takes about three months to knock down tau protein levels to the extent that you're interested in. And that's likely a result of tau half-life, right? So you're knocking down expression, but then it takes a while for the protein levels to go down to where you want. So that's going to be the first meaningful readout, CSF tau levels, that we're going to see from that study.

Okay, great. You'll also have some data coming up for the progranulin program. Do you want to maybe talk about what to expect from that readout? And, you know, I think one interesting thing that you guys have kind of previewed is that you'll be also focused on neurofilament at 12 months. Maybe give some context there as well.

Yeah, definitely. So we recently made the decision to extend that study from six months to 12 months. This was something that we had actually had interest in doing for quite some time. Recall that this program for a while was partnered with Takeda, and so we had aligned initially on a six-month endpoint with our partners. Once we got full rights to that program back from Takeda, we had the liberty to do what we felt was the best study. And you're absolutely right that we're focused on NFL. And the reason we're focused on NFL is we feel like progranuline, just to take a step back, that program in many ways behaves like one of our enzyme programs, that we feel like the active area in the cell for progranuline to go is the lysosome. The difference there is compared to our lysosomal storage disease programs, there's not super strong biomarkers, lysosomal biomarkers that we can measure clinically for progranuline patients. So, you know, we're not able to do some of that proximal target engagement type of work that we talked about for Pompeii or SJSH or Ablea. And so we're then, you know, really puts increased emphasis on the NFL readout there. And we feel like also, again, looking at our ERT programs, what we saw is sometimes NFL takes time to really, you know, hit that impact on NFL that we're interested in. So we wanted to extend that study to make sure that if there was a signal there, we gave ourselves the best chance to see it.

How well do we understand the kinetics of NFL, NFTD, GRN? I followed, and I know you followed, it's your neighbor, the Elector Development Program there. And I remember when they read out NFL, and it was so variable, they were putting it on like a logarithmic scale. And it just kind of, I don't know, it's sort of, I think that and a number of other readouts have sort of shown us that neurofilament can have, you know, very different variants and kinetics across different diseases. So how are you thinking about that as it relates to interpreting this 12-month?

It's a great question, and I think the short answer is we don't know much, right? I think in the absence of having a therapeutic that's effective and being able to see an impact on that readout as a readout of therapeutic benefit, it's hard to make predictions. And I think what you see is if there's not a lot going on, then you try your best to figure out, like, whether there's something there. I think what we've learned from our ERT programs, and we're not necessarily setting a bar for what type of NFL we want to see with progranulin for that reason, but we want to be able to see something that is meaningful and is really making it clear to us that we're having an impact on that biomarker, especially given the weight that we're putting in it.

Okay. And so if those data look promising, like you're showing a decrease in NFL, what's next on the plans there, regulatory engagement?

Yeah, I mean, I think that we are definitely going to explore the potential for an accelerated approval in that indication. You know, that's something that we haven't had, you know, conversations with regulators on yet, but something that we'd be interested in doing. And I think beyond that, it really depends on what that data readout looks like in terms of how we take that program forward.

And so what's the timing of that readout, Joe?

So that readout will come in 2027.

Okay, but nothing more specific than that?

I don't think no, yeah.

Okay, okay, great. We have like a minute and a half here left. Anything else you'd like to add that we didn't cover?

No, I mean, I think we're excited overall about the portfolio. So, you know, a couple of advanced enzymes. We talked about Ablea, Strong Launch, which is fantastic, and our Zafi program, SDSH, you know, shortly behind. You know, our third enzyme is getting ramped up. I think where we're also really excited about it is our Alzheimer's programs. So, both the A-Beta program and our TAO program, there's good reason to be excited there, you know, to really impact a broader population with our therapeutics.

Yep. Okay. Great. Thank you very much for the time. Best of luck.

Appreciate it.

Thank you. Yep.