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

Good afternoon, everyone. Thank you for joining us for today's webcast as part of our ongoing genetic medicine for generalist series, where the tagline is genetic medicines are complicated, but the conversation around them doesn't have to be. And as you might know, the goal of our conversations is to take things back to the basics and bring everybody up to speed on the latest and greatest in the space. On that note, the more the merrier, in terms of audience questions. So please don't be shy about asking questions. Feel free to enter them into the webcast interface, and I will work them in as we go. And importantly, remember, there's no such thing as a stupid question. So please ask away. And with that, I am very pleased to be joined today by Kevin Fitzgerald, Chief Scientific Officer at Alnylam. And Kevin, I'll turn it over to you to give a brief intro to yourself and your background.

Sure. Kevin Fitzgerald. I'm our Chief Scientific Officer. I've been with Alnylam for nearly 20 years, and I've been studying this new area of biology called RNA interference for probably 25 or closer to 30 years. Happy to be here with you today and happy to sort of discuss something that's near and dear to my heart, which are RNAi therapeutics, especially as we're moving into the central nervous system with these therapies.

Definitely. Definitely. Excellent. No, thank you again for joining us today, and you are the perfect person to be having this conversation with today, I think. And I should also say for everyone on the line, we will start high level, as I said, and kind of drill down into the details as we go with a focus on the CNS pipeline. So I forgot to come up with like a HELIOS-B joke at the beginning, but assume I did, ha, ha, ha, laugh. Now we're moving over and focusing it on the CNS stuff because we are very excited about this as kind of the next area of the platform, as I know a lot of people are as well. So with that, as I said, taking it all the way back to the basics, what is RNAi interference? How does it work if you were kind of some of these brand-new -- what is this cool new thing we've been talking about?

Yes. So where I'd like to start is anybody who's had a cup of coffee this morning, this afternoon, congratulations, you're naturally occurring microRNAs or RNAi has happened in you though. And so it's a way that it's a mechanism that organisms over time have developed all the way from this [indiscernible] called [indiscernible] , it's microscopic all the way to people to control the way the genes go up and down. So I think of it sort of as a switch to turn things on and off in or [restart]. So more of the ability to turn a gene up, in particular, turn a gene more towards the off position. And so you can really control the level of gene expression. And the way that it works is that there's something called double-stranded RNA. And these are short double-stranded RNAs, was really discovered, won a Nobel Prize by Andy Fire and Mello a number of years ago where they discovered that this was working in that same [indiscernible]. And then Tom Tuschl and others were able to show that this worked in more mammalian cells. And so what we've set out to do is to mimic that naturally occurring system by using a synthetic made small RNA that we actually dose subcutaneous in general, and it goes in and it sneaks its way into the liver to this particular case. And then turns the expression of a gene down at the messenger RNA level. And so it's a way to take something. If you think about sort of antibodies that everybody is familiar with, they sort of mop the floor and this would -- RNAi as a way to maybe turn down the faucet, turn off the faucet as a way to control gene expression.

I like the analogy a lot. I love a good analogy. That's excellent. Okay. And so you mentioned it to turn off the faucet, it's kind of modulating kind of the expression of some of these genes. How does it do that? It's like binding to the RNA copy and then -- or not, explain to us how that works.

So it sounds a little bit like science fiction, but we'll inject our molecule subcutaneous into the skin. It will then survive going from the skin into the blood stream. We then have a ligand and the receptor pair. So that's something that it will bind to in the liver cells as it's going through in the blood stream. And we'll grab a hold of that and then it will be swept into the cell into a compartment called the endosomal-lysosomal system. It will survive that and then go into this complex that's called the EGO complex. And that's the complex that sort of does its function. So it will bind in there. And then that complex will scan through every messenger RNA, that's being made, or every RNA, and that's what produces proteins in the cell body, and we'll find the one that it matches and then we'll clip it. And so by clipping it, it will degrade, it will turn off. And so that's happening in 280 billion hepatocytes simultaneously with this technology. And so what happens out of that, if there's no messenger RNA, there could be no protein, and so protein will go down. And so you can measure that, for instance, one of our early programs was a drug called Fitusiran, which is now known as LEQVIO. And that was for Hypercholesterolemia. And so that drug you inject subcutaneously every 3 to 6 months. And what happens is that, that goes in, does all those things that I said, ends up in the removal of a protein called PCSK9 and then cholesterol more cholesterol goes into the liver and just taken out of the blood stream, and so you lower your LDL cholesterol.

Got it. That's really interesting. And I guess, are all silencing RNAs or RNAi's, or RNA's sorry, that kind of turned down the faucet created equally? Are there different ways to go about this? And I guess what's kind of -- what's the special sauce, I suppose, about the way that Alnylam does it?

Yes. So very early on, when we tried it with just -- so I told you it's sort of a naturally occurring process. So we tried to mimic those naturally occurring molecules [indiscernible] . And so what we've had to be able to do is to chemically modify. So while we call it our RNA therapeutics, there's really no RNA left in the molecule. It's all a different chemical modification. And what that does is it stabilizes it because there are things in the body called Endo and Exonucleases that low to cleave RNA. So they chop it into very tiny pieces. And so what we do is we put chemistry across the molecule so that those Endo and Exonucleases can't chew it up anymore. So that allows it to survive long enough for its action. And so we spent a large number of years trying to figure out sort of that complex pattern of what you could modify and what you couldn't so that it would maintain its ability to go in and do what it needs to do, which is to bind Ago2 and then cleave the RNA, the up you resistance -- resistant to degradation so it had enough time to get there and do it. And so all of our molecules are essentially fully modified at this point. And then we've taken this ligand but on the end of it, it's a sugar called GalNAc, and it turns out that there's a receptor, and that's something that it binds to in hepatocytes. That's about 1 million copies per cell, and that allows it to go in. Now as we pivot to the central nervous system, a lot of that -- those learnings absolutely true in the central nerve system, but we did have to use a different ligand. So it's getting into cells via a different mechanism.

Okay. That's helpful. So as you just described, you have these modified RNAs, which maybe are not really RNAs anymore, but modified kind of to do their job, you're adding this ligand, I kind of like to think of it as like a tag or maybe a key, you're attaching to the molecule.

Yes, it's like an address.

Yes, like an address, yes. Okay. So in the liver, that's GalNAc, and you all have made a lot of progress there -- progress there in that front. So yes, talk to us about the CNS. And I guess, maybe let's start off with in terms of the modifications that you make to the RNA itself. Are there -- is it the same modifications, do you kind of have to like start all over when you think about targeting the CNS or different organ systems as you go along the way?

Yes. So there are some slightly different modifications. In particular, the ligand that we use is a long -- it's sort of a stick sort of a fat on there instead of a sugar and that allows it to distribute throughout the brain. And then some of the modifications are a little bit different on the molecule itself in order for it to be able to similarly survive in the right cells and get to [ Ago2 ] within the different parts of the brain. Now the brain is much more complicated. There are trillions of neurons instead of billion. So there's an order of magnitude difference in complexity.

Got it. Okay. That's helpful. And another really like taking it back to the basic type question for me before I turn it over to Joohwan on my team who's the expert to dig into the weeds a little bit further. But why is getting stuff into the brain so difficult?

Well, it turns out that the brain has evolved to not want to have bacterial infections or viral infections or other things that might get into your bloodstream into your brain. And so there's something called the blood brain barrier, which is essentially like a big wall that only lets a few things through. Certainly not very large charge molecules, which our RNAs are. You might imagine that viruses and bacteria have their own RNA and viruses, in particular, the one thing they'd love to do is to take over your cell machinery and make more virus. So you have an immune system that recognizes that and then you have this thing called the blood-brain barrier that basically excludes anything that's too big to charge. And so that's been -- even in developing small molecules like the types of things like aspirin in those kinds of traditional small molecules, you have to design them very specifically to get across. Antibodies can go across, but only in small quantities. And these without some sort of an engineering because these RNAi therapeutics are large, they don't really go across it all on their own. Now we've got technology now that is allowing us to shuttle some of them across that blood-Brain barrier. But currently, right now, we're actually going straight into the spine in order to get across that blood-brain barrier and into something the CSF or the spinal fluid.

Got it. Okay. Super helpful. I think I'm going to turn it over to Joohwan now to dig into maybe that [fat]. I think you mentioned attached to the molecules.

Thanks, Whitney. Appreciate it. Yes, I'd love to dig more into the address as you called it, or the ligand. Can you tell us [indiscernible] by just telling us a little bit about Alnylam's approach to CNS delivery with respect to that ligand. Is that fact?

Sure. So we use something called C16, which means there's 16 carbon chains, right? And that's -- there are 16 carbon chain fatty acids found again, naturally occurring in your body and they traffic in different ways and can get into cells in different ways. And one of the reasons that we wanted, we were looking actually for something, do you imagine, as you go into the CNS, as I mentioned, there are different cell types. There's neurons, there's something called Astroglia and Microglia versus just mostly Hepatocytes in the liver. And most of the diseases, there's huge unmet medical need within the central nervous space, diseases like Alzheimer's and Parkinson's, Huntington's disease, to mention a few others. And what -- when you look into the targets that are validated genetically in the human genetics of those diseases, what you find is that a lot of those targets are expressed in almost all of the cell types. So there's something called almost ubiquitously expressed. And so when we were looking for a delivery strategy was a little bit different than just trying to target Hepatocytes. Now we had to try and target multiple cell types. And so in doing that, we found, and we screened through a bunch of different fatty chain lengths and places to put them on the molecule. And we found that C16 was very effective getting distribution throughout the brain after a single intrathecal injection and getting into the many cell types that we were trying to target in a lot of these diseases with high unmet medical need.

And it sounds like you touched on this a little bit earlier, but in terms of testing out different versions of C16, how much of optimization kind of went into it whether were you looking at C15 or C17, or is that something that will it just found that C16 was the most ideal?

No, we looked at a wide variety of chain lengths, wide variety of where do you actually attach it to the molecule. We looked at a lot of iterations of that to try and find the one that we thought would be both safe, effective and also easy to scale. So the other thing that you have to be able to do is the molecule is, it can be the greatest science in the brain in the world, but if you can't manufacture it because when you're trying to put it into a while solution, it all falls out then it's not probably going to be a good drug. And so there are multiple things that you're optimizing for when you're trying to think about making therapies. And so we went across all of those different types of modifications. It wasn't just fatty acid change that we've tried. We have also tried other ligand receptor pairs and continue to work on that, by the way, because there are cases where maybe you're only going to get into one cell type in the CNS, one brain cell type. And so we continue our efforts there to sort of target some of them. But for many of the diseases like I said, you really want broad distribution and C16 turned out to be give us the ability to rapidly knock down the target throughout the CNS, including in deep regions of the brain like the striatum that typically have been hard to reach with antisense technology.

And just to really kind of focus on the differentiation point of this, can you tell us just a little bit about what normally has been or what would happen if you were to inject just naked siRNA intrathecally or through ICV?

Yes. So naked siRNA by itself, it doesn't distribute. It doesn't have properties like binding protein. So it tends to stay closer to the side of injection. But also without the chemical modifications, we get degraded very rapidly. And so while you'll see a little bit of activity with that maybe locally, maybe a little bit of the spine depending on the dose, it really wasn't optimal for a therapy where what you're really trying to do because it's an intrathecal dose at this point, we will get across the blood-brain barriers, just a matter of time. you want to do that injection as infrequently as possible. And that's where the stabilization chemistry comes in, so that you're looking at a once in 6 months, maybe once in 9 months or a year, possibly depending on the target injection into the CSF.

And if I was hearing this correctly, the stabilization chemistry that kind of fills into this C16 modified siRNA. It's different based from some of the previous generations of siRNA in order to enable that sort of protection from new places and that distribution.

It's a lot of the same chemistry, but there are also a couple of additional modifications that you use in order to get activity across the brain. And so it's very similar, so we can take all the learnings from the liver and we didn't have to start all over from ground zero, but there still were some learnings that we've had to do along the way for all safety and the efficacy.

Got you. And I guess looking at differentiation again a little bit more. When you're thinking about like large molecules or some of the antibody conjugate approaches that have been used for CNS delivery, how is this differentiated as opposed from some of those ventures?

Yes. So I think so far -- so we have active -- we have a lot of activity with antibodies to try and get across the blood-brain barrier as others do. couple of different targets there, transparent receptor, other targets that seem to get some material across maybe you get a 50%, 60% lowering of the target across the brain, pretty high doses and a lot of doses, IV in order to get there. So there's definitely some optimization work that needs to be done on that side. So I said it's not really a matter of if, but more when as we continue to optimize that approach. But right now, the go-forward approach is to get 90-plus percent knockdown throughout large regions of the brain is to really go in into the intrathecal space and give a dose there.

Awesome. And in terms of the regions of the brain that you are accessing, can you speak a little bit on that? And are there any reasons the brain where there might not be a substantial uptick in...

Yes. So we -- it is a little bit dose dependent, right? So a little bit higher of a dose, we'll push it a little bit further into the brain regions. But in general, we have good expression, good knockdown and rapid knockdown even within deep regions of the brain. I would say the one cell type, I think it's the Oligodendrocytes that tend to have a little bit less silencing. But in general, across the broad level of cell types, and in different regions of the brain, we have really nice silencing. And again, it's once in 6 months -- what we once in a year injection.

And I just kind of follow up to that in terms of Oligodendrocytes. What is the significance there? Is that something that you want to target more or less or...

It depends on the disease and the disease area, right? For the targets that we've chosen, luckily the Oligodendrocytes aren't highly involved. And again, it is also a matter of dose. So if you move the dose up a little bit higher, you can get more into those cell types.

And of course, you guys have done quite a lot of modifications in order to have a great safety profile. Can you speak a little bit on that in terms of the additions that you've made or safety concerns [indiscernible].

Sure. So luckily, within our clinical experience so far, we've had good luck with ALN APP and our preclinical experience to date these things are highly effective, and we're very happy with where we are in the safety profile. And I think when I think about safety, I think about 2 things. One, I think about dose and I think about dosing interval, right? So how much dose do you have to give because safety is always a product of how much drug exposure do you get? And then how frequently do you have the dose. And so very happy with where we are with very low doses of 50 to 75 milligrams giving us up to 6 months. So that's a very low dose to be giving very infrequently. And so you'll find that as you dose of drug less and less frequent, you're going to have a nice safety profile. And so that's one, I think, of the benefits of our platform, but there are some of the others that are out there is that we've been able to show this remarkable rapid knockdown of target but then also there's really remarkable duration across all areas of the brain.

And lets talk now move towards the data side. I think you've already referenced a little bit about it, I love to hear more about it for sure. How did you go about sorting through the different targets and diseases and designing where you really want to go first with the innovations?

Yes. I mean, so for me, my background actually starting early on -- you can take it all the way back to -- I talked about that [indiscernible] I used to do genetics and that's where RNAi's was discovered, and then I went on to look at human genetics. And so the way that I think about targets is I look at human genetic data first, right? And so if you think about the pipeline that Alnylam has had historically, all of our targets have had been based in validation by human genetics. And what do I mean by that? So let's take PCSK9, which was one of our first deliver targets. There are people out there that have so-called gain of function or too much PCSK9. They have more than normal. They have very high LDL cholesterol levels, they get heart attacks early. Then there are people who have half as much as normal. So they have a mutation in 1 of their 2 PCSK9 genes. Those individuals have lower LDL cholesterol on average and tend to be protected from heart attacks and strokes, right? And then there are a couple of individuals that were identified 3 or 4 that have no PCSK9 at all. And they have an LDL cholesterol of 20, they'll never get a heart attack and they appear to be healthy and happy. And so therefore, right? That target, to me is genetically valid. That means that I know that if I can knock it down 50% or more, I'm going to have a benefit in the long run in the disease of Hypercholesterolemia. I know that if I go too low, well, there's people that don't have any of it. So that's okay. And I know that it's causative because the people who have too much always get the disease, right? So those are the kinds of targets that we think about, and if you pivot to our first target, which is Amyloid Precursor Protein or APP. There are human genetics that suggest there are people out there that have mutations and APP. So they have an irregular APP where it has too much function or there are individuals that have multiple copies of APP. So they have too much function. Those individuals get early onset Alzheimer's disease. And so they can get Alzheimer's disease, unfortunately, in their 30s and 40s. And so we know there that lowering that back towards normal or below normal should be helpful. And that's been some of the hypotheses around it for years. Now nobody has really been able to do it in the way that we're doing it. So people have had antibodies to these things called Abeta40 and 42, which are fragments that come off of APP. And those fragments accumulate in something called plaque in the brain and those are so-called Alzheimer's plex. And you have antibodies that go and buying those and try and clear them. okay? And so what we're doing instead of trying to do that, we shut off the production of the protein and let the body clear it naturally. So there's always this mechanism of all of us are making those fragments, they're trying to deposit in our brains, but they get cleared. Over time, as we get older, it seems that, that clearance mechanism is less robust. So it's like a little bit like a partially clog sick right? So first, you've got water running and got water going through the sink, it gets partial to the cloud, that gets too clogged, eventually, the sync fills up and overflow. And what we're doing is really turning off the faucet and allowing the body what's left of that drain to sort of click, right? And so that's the their purpose. It's very similar to what we've been doing in the TTR space, where there's a protein that misfolds. It aggregates in these things called clog in this case in either in the heart or in the neurons. And then you shut off the source and Lo and behold, the tissues start to clear.

Fantastic. Whitney, I'll turn it over to you if you want to talk about APP.

Excellent. So yes, that was a really helpful kind of introduction, I think, to how to think about APP. And you mentioned -- you talked a little bit about what goes wrong, but before we get there, I guess, what do we know about the APP protein. What does it normally do in patients who don't have an issue, I suppose, and then we'll kind of dive in a little bit more to what happens when they do.

Yes. So what we know about APP is that normally, there are -- it's actually a very complex protein. It's got a portion that's inside the cell and a large portion that's outside of the that outside portion tends to throw off these peptides Abeta40 and 42. 42 in particular an Alzheimer's disease seems to be an issue. And we'll talk about a different disease later called CAA, which is, I call it -- it's almost Alzheimer's disease of the blood vessels instead of the portion of the -- instead of the brain itself. And so what we know is that in the disease, like I said, if you have too much of it, there are other mutations that in a different gene called Presenilin's that actually clip that APP protein prematurely, gives you more Abeta42, they also get early on to an Alzheimer's disease. So what we know is that, that protein is intimately involved in the development of Alzheimer's. And so the hypothesis is that if we lower it, both the intra and the extra side of the domain that we can help with that disease. The Intracellular portion is not as much known on what it does, but we do know that it aggregates inside of neurons. And it makes some of the structures within those neurons irregular. So we also think that lowering that could also be quite beneficial in these diseases.

Helpful. And then, I guess, relative to CAA, you kind of just mentioned it, but in Alzheimer's, so you're saying the plaques or the clogging up of these proteins happen in the brain in the neurons themselves, but as you alluded to in CAA, it's happening -- same thing happening, but just in different cell types in the blood vessels. Is that right?

That's technically right, although it's not that straightforward in that up to 20%, probably more because we haven't been doing brain imaging that long as a diagnostic. A lot of it is a mixed disease, where you will have plaque in both places, right? And so part of that is getting picked up now where the antibodies that have been recently approved in the news that bind Abeta plaque, the individuals who actually have CAA or plaque in the blood vessels actually don't tend to do so well is something called ARIA, which is -- in the brain, which they're actually not trying to avoid. And so there's a lot more imaging going on, which is actually finding more CAA patients or the more of a mixed phenotype. Our data in preclinical animals, there is a rodent model of this disease where they get plaque in both places. So they get it in the blood vessels and they actually have the -- what happens when you get it in the blood vessels, blood vessels become fragile and you actually start to bleed in the brain. First, you have a small bleed called micro bleeds, and then eventually, you can have much larger -- and they tend to lead to much larger bleeds or stroke. And so even the micro bleeds can impact, as you imagine, if you have bleeding in the wrong part of your brain, it might impact cognition, right? And so in our preclinical models that mimic both when we lower APP in those models, we actually clear out of both places. And so we cleared the blood vessels up. We actually prevent the bleeds. And then we clear that plaque out of the neurons as well and the neuro starts to look more normal.

Got it. Got it. Okay. That's helpful. Once again, this is sounding a little like TTR to me. And like there, we were talking about heart versus neurons and maybe some patients have mixed disease. Here, it's sort of the same, but all on the CNS, that is interesting...

It's very similar black and plaque, right? Plaque is just basically the accumulation of misfolded proteins, a little bit like clogging a sink, it's like a hairball you stuck down in the sink sometimes it's partially a clog. Sometimes it's you try to keep it for being fully clogged.

Yes. Yes. Okay. That is helpful. And it seems like a fairly straightforward problem, I guess, as you set it up, related to APP, but has one that a problem that's been hard to solve so far, at least from an antibody approach perspective. So I guess, can you talk us -- talk to us a little bit more about kind of what makes this disease hard to target from an antibody perspective. You touched on it a little bit in the sense of kind of your approach is shutting off the faucet versus trying to, I don't know, chew up the hair clog maybe, but there's different antibody approaches as well. So can you help us understand maybe just where are the challenges and why maybe your approach which sounds to me like kind of going further upstream maybe is the better way to do it.

Yes. So it is upstream. And again, going back to the analogy of sort of mopping the floor right? That's kind of an antibody trying to mop it all up that's already there versus shutting off the faucet. And so the antibody is, number one, they need to get across the blood-brain barrier. You're not putting them in directly. Number two, they tend to bind these plaques and create inflammation. And part of that is creating inflammation to sort of clear. And so the good news is that they are able to clear the plaque out. The bad news is that in the process, sometimes you can get an inflammatory response. The other aspect is that I talked about the sort of extracellular so that's the stuff that's outside the cell that's going to other cells, the antibodies are good at touching that. The stuff that's actually going on inside the cell, the antibodies don't go inside the cells, right? And so what we're able to do is now shut off both the stuff that's being secreted and goes outside the cell, those fragments that are coming off the surface when APP was cleared, but also that intracellular domain that we think is very important also in disease.

Got it. That's helpful. And a related audience question here. How is your mechanism different from BACE inhibition.

Yes. So again, the BACE inhibitors, if you look at what they're doing, they're affecting the cleavage of Abeta40 and 42, but they're not actually touching the intracellular domain of that protein. And there are other peptides that come off besides 40 and 42 that the BACE inhibitors didn't change. The other thing with the BACE inhibitors is a little known probably if you studied the BACE field and I was a personal [Myers], one of the things I worked on was BACE inhibitors. They don't just hit APP. So they have a lot of other substrates that they go off and cleave, and so it was never really sure whether some of the effects of those were due to the cleaving something else off target.

Interesting. Interesting. Okay. Okay. So just so I understand, and I am a little bit fuzzy on -- or I have not studied BACE inhibitors, let me say that. But so BACE inhibiting or BACE inhibition is sort of preventing the cleavage to prevent the fragments performing. Is that right?

Some of the fragments but not...

Some of the fragments performing. Yes, yes. Okay. And you're kind of silencing APP would just be preventing the production of the thing that gets cleaved to begin with.

Yes, and not just the stuff that's being cleaved, but also the other end of it. So we get...

Yes, yes, yes. Yes. Okay. So that makes sense, as you said, kind of even further upstream. Got it. Okay. And then I guess, what's different about the Alnylam APP approach versus maybe other silencing or ASO approaches for this disease?

Yes. So there aren't any other that I know, either Antisense Oligos or RNAi's APP. I'm sure some will come after we've been able to sort of show proof of concept. The Antisense Oligos have been tremendous for the field in terms of, if you think about some of the diseases that they've been able to treat. SPINRAZA has been a good drug I think one of the things of that class, however, they're full of these things called Phosphorus thyroids. And those modifications depending on the dose, can be fairly inflammatory. And so you see white blood cells go up, which are the immune cells in the compartment. When you look in the central -- in the fluid of the spine, you can start to pick these up. You see proteins go up and a few other things. And so to date, we've not seen that. And so -- and that sort of reflects similar things that you see for the -- in the liver, right, where especially the early days of the Antisense Oligos before GalNAc, they had pretty interesting inflammatory markers go up in the systemic circulation versus RNAi that's been silent. And so I think they are different. They use completely different mechanisms. So RNAi is a naturally occurring mechanism, like I said, it's Ago2 happening all the time, we take advantage of that the Antisense Oligos actually use RNAsh. So they co-ops an enzyme that has a different purpose, right? It's not disable function to do this. And so there's some differences there. When it comes double stranded or as is double stranded, the other one is single-stranded. So I think those are some of the key differences. But whatever can help patients. I hope they're both wildly successful.

Definitely, definitely. Okay. Perfect. That's helpful. All right. So we kind of have talked about figuring out delivery. We've talked about APP itself, which kind of brings us to ALN-APP. What is the target product profile of this as you all set out to develop this program?

Yes. So I mean, I think what we're looking at is to try and lower APP and Abeta40 and 42. And we won't be able to measure the intracellular domain, but we do have all of these biomarkers that we can measure the beta fragment, the alpha fragment in the CSF to know that we're lowering it significantly. And so what we're going to do is we're in the middle of doing single ascending dose to go up and see how far we can lower it so far so good. We've been able to knock it down the 75-milligram has been quite successful. We've dose escalated to 100. And we're in early onset Alzheimer's patients. So we're going to be looking at imaging. We're going to be looking at all of these different aspects of biomarkers. And it's a small trial. So trying to see some differences on imaging and cognition. We'll have to -- you have to say it's going to be very small numbers, but we're looking for signs. And mainly, we're looking it's a Phase I, it's safety and efficacy. So we've been delighted that on our first trial in the CNS, we're seeing 80-90 at a very small dose of 75 milligrams single dose that's lasting out to 6 months. So that's really about as good a profile as I would have hoped going in, given this is the first generation and our first time in this space.

So I'm getting myself because there's some beeping in the background, Apologies if you can hear that. So that sounds good. So you -- as you said, you're in a single dose study, and you mentioned this a little bit in terms of what you're...

We're actually now in a Part B multidose and...

True, true, true. Fair, fair. Yes. Fair points. You have progressed. But thinking about the single ascending dose setting, kind of what you were looking for headed into that. And you touched on this a little bit, but the alpha fragment, the beta fragment, 42 versus 40 for somebody, again, who's kind of newer to the space, like what are all of these different pieces? And what do they tell you maybe alpha versus beta 40 versus 42.

Right now, I'm looking for them all to be lower. I want them all to go down.

Okay. That's easy, that's easy.

My mechanism is to remove the protein, which from which they're all derived, right? So I'm looking for everything to be down. And so that's really what we're -- I love everything to be down, 80-80, 80-90 and see where that is now. And maybe that this drug is going to be effective at an 80-60, 80-70. But in general, if you're thinking about turning off the faucet versus mopping the floor, you want to get down where it's turned off probably to a trickle.

Okay. Perfect. And then kind of going back to sort of the genetic validation of the target that you mentioned earlier, are there patients like can you go too far, I guess, can you -- if you turn the faucet all the way off, is that a problem here?

In the preclinical models and in our safety studies in primates it doesn't appear to be the human genetics isn't quite as clear there. There aren't any people because APP appears to be involved in development early on. This is a case where there may not be homozygous individuals that you can identify. And so that has been one of the questions that we looked at very carefully in our preclinical safety models, and we were able to go to very high doses with our safety models and show over a long period of time that appears to be okay.

Got it, got it. Okay. That's helpful. All right. So kind of going back to then what you did see though in terms of knockdown after a single dose, and you alluded to this earlier, but you were seeing kind of remind us, I'll let you talk about it, but significant percent knockdown at the different doses for a certain period of time. And that all kind of met your expectations or at least check the boxes as you move forward into the next phases.

Yes. So we had up to a 90% lowering at 75 milligrams, continue to dose escalate, given the safety profile we were seeing at that dose. And that continues to go on, and then we started out to do a multi-dose every 6 months to see what the profile of that is. And so that's ongoing. So very exciting. And if you look at the duration of effect, again, after a single dose, we've got this nice rapid knockdown of all of the different biomarkers that we looked at. And really to date, we -- all of the markers that we've looked at from a safety perspective, that's something called Neurofilament Light Chain, which is a marker of how well neurons are doing white blood cells, that's a marker of inflammation and a couple of these other biomarkers they all look right. So very, very pleased with what we're seeing so far and very excited that this type of an approach could be helpful in this disease and also the disease of CAA.

Okay. That's helpful. And then you mentioned Neurofilament Light Chain. So let's dive into it a little bit because it does -- like we hear about it a lot. You mentioned it in the context of safety, but can you tell us a little bit about what that is? You mentioned it's kind of a measure of the health of neurons. But what do we learn from that? What do you learn from that, more importantly? And is it just safety? Or is there an efficacy.

Well, it depends on the disease. So Neurofilament Light Chain is a little bit of a crude measure of neurons dying, right? And so if you think about a disease like ALS, where it's a very quick progressing disease, which you'll find is that patients can have a very highly elevated level of NfL, and when you're going into a disease like that, you may look for NfL levels to come down if you're halting the progression of that disease or having a positive impact. For a disease like Alzheimer's that transitions over a long period of time, even in the early onset, they don't tend to have an elevated -- not high and elevated NfL level to begin with. That's not that many neurons that are going off over time. So there, you're looking to make sure that your drug from a safety perspective in all cases, doesn't make the NfL go up over a long period of time because that might be an indication that things aren't going the way that you...

Got you. Okay. That makes sense. All right. And then moving over to the ongoing clinical trial and in particular, related to some recent our recent announcement related to the FDA lifting the clinical hold, allowing you to initiate in the U.S. at least multiple dosing in the Part B. Can you remind us, again, for people who may not be familiar, kind of what caused that hold in the first place? And how did you convince the FDA to move past it?

Yes. So we started out doing -- you do safety packages. And when you start out with a brand-new platform that we had, we never taken into people before. So we don't really know what dose was going to be effective. And so we can do modeling from what you see in animal models and you model it out and you say, okay, well, what's the highest dose we think would be reasonable for us to have to go to based on our preclinical models. And that dose when we modeled it was somewhere in the 1.2 to 2-gram level, right? And so we had to do a safety study that was able to cover that kind of an exposure and people. Now remember, we are now at 75 milligrams versus we potentially predicted a grant. And so what happened in that study is we dosed them. We didn't know how frequent because we didn't know the duration. So we didn't know we thought it might be once every 6 months, but it could have been -- maybe it would have been once a month, right? And so you have to design safety studies to cover that, too. And so what happened at the top dose in this study is we dosed the animals very frequently at very high doses, now we had a safety finding. And the FDA had questions about that, whereas other regulators in other places like Canada and the Netherlands didn't have the same questions. And so we had to just go back, generate a little bit more data to make them more comfortable. And that's not unusual, especially in the central nervous system and not unusual in drug discovery in general now. They've now allowed us to go to a multi-dose up to 180 milligrams. And you saw the data that we had at 75 and 100. So we think, in general, for the early onset program, where we won't have to go to the doses that they've capped us at, so it becomes really more of a less of an issue.

Got it. Okay. That makes a lot of sense. So yes, you are in multi-ascending doses, as you said. So as you move there, what new are you looking for, I guess, as you move from the single ascending dose, you're looking for knockdown and kind of duration. What's the next question as you move to multiple doses?

Yes. So we're looking at -- again, when you go to multiple dose, you're looking at safety, and you're looking at safety over time. And then you're looking to really try and figure out if you -- as you think about designing a pivotal Phase III trial or a bigger Phase II trial, what dose or what 2 doses do you want to take forward into those trials? And so you're really looking to get enough data on both the safety and efficacy to choose those doses correctly. And so we're getting more data on that. We'll get more biomarker data over time to see if any of the biomarkers that we're measuring are going in the right direction, and that will give us a lot of insight into how to design those later-stage trials.

Got it. Okay. That's helpful. And you mentioned you're doing some imaging work as well. and maybe kind of going back to something we didn't actually talk about, but the early onset Alzheimer's population and kind of what the baseline image of their brains looks like and the potential time course of any changes there. I guess, can you talk us through that? And should there be an expectation that maybe you'll see something positive from a multi -- this current multi-dose study?

It's hard to know because -- so going into the trial, they do -- they are imaging positives, so they have amyloid, but it is a small number of individuals. We're trying to figure out by this mechanism of natural clearance, we don't really know what the rate of change will be. And so that's part of what we're trying to answer in this multidose and get a feeling for before we go into a larger population where you would then look for statistically significant changes.

That's fair. That's fair. Okay. So if you turn off the faucet, but the unknown is still like how big the hair clog is how fast the existing puddle is draining, so that's okay, okay. That makes sense. And understood. All right. That's really helpful. And then, I guess, the ongoing study is in the early onset Alzheimer's patients, as you mentioned. What are the next steps or plans in CAA?

Yes. So CAA is very exciting. So let me explain a little bit about CAA too around the genetics because I think -- so there are individuals, in particular, there's a group in the Netherlands, who they have mutations in APP, and those individuals actually have CAA. They get a very aggressive form of CAA, where they'll get bleeds within their 30s or 40s. It's almost 100% penetrated. If you have this mutation, you will get brain bleeds, and it's interesting on imaging, you can measure when somebody's had a brain bleed because it leaves behind, like it's almost like a little iron deposit. And so you can see those. And so you can -- those individuals are getting imaged all the time. And once they get past having 4 or 5, you know that they're then going to progress towards a severe stroke fairly rapidly in that population. And so again, it shows by human genetics that this protein is involved and causative of that disease. And so when you look in the general Alzheimer's population, there's a large number of individuals who may not have a lot of plaque in their neurons and in their brain, but they have a lot of plaque within their blood vessels. And it's this amyloid that's in this case, mainly Abeta40, not Abeta42, but certainly, it's coming off of that same precursor protein. And so in our preclinical models, as I said before, when we shut off the faucet, we see clearance within those blood vessels fairly rapidly, and they remodel towards normal blood vessels, and they don't have bleeds. And so very excited to start that trial and get into that patient population and really see what benefit we can have. So those patients are also now not really eligible for the antibody therapies that have improved because of some of the risks of those. So it's a very underserved population. And so we're quite excited about it and ready to get started.

And to follow up on that point on the safety side, and you touched on this a little bit, but why shouldn't we be worried about ARIA here, just because of the natural clearing process happening versus...

It's a different mechanism, and it's not going in and binding and inflaming a plaque.

Got you. Got you. Okay. Which I can imagine this is particularly important in the in the CAA patients, in particular, as you said. Okay. And so in terms of next steps, you are planning to -- well, actually, I should say, because you mentioned some Alzheimer's patients have plague in their blood vessels. Are there any -- do you have any hints, I guess, of CAA like early looks at CAA, I suppose, from your Alzheimer's patients, or do you expect to based on the enrollment in the ongoing.

I don't know at the moment. I mean I think as we get our imaging, imaging is actually fairly labor-intensive. But as we get the imaging I would be surprised, we have small numbers on our Phase I given what we think is the prevalence of 1 or 2 of them had some evidence of CAA, I wouldn't be surprised, but I think we don't know that yet here, we'll be imaging upfront to, again, look for individuals that have CAA. And also we'll be working in that -- a portion of those patients will be in the Netherlands.

Got it. Okay. That's helpful. All right. And then as you move into studies here, is there any reason -- or do you -- are you targeting any different level of knockdown or anything like that in this form of the disease, I suppose.

I think it's very similar.

Very similar. Okay. So the dosing work that's happening on the Alzheimer's side is pretty reads through directly to CAA.

It's going to help us, yes.

Okay. Excellent. Excellent. Okay. Really interesting stuff. I guess, as -- again, having solved delivery presumably and kind of starting to get some preclinical -- sorry, clinical data from the Alzheimer's side, as you said, how much in your mind, does that derisk the rest of CNS. So is it like, all right, as we talked about, we figured out C16 we now have shown it works in Alzheimer's. So now we're just off to the races in CNS? Or are there any kind of new technical or platform challenges as you move from target to target.

Yes. For a lot of the targets, I think if you derisk the platform, you need to risk the platform, right? And I think that's exciting because that means that every target behind it will learn from the one in front of it similar to how we have had a platform in the liver and have learned along the way and applied those learnings. I think it's early days in the CNS for us, but I think we're excited about where we are, and we do have other programs coming along in Huntington's and with our partners at Regeneron on SOD1 and ALS, and so we are building a central nervous system pipeline with genetically validated targets that we're confident in and so far, so good with the platform. So I think all of that if I had to imagine a few years ago because we didn't -- we started this endeavor, not that many years ago that we would be able to go in with a single injection into the spine and have 90% knockdown at a dose of 75 milligrams single dose. I would have been pretty happy with that and use that as a basis to build the foundation for CNS platform.

Okay. That's helpful. And you've talked about 2 new INDs by 2025, I believe. And as you just said, you kind of highlighted the SOD1, ALS as well as Huntington's as being IND enabling. So are those the 2 we should be thinking about, or are there others?

Those are 2, and there are a host of those who's behind those. And the timing of those, we'll see. But it is a very similar type of chemistry and a very similar type of platform. So that's one of the benefits of having a platform is that the iteration time of moving from one to the next to the next is quite short. And so the art and the science of it is obviously to let the clinical data in the first couple of programs emerge in time to make adjustments, if necessary, on all of the others right now. We're very happy with, where we are on the safety and efficacy side and really driving forward to build a central nervous system portfolio of drugs that we hope will really change the medical landscape of these central nervous system diseases that really have tremendous unmet need and not a lot of therapies.

That makes a lot of sense. And just to be clear, for the SOD1 and HTT programs, are both of those kind of plug and play in the sense that they are also C16, or have you talked about that?

We haven't talked about it, but they're the same platform.

Okay. Okay. All right. And just going back to something you alluded to earlier, all of these are kind of direct to the CNS dosing, but you talked about looking at our kind of continuing to work towards systemic dosing to target on CNS. So -- and you talked about kind of antibody conjugates in that regard? Have you guys laid out time lines, or I guess how much of focus is the systemic dosing to reach the CNS for you.

Yes, we're working on that ourselves and also mainly with our colleagues at Regeneron, who are the best antibody experts in the world and who better to have a partner when you're trying to make an antibody to get across the blood-brain barrier. So we're very confident that we'll have a solution to that. And that will be the next generation of these molecules that come through. And what will be interesting about it is that there's only some aspects you can change out the delivery of it, but you've learned a lot about that sequence in the SI. The Si side is -- could stay the same. And so you imagine that it's sort of a life cycle innovation on it. And just going to make sure that the dose is low enough that you're not having to do consecutive high-dose IVs because then that becomes if I think about a single intrathecal dose once every year versus a bunch of IVs, we'll see how that checks out over time. So we're working very hard to bring the dose levels down and to find the right delivery solution to get us across the [indiscernible] .

That makes a lot of sense. Okay, perfect. And maybe just a bonus question at the end, thinking about other tissue targets, and you sort of highlighted some of this at your R&D Day as well. But CNS is just one additional tissue target, and you're working towards some others. So can you remind us, where else are you focused for the -- in the nearer term, I can imagine you're probably looking everywhere, but for the nearer term. And kind of how are you thinking about accessing those from a technology standpoint?

Sure. When I started out, to come to -- Alnylam I left a small molecule company, and people over there told me you're crazy, it's too big, like, well, it's an engineering problem. I love engineering problems happen. So every tissue is an engineering problem to be solved, and that's really how do you get it into that tissue. Beauty of RNAi, I talked about a cup of coffee. Well, anybody went for a run, guess what, your microRNAs went up and down in your muscles. And so the system behaves in every organ that we've tested it in so far. So it's really about how do you get delivery into that organ. And so we're working very hard across different tissues in the body where we think we have really high-value targets that will help pensions, right? And so that's how we look at the lens of where do you go first, second and third. We have ongoing efforts in muscle and part and a bunch of other tissues.

Very interesting stuff. Perfect. All right. Well, this has been really interesting. I learned a lot. I think we -- as I said at the beginning, we're very excited about the CNS and the additional tissue types, as you mentioned as kind of the next wave of where this platform could go. Let me not -- we are also excited about HELIOS-B, and so we look forward to seeing those results in due course. But I think the minute that, that news is behind us. We'll all be talking more about what's next. And so we really appreciate you taking the time to walk through this side of the platform with us, and thank you all for joining us as well. We'll be talking to you all soon in our next installment of the series tomorrow. So thank you again, Kevin.

Thank you for inviting me.