Denali Therapeutics Inc. (DNLI) Earnings Call Transcript & Summary

Everybody, good morning, and thanks for joining us for the next session. I'm Matthew Harrison, one of the biotech analysts here at Morgan Stanley. So I just need a quick disclaimer I need to read. Please note that this webcast is for Morgan Stanley's clients and appropriate Morgan Stanley employees only. This webcast is not for members of the press. If you are a member of the press, please disconnect and reach out separately. For important disclosures, please see the Morgan Stanley disclosures website at morganstanley.com/researchdisclosures. And if you have any questions, please reach out to your Morgan Stanley sales representative. So very pleased to have Ryan Watts, the CEO of Denali with me. Ryan, I'm going to turn it over to you to make an opening comment or 2, and then we'll get into Q&A.

Matthew, great to be with you. I, of course, would love to be with you on stage in New York. What a bizarre and challenging year. Denali has had actually a pretty extraordinary year in terms of where we are as a company after 5 years. Our focus on the genetic pathways on engineering brain delivery, biomarker-driven development. We've had a successful year both presenting data on Parkinson's disease, launching our first clinical trial in rare disease with our blood-brain barrier technology and then entering into, I think, a critical collaboration for us on LRRK2 as well as blood-brain barrier targets with Biogen. And so although challenging and working remote has been somewhat challenging, it's been extraordinarily productive and looking forward to spending the next half hour with you, Matthew.

Great, great. So look, I think I want to focus on blood-brain barrier to start with because you've got some key data coming up. But maybe can we just touch on the Biogen deal, which I think was a pretty sizable and transformative deal for you guys. How are you thinking about prioritizing the pipeline now in light of that deal as well as some of the other deals that you've done over the course of the last year?

Yes. So I think our deals have been important for a number of reasons. They've allowed us to both accelerate the programs that are in the partnership, giving them the essential resources to move forward. That's the case with the Takeda partnership, with our Sanofi partnership and certainly the case with Biogen with LRRK2, but it's also allowed us to expand our portfolio. In addition to the partnered programs, we now have a broader portfolio of both all molecules as well as transport vehicle-enabled large molecules. And so I think this deal, we're at a critical point with LRRK2, selecting a lead molecule, designing the late-stage clinical trials, and we prefer to do that with a partner. And so it's had a very immediate impact on our portfolio, again, going faster, deeper and broader.

Great. So as you -- you've got an important readout coming up on blood-brain barrier hopefully towards the end of this year with ETV:IDS, but you also put out some, I think, important preclinical data earlier this year. So maybe just start with that. What does the preclinical data tell you about the ability to drive higher concentrations in the brain in humans? And what sort of level of confidence do you have given that data?

Yes. Great, great, Matthew. So for those of you on the call, we've also distributed a PDF that has slides that I will speak to as well. So if you go to Slide 17, this is the first of 2 papers published back-to-back in Science Translational Medicine. So these were published on May 27. And basically, the 2 papers are as follows: the first is the actual invention of the transport vehicle technology. So it's engineering the Fc portion of an IgG to bind to transferrin receptor, which is required for transporting iron. And so transferrin receptor is expressed at high levels in cells that are required both to transport iron or a high metabolic need. And in the first paper, we laid out the engineering technique directed evolution to the Fc. So it's the natural number of amino acids. It allows us to basically get any antibody across the blood-brain barrier or to make Fc fusions and to get them across the blood-brain barrier. We had non-human primate as well as an engineered mouse model in that paper. And actually, the engineered mouse has the transferrin receptor of human binding domain that allows us then to cross to all of our various animal models, including the Hunter syndrome models, the Alzheimer's, Parkinson's models. We can take our clinical candidates, test them in those mouse models and understand basically the dose that's required to drive efficacy and then scale that to humans. And I think one of the most -- probably one of the most important pieces of data in that paper is validation in the nonhuman primates. And so we looked at both the duration of response as well as the robustness of the response after a single dose. We actually looked at nonhuman primate brain 2 days after a single dose and see about 30- to 35-fold increased uptake across the various brain regions. So that was paper number one. It represents about 4 years of work and both from our engineering side as well as the biology team. And it has essentially laid the foundation for what is the right affinity, what is the right sequence for the clinical candidates. The second paper is the application, and this is now Slide 18, but it's the application of the transport vehicle technology to Hunter syndrome and enzyme fusion. And so we can go into more detail here, Matthew. But that basically is the data that is the rationale behind our ongoing Hunter syndrome trial with DNL310. And just in terms of nomenclature, ETV is enzyme transport vehicle and then of course, IDS is Iduronate 2-sulfatase. It's essentially an Elaprase Fc fusion or like Elaprase, it's a natural enzyme sequence that we can get across the blood-brain barrier to basically rescue the neurological symptoms in Hunter syndrome, which are not treated with current therapies, specifically with Elaprase.

Perfect. Yes. I think it's good, and I think I want to spend a little bit of time talking about data expectations towards the end of the year. But maybe just before we do that, another high-level question, which is what's the read through? Let's just say, top level, you say, we have so much Elaprase in plasma, and we measure it in the brain, and there's a 30 -- I mean there's an improvement over what you could get in if you were just giving it in the plasma to the CSF with your platform. What's the read across to other targets that aren't enzyme targets that are maybe antibody targets or things like that? I mean do you see it as essentially proving that you can get any sort of molecule in the same concentration into the brain? Or are there caveats that we need to be aware of based on this data?

Yes, excellent. Excellent question, Matthew. So I'll reference for those that have the slides in front, I'll reference Slide 19 to set the context of the other modalities in addition to the enzyme transport vehicle. So we have 2 antibody -- we have a number of antibody programs, but disclosed, obviously, we have a TREM2 antibody, we have an Abeta antibody now with Biogen Tau antibody using the transport vehicle. These would be antibody transport vehicle. We have progranulin fused to an Fc. That's the protein transport vehicle. And then recently, we've validated that we can deliver ASOs across the blood-brain barrier with what we call the oligo transport vehicle. So these are the, I'll call them, the submodalities of transparent -- of the transport vehicle. And what's really important with this initial data is that it validates transferrin receptor as a pathway to the brain in humans. And so we see this as critical data to unlock the other programs. And I think as you point out, Matthew, it definitely increases probability of success of all subsequent enzymes. But it also allows for the initial validation of transferrin receptor in humans for the other transport vehicle modalities, the antibody protein and oligo. I think with varying degrees of validation, enzyme, it's almost 1:1. I'd say the proteins for progranulin, it's close to 1:1. And in fact, probably the ASOs, it's likely also a pretty strong correlation. I think with antibodies, the dynamic is that you have a different dosing paradigm. And so there, it's still validating for transferrin receptor, but it will be a target-by-target requirement. So for example, with Abeta or aducanumab-like molecules, we're going to need to understand dosing and dose levels. Our guess is that we'd probably get the same amount of antibody in brain with 1/10 or 1/20 the dose. But then of course, you think about is the monthly dosing, and so we see that as really important, validating this initial data. And I think we'll get to this, but our goal is to achieve greater than 50% reduction of substrate in cerebral spinal fluid in humans because we know when we achieve that, we essentially have a correlation with complete rescue of lysosomal function in Hunter syndrome. And that, I think, exceeds what we would need to validate transferrin receptor itself, in fact. And I'm going to go back to that monkey data. And what we see is that transferrin receptor-enabled technologies, like the transport vehicle, gets across the capillaries. So every brain region, we're seeing the same 30-fold increased uptake. So when we see that CSF reduction, we essentially know that we're having basically broad biodistribution with the transport vehicle.

Okay. And I think that's probably a good point because obviously, some enzymes have been dosed into the CSF, and they probably didn't get broad distribution across the brain. So maybe just highlight for people how you're measuring and what you think about biodistribution and why that means other examples where you dose in the CSF might not yield the same result.

This is, in fact, a critical difference and a point that we've worked on for a number of years, both at Denali, but also previously, when we were working on blood-brain barrier at Genentech, we looked at basically mini pump delivery of antibodies into the CSF. Obviously, there's a lot of work on enzyme and even AAV delivery into cerebral spinal fluid. And what's really fascinating, and I think this -- I'm just going to point specifically to Elaprase and the Shire intrathecal delivery. That intrathecal delivery is in a lumbar region, so towards the bottom of your -- of the spine in these young boys. And what it requires is basically from that point that the enzyme diffuses throughout the entire brain. This is very different than using the transport vehicle and using transferrin receptor where you have systemic delivery, you have high circulating blood levels and then you get across every capillary wall. And so in our experience, we were getting a very small percentage of large molecule distribution throughout the brain when we have an intrathecal delivery, and especially when brain regions that are required are distal. It's essentially routing and motion, right? The molecule will travel with cerebrospinal fluid and then just relies on the fusion as opposed to direct delivery across the capillaries to the neurons. And I think what's important is the pharmacodynamic readouts in that Shire trial were measured at the point of enzyme injection. So it's biased towards basically high concentrations. The real question is if you were to pull cerebrospinal fluid from the central nervous system, one of the ventricles, what would be a reduction, it would likely be substantially less. That's been our experience.

Okay. Got it. So maybe on the specific study and what we're going to expect to get by the end of this year. Maybe you could just talk a little bit about what biomarker data investors should expect to get and what you're going to see as the success?

Okay. I think for that, what I'll do is reference several slides. And for those that don't have the slides, I'll verbalize as well. So please reference Slide 22, which is basically the first experiment we did comparing head-to-head with Elaprase. And point number one is that in periphery, so we looked at liver Glycosaminoglycans, you get 95% reduction with both Elaprase and ETV:IDS. In this clinical study, which ultimately we want to enroll about 16 patients, we're washing out Elaprase, and we're giving just ETV:IDS. So the first point is that we will essentially be able to look at urine gags for replacement of Elaprase. So we have an expectation that we'll be equally or more efficacious in the periphery than Elaprase has been. The second is looking then at brain Glycosaminoglycans. So in animals, we can obviously look at brain homogenates. So we look at the brain and then we look at the CSF and we correlate the 2. And if I now reference Slide 23, you can see an example of this where we give a relatively low dose. So 1 mg per kg weekly or 3 mg per kg weekly, and we can achieve greater than 70% reduction in brain and around 50% to 60% reduction in CSF. And that's actually the bar that we've set as the bar we want to achieve for Hunter syndrome. And the reason for that is if you actually look at the neurofilament levels, which we're the first to show that neurofilament is elevated in Hunter syndrome. We just actually published a technical paper looking in Hunter syndrome patients and see an elevation of neurofilament. But when we reduce CSF gags for 13 weeks by 50% in CSF, we completely block neurofilament. So we block neurodegeneration. And that's essentially the framework. So the study itself is a Phase I/II study. It's enrolling now. We have a number of patients enrolled [Audio Gap] is that we want an interim readout by end of the year, looking at Glycosaminoglycans. So this is either a 4-week or 8-week data, but it's a 6-month study. And the way to think about our program is really in 3 levels of validation. Level 1 is the target engagement. So it's looking at Glycosaminoglycans and CSF in short-term dosing with the bar trying to achieve greater than 50% reduction. Step 2, we'll be looking at downstream lysosomal biomarkers and neurofilament. That's after 6 months of dosing. So that's probably mid-2021. And then step 3, of course, is cognitive rescue or rescue of neurological symptoms. And we obviously plan to extend the Phase I/II, but then initiate a Phase III based on that Phase I/II data. That, of course, would be registrational with neurological endpoints.

And as we think about broader read across to the platform, in that data when you look at the gags and the CSF, I mean do you feel confident because it sounds like you're not going to be measuring Elaprase gag reduction in CSF. I mean I assume the literature is pretty strong on what that level is. And so I guess what I'm trying to ask is, how are you going to correlate how much of an increase in Elaprase that you can get across the blood-brain barrier, so you have some idea about the PK for the rest platform?

So PK is challenging. CSF -- so CSF is a great measure of pharmacodynamics, but it's not a measure of pharmacokinetics. And the reason for that is that when you deliver across capillaries, especially with the enzymes, these get internalized into neurons. So we measure PK in brain homogenate. We don't measure CSF. So the driving force for the decision is the actual pharmacodynamic response. So it's the product of the enzyme activity, right? So I think this initial interim data is critical read-through for the platform. This subsequent data, the 6-month data is really around validity in Hunter syndrome, what's the probability that it will be efficacious for Hunter syndrome. We obviously are going to look at a number of lysosomal biomarkers as well as neurofilament throughout the entire study, but our focus in the interim is this Glycosaminoglycan reduction. Now the great thing is that in the animal models, we can look at brain homogenates and CSF and make that correlation. That's where our confidence is placed on looking at gags as the primary readout from humans.

Got it. But -- and sorry to dwell on this, but from an investor standpoint, investors are not going to get a read that says we got 30 more concentration than you would have otherwise if you would have dosed Elaprase systemically without.

Yes. That's -- I think that's correct. I think what we will get is how efficacious was DNL310 or ETV:IDS. Did you achieve the biomarker goal? And that we can actually see -- we know the correlations from monkey and we know the correlations from mouse. And in order to achieve 50% reduction, you need that 20-fold increase of -- it's actually somewhere between 10- to 20-fold increase of enzyme in brain. But that's basically going to be the biomarker target engagement readout that will drive that decision.

Got you. And then sorry, one final question. When you look at and I'm just not familiar with the literature here, but when you look at Elaprase dose systemically, what's the reduction in gags in the CSF that you can achieve?

There's essentially no reduction of gags in CSF. It's given it roughly like a 0.5 to 1 mg per kg and there's no reduction in gags. And what we'll be looking at is CSF levels before ETV:IDS and after. And remember, these patients are on Elaprase. So they're on Elaprase, they stop Elaprase and then they go immediately on to ETV:IDS. So you'll be measuring essentially CSF levels on systemic Elaprase.

Okay. Perfect. So I want to jump to LRRK2, but maybe just one final question before we do that. Just remind people time lines for other blood-brain barrier programs that you mentioned in terms of bringing them into the clinic.