Very much everybody. It's my pleasure to be here with Holly Kordasiewicz to talk about the Ionis neuroscience pipeline. Holly, instead of asking you to just give some prepared remarks, do you mind if we just get right into it? Is that cool?
Sure.
Okay. All right, sweet. Thank you. Okay, I am super excited for this tau readout coming up later this year. I think one of the things that's hard to really answer in research about tau is the whole correlation causation question.
which was a question with amyloid for a long time. Obviously, the correlation piece was somewhat disproven, as it relates to tau, can you just talk about tau as a target and put it into context, then the evidence for tau as a causative part of the biology of Alzheimer's disease?
Yeah. First we can start with the correlational pieces and the pathology. Tau pathology actually correlates better with cognitive decline than amyloid does. For there, that bit is already stronger. There's two ways that tau has been implicated in Alzheimer's disease. The first is, of course, the pathology and that correlation with the pathological spread of tau correlating with cognitive decline, also the function of normal tau. The endogenous tau, if you lower that, you take that out in animal models, you can protect against AD and Aβ driven AD. Those two elements taken together suggest that tau is an important target for Alzheimer's disease.
Okay, great. Sorry, I muted myself. Okay, makes sense. I guess on the safety side, what do we think the role of tau is actually in healthy brains? Do we think it has a role in adults? Do we think it's going to be okay to lower all forms of tau at the gene level?
Yeah. The short answer is yes. We've seen absolutely nothing to date that suggests that partial lowering wouldn't be tolerated. There's nothing in the animal models or genetics. In fact, as I mentioned before, if you take a full knockout mouse with no tau and you cross it with an Aβ mouse, the Aβ mice do better.
Yeah.
They have less disease. It also protects against excitotoxicity, so protects against genetic epilepsy. It also protects against chemically driven excitotoxicity. Tau full 100% knockouts actually have a protective function in the animal models. In human genetics, loss of function isn't associated with anything, any diseases. It's all gain of function or splicing mutations that are associated with tau-driven disease. There's really nothing to suggest that there's going to be an issue with tau lowering. That said.
Yeah
Of course, we want to be absolutely careful and continue to monitor this as we go through our clinical trials.
What do we think causes tau misfolding and aggregation, and when do we think that happens in the disease?
Excellent questions. We don't exactly know. It probably happens early, and we do know that it shows up early in the disease course, and that it progresses throughout the disease course. We also don't know how much is driven by tau pathology, and then, as I mentioned, that other role of tau and the endogenous tau function leading to potential increased excitotoxicity, and which of those elements are contributing more to the disease. It's possible it's both. It's possible it's one driven at a different time period than the other.
Do we think it happened, like with Aβ, right? This is one reason why I'm really excited for the presymptomatic readouts. We think that that starts at zero minus 10 years, whichever way you want to directly do it.
Yep.
Do we think tau comes into play presymptomatically too? Does it really come into play when you have MCI appearing?
Yeah. You can have tau pathology early on. It is typically thought of as happening after and downstream of Aβ, but that's not consistent across all patients.
Okay.
It's not that simple of a pathology, and it's going to be different in different individuals based on what their underlying disease is. If it's more tau-driven, if it's more Aβ driven.
Right
If it's more inflammation-driven, what their APOE phenotype is.
Right.
All of that is going to change this.
Was it surprising to you that the antibodies for tau showed nothing? I understand that they don't lower intracellular tau, but the whole idea of stopping the spread from area to area, are you surprised that they didn't even show a trend?
No. There is some hints of stuff in the antibodies.
Is there?
I don't think we can diss them that much.
Okay.
No, because trying to catch the tau pathology as it spreads between neurons is really hard to do. We don't know exactly how it's transmitting from neuron to neuron. If it is in that extracellular space so that it can be captured, if it's moving through vesicles or other means between those neurons. To actually be able to capture that, to be able to prevent it from getting into those neurons completely so that you prevent that spread is hard to do. If you think about how tau seeding works, if you get a small seed in that neuron, then all that endogenous tau that's in that neuron can then misfold. That's been shown over and over again in lots of different systems.
To prevent all pathological tau from entering into a neuron, to prevent that neuron from then misfolding its own intracellular tau is really tough to do.
Yeah. Okay. Maybe talk about the data you have on PET, how confident can you be that you're lowering tau enough in the right areas of the brain that matter with intrathecal delivery of an oligo?
Yeah. In our phase I/II study, we were able to look at tau PET, we were able to actually see a reversal in tau PET. Not just preventing a progression, but the tau PET that was there before treatment went away. Because we were able to see that reversal of tau PET, that gives us confidence that we're targeting the brain regions that matter. All the regions that had tau pathology have that reversal. Because of that consistency, it gives us a lot of confidence. Now, that said, we know how our oligos distribute. We have a lot of preclinical data.
We have a lot of experience with IT oligos. This is something, getting distribution to the important brain regions was something we were confident going in. The tau PET is just really nice confirmation of that for the technology, also exciting because it shows that reversal, which had never been shown before in an AD patient.
Right. Yep. No, it's amazing.
Good.
You lowered tau PET maybe 60% in the hippocampus. Is that the right ballpark?
Yep.
With amyloid, I think we learned over time that 70% is not enough, right? You remember the whole aducanumab post-hoc thing.
Yep
which was a mess for a variety of reasons.
Yep
with amyloid, it's like you got to obliterate it. Do we have confidence that 60% is enough in the key area of the brain? What would you point to?
That's the question that we're answering right now in the phase II study, in the ongoing CELIA study. That's the exact study we're answering. We have looked, Biogen did a really nice post-hoc analysis from the phase I/II study where they compared the clinical data to an external control. They used the TANGO studies. They did a propensity score matching, comparing it to our BIIB080 treated, our tau oligo treated individuals. All those individuals, it favored BIIB080 in terms of all their cognitive performance and functional performance. That said, it's a post-hoc analysis. It's small Ns, but that was enough to give confidence that this could be in the right ballpark for having a benefit.
Yeah. Okay. There are every six and every three-month arms in this study.
What would you say your confidence is that every six is going to be in the right knockdown reduction range, given that it's been so hard to get uptake with these Aβ antibodies, and people obviously wonder about intrathecal?
Yeah. You just got to that question that we were asking, is how much tau do you need to lower, and to what extent to have the biggest effect size? That's the key question.
Right
That we need to answer. We have a lot.
Right
Of experience dosing drugs in the CNS. We also.
Right
know that our tau oligo lasts a long time.
Yeah.
From the phase I/II study, we stopped dosing, then we looked at a six-month recovery, it was flat in the high-dose groups.
Okay.
There was no recovery over those six months.
Interesting. Okay. That suggests every six months should be-
Yeah
good enough.
We had that gap between when we stopped the study and when we started the LTE-
Right
specifically to look at this, and it didn't recover. With that gives confidence that it could be enough. Again, if you need really significant reductions, you might want that quarterly dosing to get stronger reductions. That's why, just to make sure we had all bases covered, we have both included in the study.
Are you doing a loading dose for the every six arm?
No.
No.
No loading.
No loading dose. Okay. Interesting. Anything else you'd add on tau?
No, just that it's really exciting. The data readout is mid-year this year.
Yeah
I can't wait to see it.
Yeah. I guess, I have Biogen on the panel tomorrow, and I'll try to probably unsuccessfully pin them down on the bar for success. It's a big study, right? It's placebo-
Yep
controlled. It's a long study.
Do you feel like this is well-suited to really answer the question and also show something statistical if you have an effect size that's in the ballpark of Aβs?
Yeah. This will do what it was supposed to do, which is to replicate the biomarker data that we saw in the first study and then allow us to design a phase III study. It has all the different elements to it. We have the tau PET, we have the CSF markers, we have, of course, APOE carriers, non-APOE carriers, so that we can look at the different patient populations and how that they respond, so that we can then design subsequent studies if there are benefits seen here. As you mentioned, it's a big study, it's a long study, so I think we have the right timelines and we have the right patients to be able to ask the question.
Yeah. Were patients in this study, is it basically the same population as like lecanemab and donanemab?
Yeah, it's the.
Phase III, phenotypically?
Yep.
Okay.
It's the original populations.
Okay. All right. I can't wait.
I know.
Maybe switching gears to Angelman, do you want to just sort of set the stage and talk about the phase III and how it's progressing?
The program's progressing really well. We recently refined the phase III REVEAL design to focus exclusively on the 80-milligram cohort. We initially had 2 dose levels, but based on the positive efficacy and favorable safety tolerability profile from our phase I/II HALOS study, which was an open label study that had both dose levels in it, we decided to focus on the higher dose level. This means we can now achieve the objectives for the study with 158 patients instead of the 210 we were originally planning, so a little bit smaller. That enrollment for that study is expected to complete this year so that we also have FDA Breakthrough Therapy designation, that means read out the following year in 2027 because it's a 12-month study.
Makes sense. We'll get the Ultragenyx data first. How are you thinking about the read-through from that readout to yours?
Yeah.
What are the limitations of that read-through?
Yeah. For Ultragenyx, the thing I'm most looking forward to seeing is the placebo effect in this patient population.
Yeah
We really don't have data on 12-month placebo effect in these individuals. Of course, we can model it, we can do all sorts of things, and we did do that as we powered our study, but to really see what that placebo effect is at 12 months will be really nice. That's an important question for the whole field, not just for us. Other than that, the molecules are very different. The Ultragenyx molecule is a different chemistry. They have dose-limiting toxicity, so they're dosing it extremely low. Because of that, I think the read-throughs are going to be limited.
Yeah
the placebo effect.
Beyond the open label data from both companies, which obviously looks directionally supportive, I guess, asked another way, right? With tau, you can look at tau lowering. With SMA, you can look at other biomarkers. In Angelman, do we have any hard evidence or anything we can ground ourselves in in knowing that you're normalizing the actual underlying genetics?
Yeah. Again, we have to go back to our knowledge of drug distribution, what we know about the preclinical modeling, how we understand our CNS oligos, which we have a lot of experience with. Based on all of that, we are at the mid and the high dose levels, achieving nice distribution to all the brain regions to get that restoration of UBE3A levels. We've been able to do that for other programs where we do have nice CSF biomarkers for, so I'm confident in that that's going to translate as well. That's also supported that we're getting the right target engagement and domains by all what we're seeing in the clinical endpoints. The important thing from our HALOS study is that we're seeing benefits across domains and across tests. It's not just one test, we're not cherry-picking any data.
If you look at the Bayley, which is physician administered, if you look at the Vineland, which is parent reported, if you look at the CGI, which is physician reported, all of them tell us the same story. They tell us the same story, they tell us the same across different domains that we're having a benefit. It's that consistency in the data that's giving us confidence here, since we don't have that very hard, crisp biomarker that we have for some of the other programs.
Yeah. Okay. Makes sense. You want to talk about the endpoint selection for phase III? I guess maybe just in the backdrop of this, there's a lot of consternation right now about the FDA and rare disease and alignments and is this align-- Whatever. Maybe just talk about your regulatory dialogue. Obviously, you're doing something different than Allogene and your confidence that this is kind of truly locked down.
Yep. Yeah. Just to remind everybody, our primary endpoint is expressive communication as measured on the Bayley-4. That's a physician-administered assessment. We chose expressive communication because it is by far the number one thing that is most important for parents and caregivers. It is also something that if you're administering the Bayley, the individual has to communicate during the testing. They have to use gestures, sounds, words to score on the Bayley-4 for expressive communication. It's a very crisp endpoint that's going to be less subject to learning or bias or placebo effects or things that can happen than some of the other endpoints that are variable. It being a crisp endpoint, it being the thing that matters the most, and it's also the thing that we had the biggest effect on in our HALOS study.
It also has the lowest natural history, so there's really no change in expressive communication over a year time point in this Angelman patient population. It gave us the biggest delta. Taken together, it's the obvious right endpoint for us. We proposed this as the primary endpoint to the FDA. They agreed with no debate. One thing they did ask us to change is the Bayley-4. We had been proposing to use GSV scores. It's a score that has to be imputed, because of that, they wanted raw scores. We knew that they were going to likely ask for this because we know they didn't like it going in. We conceded to that. They also asked that we not include parent input on the test, that we really do just focus on the physician-administered assessment, which is fine.
We agreed with that as well. Everything that the FDA recommended, we agreed with. These were things that we knew we were probably going to have to concede to going into the discussion. We were happy to make those concessions. With that said, we also have a controlled study. This is a controlled study, so we have that-
Yep
head-to-head control. You mentioned some of the other debates going on. Those are trying to use natural history comparators or external controls. This is gold standard controlled study in the patient population using the endpoint that's most meaningful for patients, doing it in a way that's directly consistent with FDA guidance and their feedback. Because of that, I'm very confident in where we're at. We've had a long, lovely relationship with the FDA, very collaborative, going all the way back to the days when we ran the SPINRAZA trial for SMA.
Yeah.
I'm confident in where we're at and that we have the right study design to be able to answer this question definitively.
Yeah. Okay. That makes a lot of sense. Do you want to just talk about, again, you mentioned the point on placebo arm, right? That being one of the natural questions. What's the effect size implied by the phase I, II versus natural history? When you did the powering in phase III, how much cushion did you give yourselves?
Yeah. We were pretty conservative in our powering because as we discussed about that placebo that we don't know. We do have 90% power to see the effects that we're expecting based on the HALOS data, we have powered that fairly conservatively, as I mentioned, because of the HALOS.
Okay. Not giving numbers around it.
No, we're not giving numbers.
Okay.
No, this is a pretty competitive space.
Is it? Okay.
Yeah.
I'm just joking. Okay. Sounds good. Anything else you would want to add on Angelman before I move on?
Yeah, just two other things. We did add the under two cohort to our HALOS study. That will be open label. This is a really important patient population because these are the individuals who are first diagnosed. These are our youngest people in this patient population, and the ones who could potentially have the biggest benefit because this is neurodevelopmental. Very excited that that enrolled incredibly fast and is moving forward. We also have our CHAMPION study that we've announced. This is to treat individuals with UPD, ID genotypes. This is the final genotype that we haven't treated yet. We've treated mutation and deletion patients in both our HALOS and REVEAL study.
You can also get Angelman syndrome from UPD, ID mutations. Those we're going to be capturing in the CHAMPION study so that our total data package will have treated all age groups and all genotypes with Angelman syndrome when we go, hopefully, with a positive study to file.
Yep. Okay. That sounds good. Maybe just switching gears. Do you want to talk about the platform? Obviously there's been this explosion of the number of companies that say they have a Brain Shuttle.
Maybe not all of those are of equal quality or realness, but you guys, I remember during your R&D day in COVID, right? You had even animal data for transferrin and muscle, and that was a long time ago. Where is Ionis in this space, and what are you doing to make sure you don't fall behind here?
Yeah. We're using, as you mentioned, novel targeting moieties and conjugation platforms to broaden our reach of our RNA medicines to new tissues, including muscle, also the brain. We've exclusively licensed the Vect-horus VHH Nanobody platform that allows us to systemically deliver our RNA therapeutics, both ASOs and SIs, across the blood-brain barrier using transferrin-mediated delivery. That's work in progress. We shared at Innovation Day last year a really impressive non-human primate target engagement data. It's the best that I've seen out there, so very happy with it. We're moving that forward into the clinic.
Is there anything more specific you can say about how that would fit into a collaboration with Biogen? Do you need them to be on board with this to move it forward for like-
We're-
tau or one of these targets?
We're pursuing it now for an Ionis wholly-owned program. If it is with a collaboration target, we would, of course, need the partners, and we'd be able to explore this as well as anything else we're working on. We're also working on Bicycle technology, this is really exciting because these are really small peptides. We've already advanced this clinically for muscle targeting, and we're working on this for BBB as well. The exciting thing about that is we could potentially get it down into a subcu auto-injector, which would be incredible.
Yeah
For neuro.
That would be huge.
That's what we think.
Okay. Is that something you're working on for tau as well?
Yep. We're working on it for all the targets that are interesting for the CNS, of course.
Okay. I got one question from someone who's listening in on Angelman. What would the Bayley-4 data in HALOS look like if no caregiver parent input was allowed?
Yeah. Fortunately, we have the Oak Hill Bio data that was published uses the Bayley-III, and they see a very similar response that we see. The Bayley-III doesn't use caregiver input, it's a very similar response without it, which is why we were happy to concede to that.
It's not analysis you can perform with your own data?
No, it's not. Not well.
Why is that? Okay. Just like the scales don't work that way?
They don't capture it as consistently as they should, the sites and everyone who's doing the test, exactly what came from the administration and what came from the discussions with the parents. We are capturing that, of course, now in the REVEAL study, but in the HALOS study, we hadn't separated that out and requested that information while the test was being administered.
Right. Okay.
Asking for that information post hoc is not useful.
Right. I understand. That makes sense. What CNS program or programs have we not talked about that you think at this panel a year from now or two years from now you think could be a major focus for people?
Yeah. Oh, my goodness, we have so much stuff going on. We haven't mentioned zilganersen for Alexander disease. We had a positive phase III there last year. That NDA has been submitted, and we're looking forward to a launch later and approval later this year. That's really exciting because that'll be our first independent neurology launch for Ionis. Of course, it's always wonderful when you have disease-modifying positive phase III data in a-
Yeah
population that doesn't have anything. That's something to see and watch as we get through the approval and then the launch. There's also salanersen. This is in partnership with Biogen. This is basically like a yearly SPINRAZA. It's using our new NMA chemistry to increase the potency of splice-modulating oligonucleotides. That's starting pivotal studies this year. The data from the phase I study is just absolutely beautiful, where they can reduce neurofilament to normal levels and keep it down for a year after a single dose. That's a really remarkable chemistry and program that's moving forward that I think is going to be generating a lot of exciting data. We also are doing a similar program with Dravet using that same NMA technology. That's going to start in the clinic soon. We, of course, have ulefnersen for FUS-ALS.
This is a potential second ALS indication for Ionis. This is another genetic cause of ALS. That phase III data reads out this year. There's also the Huntington, Tominersen phase II data that reads out this year. That's in the lower dose, younger, earlier stage patients. We have multiple mid-stage neurology programs in our wholly owned pipeline that are going to read out next year, including aldesleukin as well as prion. A lot of really exciting stuff going on here-
Yeah
guys.
On Alexander disease, the data there, it's really awesome. How have you been setting an expectation, if at all, on the market opportunity for that drug?
There's just a few hundred patients in the U.S., it is an ultra-rare disease, it is a small opportunity.
On salanersen, the chemistry you use there, is that the kind of thing you think could be extrapolatable to all types of targets, including targets that you knock down, or is it going to be more constrained?
It's only splicing.
Okay.
It's fascinating.
Only splicing.
It only works for improving the potency of splicing, across the board for our splicing, it so far has always performed.
Interesting. SMA, Dravet, diseases like that.
Yep
The right thing to think about.
Yep.
Okay.
Yep. Because they're already uniform, they're fully modified because we're modulating splicing.
Right
They don't have that DNA gap in there, they're already really long-lasting molecules.
Right.
You add this on top of it, and that's where you get to that yearly dosing.
Okay. Maybe just lastly, for any new IND you file now in CNS, why wouldn't that just be a Brain Shuttle program at this point?
We still have to show that technology works.
Right
The IT is working. We understand it. We know how to develop these and develop these quickly. It's a known, proven technology, and I hope that we'll be in the same place with Brain Shuttle not too terribly long from now, and then that'll be the answer, but we're just not there yet.
Okay. Interesting. Do you feel like from when you look at your shuttle technology and others, in your mind, is there a material difference in the expertise in targeting transferrin receptor? You know what I mean? I guess GalNAc and LICA became more commoditized pretty quickly, but I don't know where we're at with TFR1.
I think it's going to end up in the same place. Looking at the data, we've played with a lot of these different things. The biggest difference that we can find is the size of what the ligand is and using a different size for more convenience in delivery.
Right.
Other than that, they all tend to perform pretty similarly. We of course make everybody else's molecules and test them.
Right.
I think it'll be more like GalNAc. The biggest thing, though, is still understanding your cargo and understanding the safety profiles of your cargo and what you need to find good ASOs and siRNAs.
Makes sense. Okay, great. Anything else you want to touch upon, Holly?
No, I think we hit it all.
Okay. What about muscle? What's next in-
Yeah
muscle for you guys?
Yeah. We're moving muscle programs forward, both skeletal as well as cardiac. We had our first cardiac muscle program using our Bicycle technology, which I mentioned. These are the very small peptides so that you can do subcu delivery for them. That's advanced into clinical testing. We have another that'll be advancing later this year. It's moving.
Great. Awesome. All right. Well, thank you so much for taking the time. Really appreciate it.
Yeah. Happy to.
All right. Thanks, everyone.
Thanks for having me.
Thanks for listening.
Yep.