Alnylam Pharmaceuticals, Inc. (ALNY) Earnings Call Transcript & Summary

Great. Thanks, everybody. It's my pleasure to be moderating this panel with Kirk Brown, who runs CNS research at Alnylam. I'm going to mention this to people now, so they don't just [ ping ] me with questions that this is a neuroscience-focused chat with Alnylam, and I'm not going to be bugging them about the TTR PDUFA.

But with that, Kirk, thank you for taking the time. And maybe just set the stage for people and give a brief review of Alnylam's efforts in neuroscience, the different programs you guys have in the clinic and lead preclinical and then we'll dive into each. So I appreciate it.

Sounds good. Thanks, Paul, for having me on. I appreciate it. So little on the CNS portfolio at Alnylam briefly. So the sRNA platform for CNS has built off a foundation of sRNA developments and innovations targeting the liver, essentially focusing on making an sRNA more potent and durable and more specific. So we've applied all those learnings to our C16 platform, which enables the delivery to the CNS in multiple cell types across the CNS. With our partners at Regeneron, Alnylam now has a growing portfolio, which leverages the C16 platform. And it does utilize intrathecal administration, which we've seen fantastic distribution across the spinal cord, the brain and into the deep brain. For programs, Mivelsiran is our lead program in the CNS. It targets amyloid precursor protein or APP for the potential treatment of Alzheimer's disease as well as cerebral amyloid angiopathy. This program is currently in a Phase I study at early onset Alzheimer's disease, where we shared initial data from both the single ascending dose and multiple dose parts of the Phase I. And we've demonstrated potent reduction in the biomarkers of target engagement as well as encouraging safety profile in the early studies to date. And that Phase II is planned for later this year. Mivelsiran is actually also concurrently in a Phase II in cerebral amyloid angiopathy, which is a major cause of intracerebral hemorrhage, which is the most severe form of stroke. HCT is another program of high interest to us and high unmet need for the treatment of Huntington's disease and this program is currently in Phase I in enrolling patients. Then we have Regeneron who is leading the SOD1 ALS program, and we have numerous preclinical programs and developments such as MAPT for tau and tauopathies as well as alpha-synuclein for Parkinson's disease. And we've recently shared at R&D Day, a brief update on the delivery progress of our sRNA platform, and we've shown that we can deliver across the blood-brain barrier to nonhuman primates. So really, we're particularly excited about what we're seeing in the platform, both with our C16 platform and clinic and also what's to come in the future with the sRNA delivery in the CNS.

Great. Let's talk about APP as a target. I mean, I think CAA, it's intuitive to me. I mean, the monomers are toxic and you're 1 step away, I mean, the analog to TTR, I think, is really good. I think for Alzheimer's, the question we've been trying to wrap our minds around is why isn't APP too far upstream, right? I think about solanezumab, the targeted monomers, very marginal benefit, right? You think about BACE inhibitors, right? And I guess in a world where amyloid starts accumulating 10 years before anyone has disease anyways, like I feel like if you're not hitting APP when someone is 30, like are you really going to show a big effect? Obviously, you guys have a sophisticated view on this. So what gets you excited?

Yes, great question. I mean, first and foremost, APP is a genetically validated target for AD and CAA, right? We know that mutations in APP itself as well as the enzymes that process it, drive early onset AD and autosomal dominant AD as well as CAA. We also know that duplications or even trisomy 21 in Down syndrome also produce early onset AD and CAA. So APP is attractive to us for many reasons. There are clear biomarkers of target engagement, which can be measured in Phase I. It's cleaved into multiple soluble products such as soluble APP alpha and beta, which are measurable in the CSF. What we know from other approaches such as antibody-based approaches that they've targeted a variety, as you mentioned, of amyloid states and assembly states, monomers, oligomers and others and have had little success halting the disease progression. Honestly, we feel like targeting far upstream at an mRNA level, that's something that others cannot, which are addressing all downstream products and assembly states of the amyloid as it accumulates into plaques or in CAA aggregates along the vasculature. Importantly, RNAi can do things that an antibody-based approach cannot do. It works well, both intracellularly and extracellularly and addresses those manifestations of the disease, particularly intracellular toxicities that have come to know from neuronal health. So we're excited by that. Regarding BACE in Gamma-secretase inhibitors, I mean, they're targeting APP production, but at an enzyme responsible for processing where they're not potentially the most specific or sole substrate of those agents, whereas with an sRNA, it's truly ultra specific, right? The substrate or target in their case is a complementary messenger RNA sequence of APP. So it is incredibly specific to that messenger RNA, therefore, allowing us to have confidence in all reductions of downstream amyloid both in the plaques and in the aggregates, as you said, along the vasculature. And we've seen that preclinically. We've shown preclinically that you can lower APP messenger RNA and see changes in both brain parenchyma as well as aggregates along the vasculature in preclinical models of CAA and AD.

Yes. Yes. Okay. So how can you prove out this hypothesis? I mean before running a Phase II/III study with an 18-month cognitive endpoint in 500 patients, like walk us through the Phase I/II plan and when do you think we could expect to see something on pTau or PET imaging or some sort of measure that kind of validates again that this isn't too far upstream that this is having an impact on pathogenic amyloid species?

Yes, great question. I mean we know from at least the early biomarker assessments of amyloid beta 42 and 40 that we are similar to soluble APP alpha and beta reducing them in the CSF. We're seeing clear target engagement. We see dose responsive effect...

People listening in and where is -- what do you get -- when you measure amyloid beta 42/40. What part of the pathway are you...

What part of the pathway? We're such a...

Monomers like what -- I think you're still talking about the early nonpathogenic subspecies. Is that right? Or is that wrong?

Yes. That's a fantastic question. I mean it's really the critical question of like what is the source of the amyloid, right? Is it coming from plaques? Is there a natural turnover because we're not aggressively targeting the plaque for clearance, which also potentially protects us from some potential ARIA-like effects. But we are seeing those clear reductions, both of amyloid beta 40 and 42. But to the question of when do we expect to see, we are collecting important disease progression biomarkers. We're looking at amyloid PET as well as exploratory biomarkers. And we just recently modified the study to include an extension to capture longer-term knockdown over the course of 36 months, which will give us a pretty good clue into what's happening both on exploratory biomarker front, PET imaging and also on biomarkers such as tau and neurofilament and others. And we're pretty encouraged by the safety signal we've seen so far, right? I mean after both single and multiple doses, we see essentially no impact on neural information. No changes in CSF protein or white counts, and importantly, neurofilament is holding even after multiple doses, which is quite encouraging for our first foray into this space with CNS.

Right. Okay. Any update for people listening in on the timing of when we can get that data on things like tau and PET?

Yes. So we're planning some additional data readouts later in this year. Stay tuned for the...

Yes. And what would be the duration of follow-up by then for these patients?

Yes. So as I said, we've expanded the scope, the open-label extension now multi-dose will go out through 36 months.

So you have patients out beyond a year. Like basically, I'm just trying to -- like it will be long enough to really be able to get an answer on these things?

Yes, yes. Yes.

Okay. Great. Do you want to talk a little bit about CAA, I mean, I feel like this is overlooked. When we think about APP, right, this question of proximity to the pathogenic targets, it's less salient, right, because the monomers are toxic. So maybe just give people a little bit of a background on the disease, the work you guys are doing there and what are -- beyond biomarkers, what are the right endpoints to focus on?

Yes. So as I said, we started a Phase II, the cAPPricorn study, looking at efficacy, safety and pharmacology in CAA, both sporadic CAA, which is much more prevalent as well as a rare Dutch-type CAA, which is the genetic sort of ultra-rare population, which affects folks at a much younger age, where they see recurrent strokes in the ages of 40 and 50s. And so we'll be looking at both and maybe to your question, both hemorrhagic and non-hemorrhagic manifestations of the disease. We have those 2 cohorts that I described, which will be randomized in a double-blind versus Mivelsiran and Placebo for 24 months. And then we have an open-label extension there for 18 months to assess additional safety and efficacy. To your question around CAA, yes, it's certainly underrecognized cause of stroke, but the CAA pathology itself is quite common. It's been argued that CAA pathology is seen upwards of 20% of the population and actually higher so in those individuals with Alzheimer's disease in those comorbidities. And really only a subset will have imaging abnormalities.

Yes. Okay. How prevalent do we think CAA actually is? It would seem like, at least today, it's -- the diagnosis rate is probably very low. Is that fair?

Yes. I think it's improving, right? I mean there's the Boston criteria 2.0. But there's certainly an unmet need for patients with CAA to reduce stroke. So I mean it's the second most common cause of intracerebral hemorrhage after hypertension. And there's approximately 80,000 or so new or recurrent cases of ICH each year, which a significant number of those will have CAA or driven by CAA.

Yes. Okay. Anything else to add on CAA before we talk about Huntington's?

No, I think that's good. Thank you.

Okay. Yes. Maybe give us an overview of your guys' efforts in Huntington's. And look, from my perspective, right, I mean there's -- like I would love your perspective on just Huntington's in general because on the one hand, right, it seems like there's so much to be excited about biologically, right? Like it's autosomal dominant. We understand the bad actor. On the other hand, right, it's been really challenging. I mean you have this tominersen failure, which the drug did worsen placebo on 6 endpoints, right? Like there's been challenges with the CSF assay with Huntington, like I feel like it's -- I feel like there's truly 2 sides of the coin here. And so in the backdrop of that, why is this an indication that Alnylam decided to pursue?

I mean, you touched on. I mean, it's an attractive target for genetic medicine, right? It's a high unmet need patient population. And we know that we're able to deliver well to the brain and the deep brain. So Huntington's is an attractive potential place for us to have disease-modifying therapy. And we're really excited about our Huntington's program for a variety of reasons. And you mentioned some of the other players or failed programs. Our approach is quite differentiated from those. We're including what's known as an exon 1 targeting strategy, which is a toxic fragment that is produced from the expanded repeat, the greater the expansion, the greater the propensity of this alternate splice variant exon 1 toxic peptide to be produced in addition to the toxicity that's associated with the mutant Huntington itself. Secondly, we know that we were able to hit Huntington's quite well and deeply. We've seen that we can show very robust silencing both in the brain and in the deep brain, and we can do so using our C16 platform, which is incredibly encouraging with what we've seen so far in clinic as far as the emerging safety profile as well as the duration of activity we've seen with APP lowering. We think it's likely that it would be something around Q6M or even less frequent dosing in clinic. So fewer intrathecal administrations, potentially even a better safety profile there for patients with HTT.

Okay. Why is there such a controversy in this space around the right isoform or what is the actual driver of the toxicity? Like is it full length? Is it the exon 1 fragment? Like where does this all kind of originate from? And why isn't there like unequivocal agreement in the space?

Yes. I mean it's a fantastic question. I mean there's a growing body of evidence now supporting both from preclinical models through multiple preclinical models showing you that the sematic instability that occurs in HTT with this expansion of the CAG repeat does produce a much greater propensity of that shorter isoform. Preclinical models, longer CAG repeats correlate with increased expression of the 1A fragment and that shorter 1A fragment is highly aggregation prone and therefore, highly toxic in these rodent models of disease. And so multiple labs now have shown that lowering Huntington using the shorter exon 1A targeting approach prevents the protein aggregation and other Huntington's phenotypes much more so than just lowering the wild type or the mutant Huntington itself. So literally addressing this aggregation prone effect much more directly by targeting the exon 1 fragment. But our approach and just to clarify, we are not exclusively targeting the shorter 1A fragment, right? We're also targeting the full-length mutant fragment. So we should have this likely potential to address this question quite directly, the combination of lowering both the 1A fragment as well as mutant Huntington potentially to have some benefit here.

There's still an open question at all as to whether knocking down wild-type Huntington is safe.

Yes. It has been around for a bit, but I've grown more and more confident with our preclinical work leading up to the program heading to clinic. We've shown both in single and multi-dose nonhuman primate studies that we can lower the wild-type Huntington protein in the brain and deep brain striatum caudate to the levels of 80% to 90% at the protein level and hence, no adverse events or toxicity noted. So -- and these are long-term studies, not just 1 month or 2 month studies in nonhuman primates. Long duration of silencing of Huntington's protein in the brain and deep brain in our preclinical workup. So I think if we were to see something with -- I mean, 90% knockdown is quite robust.

Right. Yes. Okay. What does the development path look like in Huntington's? And beyond showing knockdown in the CSF, what do you see as the clearest or easiest way to prove that there's some downstream clinical benefit? Is it neurofilament? Is it something else?

I mean, yes, fantastic question. I mean the initial study, we're looking at PK/PD and widespread engagements. We're going to look at biomarkers of target engagement in the CSF. We are going to look at neurofilament in the HD population, which has been shown to elevate over time. The thinking there is we should be able to hold it in place or level that off, which should give us some confidence that we're seeing at least preserve health of the neurons as the disease progresses.

Yes. Okay. Okay. Well maybe I'll turn it over to you. What else would you like to talk about in your CNS pipeline?

I mean we're excited by a lot of the different programs that are coming down the line. We have, as I said, the SOD1 program that is run by Regeneron currently in Phase I dosing patients now. And I would really direct you to them for specific details on the study and readouts. And we also have a partner program with them targeting alpha-synuclein for Parkinson's disease. A little bit shortly following behind these sort of the C16 targeting MAPT for tau lowering, which we believe has broad utility across wide range of primary and secondary tauopathies, Alzheimer's disease, PSP, dementia and others. I mean, tau is a pretty high-profile target of interest in AD space. And we believe a genetic approach like ours has a potential to have a great impact, both intracellular tau as well as extracellular.

Yes. Tau is really exciting. Do you -- as we think about the relevance to Alnylam, like how important is the data from an ASO approach that's coming sometime next year.

Yes, it is important. I mean I think it's -- they've shown that you can lower tau to a certain level in the CSF and in the brain. It will be interesting to see what they're able to share with everyone, is it first half of next year. We are following that program.

Yes. Okay. And as it relates to tau specifically, right, like what gives us confidence that targeting tau or targeting intracellular synuclein. Like is this going to be safe? I mean these -- like tau has a role, right, in normal brain development, neuronal cell integrity. What can we do to kind of diligence that question?

Yes. It's a fair question. I mean none of these treatments are knockouts, right? They're not permanent changes, right, and we can track a dose-responsive change in tau both in the CSF and you can visualize it through imaging. And we'll obviously track safety biomarkers in these trials as well.

Yes. Okay. As we think about you guys positioning longer term, do you guys have efforts internally? Or are you looking externally at all at brain shuttle-based approaches, right? I mean, I think the interesting thing about what you guys have accomplished with your conjugation approach is you can get broader biodistribution, deep brain delivery, which is great. IT is still not ideal. And it feels like we could be on the cusp right of delivering the brain with transparent or something else via IV. Like is this something that Alnylam is going to be participating in?

Yes. Great question. I mean I do think, first, I want to say that we're pretty thrilled with what we've seen so far with our first-generation C16 platform in clinic, right? The beauty of this platform is that it's both modular reproducible. We can drop on any sequence to the C16 technology and know that we can confidently deliver to the spine and the brain through a wide variety of cell types, and that's an important distinction with the C16 profile, right? It's quite cell type agnostic. But we also believe that there's going to be a future where systemic delivery of sRNAs to the CNS is also something that should be an option depending on the indication. And we're going to take this approach sort of as a CNS target indication dependent decision. But we're certainly playing in this space. My colleague, Vasant Jadhav at R&D Day shared some early data from us a few weeks back showing that we're able to show a pretty robust silencing across the BBB in nonhuman primates, which showed nice distribution both to the cortex, which is common with IT dosing but also to the deep brain at comparable doses. So we're encouraged by what we're seeing there. But we also are planning it somewhat as biding our time wanting to have essentially the best-in-class approach, it's a BBB delivery, which includes a variety of different targeting approaches that we're exploring both internally and through partners.

Yes. Okay. Kirk, I hope you don't mind, I'm going to ask you 1 question that came in. I have filtered out any TTR questions. So even though I thought it would be funny to ask one naively. But as it relates to Huntington's, what do you guys think about the regulatory environment and the potential for accelerated approval? And when are we going to get the next data readout from your program?

Great question. I mean I think we'll try to move as quickly as we can, but we're going to let the data guide us, right? We want to make data-driven decisions. If we see encouraging knockdown and potential, like I said, the biomarkers of neurofilament, we see some halting progression of neurofilament elevations over time. We'll try to move it as quickly as we can. But again, this first step is safety, tolerability and target engagement.

Is there any reason to be concerned about the CSF assay with Huntington's. I mean I think with PTC, right, they -- correct me if I'm wrong, I think they're looking at peripheral Huntington. And then uniQure, right, they've had some noisy data. I know Roche Ionis seemed to have it figured out and then, of course, that program was -- so just like, yes, your confidence in measuring Huntington. We can't measure the exon 1 fragment, right? So like what does biomarker data look like?

Yes. Biomarker data, at least in the Phase I, will be mutant Huntington in the CSF. The C16 platform is unlikely to see changes of any sort of biomarker of Huntington and the Huntington target plasma.

And you're comfortable about like the CSF assay is like reliable and has like a tight variance? Because I thought uniQure had some issues. I don't know if it's the same assay.

Yes. I mean it's a fair question. Is it as tied to some of the biomarker assays we have for APP, potentially not, but I think it's engaged -- it shows enough target engagement and confidence that we've seen at least in assay optimization to date to believe in it. So we're confident moving forward with the mutant Huntington assay in CSF. It's a good question.

Yes. Okay. Okay. And then one other question that just came in, and I guess, for Huntington's or for tau. Can we be comfortable that preclinical models for these drugs, these targets and maybe for idiosyncratic tox are good predictors of human safety?

Yes. I mean the preclinical models are often very exaggerated states, right? I mean whether it's APP or Huntington or tauopathies, but we have certainly in tau shown that we can reduce aggregates in tau. We can see reduction of aggregates in amyloid accumulation preclinical models of AD. And same goes for the exaggerated settings as the repeats of HTT. But as far as safety is concerned and engagement, we're able to show that in nonhuman primates and are confident with that moving forward.

Great. Anything else to add before we wrap up, Kirk?

No, thanks. This has been fantastic. I appreciate you having me on.

Yes. Thanks, mate. I'd love to talk about neuroscience. So I appreciate it, and thanks, everybody, for joining. Yes, we'll see you on the next one. Thanks.

Take care. Thanks, Paul.