Korsana Biosciences, Inc. - 1781018 - KORSANA BIOSCIENCES INC
I'm Michael Yee, Senior Biotechnology Analyst at UBS. Thank you for joining us on our next session. I'm excited to have the President and CEO, Jonathan Violin, from Corsana Biosciences, up here with us. Corsiana has a very exciting blood-brain shuttle technology platform, and the lead program is an A-beta TFR blood-brain shuttle candidate that is very exciting for the treatment of Alzheimer's disease. Jonathan if I may give you an opportunity to just tell us a little bit about Corsana the platform technology and why you think it could be better than other platform brain shuttle technologies and then maybe we can delve into some of the questions. Sure and thank you very much for having us here, a real pleasure to join you so we founded Corsiana because we see a huge opportunity in Alzheimer's disease, we have the first two disease-modifying therapies approved, and yet they leave a lot to be desired in terms of their clinical performance, their safety, tolerability, convenience. And now Roche, to their great credit, I think has really broken open the field in termsof what's next. Their Trunctinamab, which is a shuttled version of Gentinarumab, so took an antibody that binds and clears plaque and attached it to a moiety that binds to the transferring receptor in the right way, and we'll probably talk more about that, that increases access to the brain, crosses the blood-brain barrier, that delivers faster, deeper plaque clearance, dramatically reduced rates of ARIA, so a better therapeutic profile. But in our view, it is not the end of the road. It's intravenous. Patients need to be pretreated with corticosteroids because of high rate of infusion-related reactions, and there's a modest rate of anemia, 10% or 20% rate of anesthesia in the Phase I-II study. So we think there's more to do in this space, And so we designed our lead molecule, KRSA-028, in partnership with Paragon Laboratories, in a way that we think can deliver what we hope will be best-in-class profile. Fantastic. So I agree with you. Obviously, we have total proof-of-concept validation, FDA-approved drugs from first-generation—I say first- generation just because they're approved—A-beta antibody products from Biogen and Lilly. And they're on pace to do a billion dollars and growing. but they do have liabilities. And you guys are enhancing upon that by using the TFR, the transferrin receptor, to shuttle an A-beta program through. So, appreciated from Paragon Labs. Can you describe the profile of the candidate before we get into the data? What has been enhanced on your antibody? Is it a better A-betta antibody? Is the TFR fragment better? Is the overall construct better than Roche-Trentonimab? So we designed KRSA-028 from scratch, thinking about all the parameters that could make it very difficult to beat, that could deliver a best-in-class profile. So I'll walk through the individual components. So in terms of the IgG backbone, the amyloid binding component, our molecule is selective for pyroglutamate forms of amyliid. So that's a post-translational modification that's enriched in plaque. So you can think of this as more plaque-selective. That would put it in the same category as dinanamab, casunla, from Lilly, or their next-gen remternitug, which at least numerically clear plaque a little faster and numerically a little better slowing of cognitive decline. So we thought, given the existing clinical data, that was the place to start, was targeting pyruvate amyloid. So that's the IgG backbone. The shuttle, the portion that binds the transferrin receptor, well, it's a different sequence. pharmacologically, it operates very similarly to the trontinumab shuttle, and that's by design. So what trontiumab has taught us, and this is the culmination of two or three decades of work on how to design these shuttles, it's not just the target that matters, the transferrin receptor. Epitope matters. We don't want to disrupt transferrin binding. Otherwise, you disrupt iron homeostasis. Affinity matters. If it binds too loosely, nothing gets shuttled. If binds too tightly, it gets stuck. So there's a bell-shaped dose response curve. And so when we looked at the trontinumab data, the shuttling component itself we think works great. There are other features, which I'll talk about in a moment, where we thought we could deliver a differentiation. So our shuttle, while it's a different sequence, same target, same epitope, nearly the same affinity as trontumab for the transferrin receptor. So where some of the differentiation comes in then is in the FC region of the IgG backbone. Andso we wanted to achieve three things to help deliver, again, a difficult-to-beat profile. First, we wanted to extend half-life. The goal there is just to allow a lower dose, a less frequent dose, to get to a low-volume subcutaneous presentation. We'll be starting clinical development in sub-Q. We plan to launch commercially with an auto-injector for at-home use. So we wanted half-Life extension. We wanted to preserve phagocytosis. That's the mechanism whereby the approved products clear plaque. Once amyloid gets in the brain, binds to the plaque, phagocytosis by microglia is how those plaques go away. There are some other entrants in the field that have foregone phagocythosis in an effort to, they've eliminated all effector function. And while that may be a safe approach, we think it takes a lot of risk on efficacy. So preserve phagacytosis. So you have active FC? Active for phagcytosis, but then we've dialed out CDC and ADCC, two other effector functions. Because we hypothesized and proven, at least in preclinical species, that that would be responsible for acute reticulocyte depletion, one of the adverse consequences of trontenamab. And so it's that combination of half-life extension, preserving phagocytosis, and reducing CDC and ADCC that we think derive a lot of our differentiation. Okay. And so, in terms of the YTE modification to make it longer half-life, the benefit of that part is not necessarily to make a multi-month dosing, meaning less frequently than once a month. you do plan to do once a month is the ability to have the drug hanging around longer such that you think that it can be a much lower dose. That's correct. Yeah. Okay. Yep. And I should say we've not disclosed what exactly the half-life extension FC modifications are. So we've engineered a number of point mutations in the FC region. One is to extend half-Life. The way YTE works is by modifying the pH dependence of FCRN binding that helps recycle antibodies. So we use the same mechanism, we just haven't disclosed which point mutations they are. Likewise with the effector function. But you're right, by virtue of the half-life extension, we have a lot more drugs circulating peripherally. The shuttle then gets a lot mehr drug into the brain, but we're not aiming to deliver faster or deeper plaque clearance than trontinamide, partly because most patients are amyloid negative by three to six months. It's the fastest plaque clearance that's been seen. we don't know if it's possible to deliver faster plaque clearance. So instead, what we know we can control is the dose, right? And if we get to a low enough dose, that will enable what we're targeting is a monthly low volume. Yeah, I don't think beating something within the first month or two months is going to matter in a long-term chronic disease. Now, that said, we do want to have positive initial data sets in terms of efficacy and safety. So let's walk through that. So you design the compound to have great shuttling into the brain that has TFR receptor affinity to drag it through the blood-brain barrier at an affinity that is similar to trontinamab. So you do believe that mechanistically it should be getting in the same amount of antibody. However, you believe that, and you believe, I guess I would say, that the antibody should be delivering the same amount of plaque clearance as trontinamab. Do you agree with that? I would see at least as good plaque clearance as trantinamib, yeah. And that's partly the potential benefits of targeting this pyroglue epitope. If it's more plaque-enriched, it may clear plaque faster. And remind me, and I think you said so, but denonimab, i.e. Lilly's product, which is FDA-approved, is also targeting a similar epitope along the pyroglutamate form such that while it's not the same epitome, it is the same concentrated effort to go in towards pyrogutamates. So you are definitely going towards reducing the plaque. And so it should be similar to Lilly. Well, except that that is an unshuttled animal, right? But in terms of the core pharmacology once it gets to plaque in the brain yeah it should work right just like the nanomab or remternitog which is lily's next gen yeah that's in phase three and i mean that over say six months that they're using uh they have reduction i'm going to throw out a number like 50 but we'll get to that point but they have to give a lot of drug because they don't use the shuttle so i'm just saying they're doing it in a way that's basically jamming a lot of antibody into you through an iv once a month to try and get 50 plaque reduction in the brain after six months. Your compound hopefully will do it in a much lower dose and a much safer therapeutic window by using the blood-brain barrier transfer receptor. So I said in a bar of 50%, but we'll get to that. So you've done all that. Now, if that efficacy goal of getting in a small amount of drug to be able to put in a sub-Q, get 50% plaque reduction in six months, We've got to do that safely because all these other antibodies do have ARIA. And then the Roche Trontinumab case has a little bit of anemia. I'll say a little but unclear what's going on with the anemia, so how do you feel about safety with your program, and why do you think you will not get ARIA, and why do think you would not have anemia? Yeah, so well let's start with ARIA and I think a lot of people when they first see the trontinamab data and see the very rapid and profound plaque clearance, think, okay, well, that makes sense. You've added, you've enabled higher brain concentration, so the drug should work better. More surprising is that the ARIA rates are dramatically lowered. So from around 20, 25% in the approved products to single-digit percentage. And keep in mind, placebo patients show a couple percent in this population as well. So a dramatic improvement in ARIA. Why is that? Well, it turns out, and I think this makes a little bit of intuitive sense, when you lower the dose, because it requires a much lower dose of Trontenamab to clear plaque, you spare peripheral exposure, and with these unshuttled antibodies, you have to push the dose up so high that you get a lot of crowding of antibodies, and they're normal antibody-driven immune functions, so you're causing local immune responses at the site of plaques on the vasculature. So, if you can dramatically lower the dosage with a shuttle, which Trontanamab has shown, you're just sparing that local immune response as driving this. So you think it actually has to do, to some degree with just the sheer amount of antibody that they're jamming in, which is crossing the vasculature, and that that amount is just causing an immune reaction along where they're binding the plaque because it's so much. On the cerebrovascular, yes, that's right. That's right, and there's some pretty interesting mouse data that suggests that's the case. You can see with these transferrin receptor shuttles, instead of getting huge amounts of antibody in the large vessels of the brain and crowding into the brain through those entry sites, you see very diffuse entry penetration throughout the brain. So it's just a safer, even more distributed way of accessing the brain Maybe analogous to that, do you believe that specifically that the transferrin receptor blood-brain shuttle, which Roche is doing and has put up good data, that the release of the antibody more into the white matter and not necessarily along the vasculature? So a naked antibody, you get an aria because it's going right there, literally crossing through, you know, the gaps within the blood rim bearer, and then that is binding plaque right there along the blood vessels, the microvasculature, and that's what's causing the aria, whereas a shuttle may release it more in the white matter. And I'm not actually sure of the science, but it's diseluting, it's just disassociating from the receptor. the the um crowding problem from very high peripheral exposure trying to push enough drug in the brain exactly as you describe um with shuttles and again you can see this pre-clinically there's transferrin receptors throughout the blood blood brain barrier so all the capillaries the brain are allowing entry just much more diffusely which is allowing a more even penetration throughout the brain okay so that that should reduce the re and again roche has shown that Okay, so if these things are playing through, you could have a potential low-volume dose in a sub-Q pen type of dose and have a product that could be higher efficacy, better safety for Alzheimer's. Now, to do that, we need to run some studies. So tell me about the phase one and what is the design of the study and when will we get some data and what data do you want to show to prove to investors that you have something that is working? So our plan is to run a phase one to study that's intended to be phase three enabling. And so to do dat, it contains three parts. First, single ascending dose and healthy volunteers. We're in that in Australia and New Zealand, for which we'll file a CTA by the end of the year. Second part is a multiple ascending dose in patients, where we can get our first read on clearing of plaque. For that, we expect enrollments almost entirely to come from the United States, so we'll fire an IND early next year to support that. And then thirdly, just as Roche did with trontinumab, we can take doses from the multiple ascended dose, because that's essentially dose-ranging information for us, and we can enter expansion cohorts for one or more doses to increase the sample size enough to adequately characterize safety that we can go to global regulators to get buy-in for phase three. Again, we've got great precedent for integrated SADMADs. In the case of Roche, we have seen a MAD with extension cohorts that has been to phase three enabling. So that means that this phase one, two study can deliver a lot for us in the next three. And what are the timelines for the data disclosure? So first data disclosure, we're targeting mid-year, next year. So that's for the single ascending dose healthy volunteer. Data would include, so this is by the sub-Q route. So first seeing that we achieved desired exposure by the sub-q route, that we've got good local tolerability. Of course, general safety and tolerability, including our first read on anemia, standard CBC. Single dose, so it won't be the end of the story, but a first read at least. And importantly, we are also taking CSF samples. So we can assess if the shuttle is doing what it's supposed to and delivering extra drug to the brain. And we can look at Roche's data for gentanerumab versus trontinumab, so the unshuttled versus shuttle drug, where they showed very nice increases in CSF levels with the shuttle. So in healthy volunteers, after one dose, this data will be in mid-year. Correct, yeah. Multiple different doses of single dose, so dose escalation. and what is good data for a single dose showing that it's getting through the blood-brain barrier? It's a ratio of the compound in the CSF versus the periphery, and what did Roche show so that we could match it up? Yeah, so it's important to recognize that CSF is not the parenchyma in the brain, so the concentrations we measure in CSF are not the same, so it'S not like it's going to directly drive modeling or brain exposure, but it's an indicator of shuttle functionality. So in the case of Roche, they showed a seven or eight times higher CSF to serum ratio. And so we're going to be looking for a CSF-to-serum ratio in the ballpark of Trontenamab. That will tell us the shuttle is working as designed. Right. So seven to eight times means that that obviously went through the blood-brain barrier, got to the CSF. That would be the level. Correct. Could you have more, less? It should be the same, I guess. And you'd want to go until you get to seven to eighth times? Like how would you think about what is the right dose? Yeah, since it's a ratio, that should, in theory, be dose-independent. So really we're thinking about, okay, if the CSF to serum ratio is in the ballpark of trontinamab, then we can go look at the serum concentrations. We already have modeling based on our non-human primate data suggesting that something around 150 mg sub-Q once a month, given a crude assumption for sub-q bioavailability, with that SAD data, we'll have human bioavialability data as opposed to our non-human primate data now, and we'll refine the serum PK. So all that will then refine our predictions of, well, is it really 150 mg once a month, or is it a little more, a little less? How do you think about that? Because trontinamab is a shorter half-life. So if they gave a dose IV, it would look like that and then probably come down pretty fast. So that's an amount of drug in a given month. It's not a long and slow release. sits high and then comes down, but I guess it's there during that month, and then your drug would be sub-Q, so I got to think about the BK, it goes like that, and it's long half-life, so much different shape of the curve, and so would you want the same C average over the course of that month, because the way they pick their doses is that they're giving basically enough drug that's safe. Then when it was safe, they put it in Alzheimer's patients in phase one to see if those doses that were the highest doses safe gave plaque reduction. Then, when they had the plaque reduction, they hoped that that would lead to a clinical outcome. Of course, now we know it does. So, you're kind of working backwards and saying, well, they basically gave as much drug as possible that was safe. They got a lot of plaque reduction and showed cognitive benefit. Awesome. You're kind of saying, in order to pick the right dose that's going to prove that, we just want the same amount a drug in the system. So I think you're highly... Appreciating you will get to Alzheimer's patients too and then model. Exactly. Go ahead. Yep. So, well, this is why we run the experiments, right? We're going to evaluate multiple doses. So we'll be able to learn. But I think what you're underscoring is that the field has learned a lot. So there's a lot of precedent we can rely on. So one great example, lacanumab was initially approved IV, now it's sub-Q. And the IV to sub-q program told us that it is the C average of the AUC that is really the right parameter to model both for efficacy and for ARIA. That's true, right. So because Lekembe basically had an IV dose, it works, it's approved, had positive phase three cognitive benefits, and then the FDA did allow approval of just matching the C-average with a sub-Q. By the way, they give a lot of sub-qs. It's like two per week, I think. It's a high dose, yeah, again, because it's not shuttles. It's not a very efficient antibody, but those eight injections over the course of a month was approved, and the FDA agrees that that's valid. Now, I do recall that also those doses did have plaque reduction. I have to remember, did they do have plaque production? I remember that ARIA that was a little bit higher, but yeah. I don't know, did that have plaque redemption? Yeah, pretty much all the sub-Q LeCambi data looks just like the IV LeCmbi data, and small amounts of noise, but it was compelling enough that it's pretty clear that it is not C-average or C-troph that we are targeting, it really is C-average. So in your phase one, getting back to the data, in Healthies you will look for doses that are safe, that you believe you've modeled out that are going to be predictable about growth C, but in layman's terms, you should have confidence because the C average, the amount of drug that's in the body over the course of that month, is similar to what Roche is showing in their IV. Do you agree with that? That is a large part of what's driving the modeling, yeah. And so as we clear doses in the single ascending dose, we can then start dosing patients in multiple ascending doses. So as we clear sad doses, we'll start dosing people. Ah, so they would be kind of overlapping. So like when you give disclosure then in the mid-27, would you have already started some mad dosing already? Conceivably, yeah. So in terms of what we'll have to disclose, so that first disclosure will just be the healthy volunteer single ascending dose data. That's right. Multiple ascending dos. We're targeting what we're calling interim proof of concepts. So this would be data from one or maybe two cohorts by the end of 27 or first quarter of 2028. Here what we're looking for is using these PET tracers to look at how well we clear plaque. They're remarkably sensitive and precise tools that allow us to see plaque clear very, very quickly. And so what we'll be looking for ist just to know if we are on a dose-response curve that can match the trontinamab rate and depth of plaque clearance. Right. All via the sub-Q route. So if you want to look at it, you can pull up the Roche data. Their phase 1-2 data, which was a couple of years ago, shows approximately 50% plaque reduction. You might have helped me out with these numbers. But it was in a faster onset, and they do have strong, again, about 50% plague reduction at six months. It's more than that. Almost all patients are getting amyloid negative, which is 24 centiloids or lower, within three to six months. Okay, so A, you're saying amyliod negative. Well, that's one way of reporting it. Yeah, that'S a percent of people that get below amyload. Yeah, we'll be looking. For Roche? So the great majority are amylloid negative by three to seis months. So it depends on which time point we're looking at. We haven't disclosed the details yet. But a very clear dose response for the depth and trajectory of plaque clearance. As early as 50 days, you can see remarkable clearance. That's right. So like the first Lillian Biogen are showing about 50% reduction of the amount of plaque on average over six months. Roche is significantly more than 50% and a faster onset. Yeah, it's about, to get to amyloid negativity is about a year faster with trontinumab. Okay. And so your data will be at the end of 27, you will say, okay, we've dosed one or two cohorts at a dose that presumably you think is going to be around the C average. And that after a couple of months, you Will look at the plaque reduction of those patients. I said like a few months, not all six, and that you will show that it's the same amount at that time point as somebody else. And again, if you guys think about like, I don't know, obesity studies, if someone shows three month or six month data that's along the terzapatide curve, you know, yeah, we believe you're on that track. That's what you're trying to show. Exactly. It's really the trajectory that we're looking at. Yep. Downward trajectory of the track. Okay. So if we get that data at the end of 27, and you basically say, see, it looks exactly like trontinamab, then can you quickly enroll a phase three cognitive CDR, some of the boxes study? What would you do? Well, as I mentioned before, we would plan to run expansion cohorts on one or more of our doses just to adequately characterize safety, also confirming the dose selection, right? The multiple ascending dose really is dose ranging, but you want a bigger sample size to know that we've adequately characterize the benefit risk to go launch a phase three. So once we finish the full study, the third part, including the expansion courts, then yes, we believe we would be in a place to start a global phase three, and that should be in 2028. You think it could start? We haven't given guidance yet. Now, while you're executing on that, there's other data points going on in the Alzheimer's space that are important, because a they could validate other populations. So I call it pre-symptomatic Alzheimer's, like trailblazer 3 may be reading out from Lilly. I cover Lilly, so our view is that it could come in the first half, 27, but Lilly has guided 2027, and that could show that we could double the market opportunity because we're going from symptomatic Alzheimer'S patients where these drugs are approved to earlier, healthier, I say healthier because they have not lost as much memory, and that Could double the TAM. Now, importantly, that could be important. Importantly, that could be valuable for these drugs because these are healthier patients where you wouldn't want ARIA. Exactly. Okay, because they do have actually a lot of amyloid. You actually have to have a bunch of amyliid to get into the studies, but they haven't lost their memory yet. So it's based on the sheer science that you're having an anti-amyloid antibody binding plaque, but the person is fairly healthy, that can still cause ARIA So if you could do that without the ARIA, that would be important So we're incredibly excited about the opportunity to treat preclinical or pre-symptomatic Alzheimer's, right? Again, patients who have plaques are not yet symptomatic. There's a lot of retrospective evidence suggesting that the earlier you intervene to clear plaque, the better the cognitive benefit. We just haven't seen that in a randomized controlled trial. The Trailblazer L3 will be the first, and we're quite optimistic about that. I should say that everything we've talked about so far, our planning to date, is all in the early Alzheimer's disease population. That's where the biology is proven. Hopefully, Lilly, maybe you sat behind them, will prove the biology in the preclinical population, and you're exactly right. There's still ARIA. We've seen the randomized data from Lilly, from the Trailblazer ALS-3. It looks like the rates are about half. The rate of ARIA is in early Alzheimer's disease, which makes sense because there's about half the plaque burden as well. And if you think about it, these are patients who have plaque, but they're not yet symptomatic, so they're going to be even more sensitive to safety concerns like ARIA, inconvenience like having to go for an IV infusion. Headache. I mean, but risk of ARIA is not good. We don't want that, particularly, again, in a population that's healthier. So the FDA would be more sensitive to that. So let's see that data because that could be positive. UBS analysts believe it will be positive, so that could double the opportunity. But also, Roche, which we go back to Roche in the traditional symptomatic setting, is in the phase three, and they're going to read out their phase three for Trantinamab using the shuttle. And when is that expected to read up? So they announced completion of enrollment a couple months ago, and they guided to data first half of 2028. Clearly a lot of enthusiasm for this new mechanism. The trial enrolled very quickly, or trials, it was two parallel phase threes. Because Trantanamab clears plaque a year faster than the approved products, the hypothesis is that could deliver better slowing of cognitive decline than the improved products. We'll see. from our view, we hope so, that that's best for patients. From our view of what we want to do with KRSA-O2A is just make sure that we deliver the maximum efficacy possible with this mechanism. So whatever bar trontinamab sets, that's what we need to at least match. Well, it's interesting because you say faster reduction. I guess we're using the similar concept that it's faster and, on average, greater? Well, It looks like... It's definitely faster because you're getting a lot more antibody in quicker. I feel like they have more people that are going undetectable. I guess your point is, well, literally, as they keep dosing, they do get a lot of people to that level. It just takes more time. Most patients who received an enanamab will become amyloid negative. It just take a lot longer. Right, like what, 12 months? Yeah, 12 to 18 months, as opposed to 3 to 6 months for trontinumab. So if you think about— A year earlier, is that what you're saying? Yeah, so if you can spare a year of pathophysiology, hopefully that delivers some additional benefit on the clinical endpoint. Okay. And so that data will be first half 28. I would say that could be the most important industry catalyst that could prove that shuttles have similar or better efficacy and in a safer profile. Better efficacy, better safety. Agreed. It's an exciting readout. Okay. Now, if all that's true and Roche works and that validates TFR brain shuttles, there are more than a few. Is there any comment you could make in our last minute about how to think about, okay, but AbbVie, they've acquired one, although I'm not sure it's TFR. Denali, they're presenting tomorrow here at the conference. They're in a phase one. How do we sort all of those out? Yeah, well, you can argue this is one of the biggest unmet needs in medicine. There should be a lot of entrants. And you're right, we don't have time to go into all the differences today, but there are a lot entrants testing different hypotheses for how to improve on Trontinumab. We think that with KRSA-028 that we've taken the least risky approach to delivering a best-in-class molecule. And we're happy to see others like Denali has some very interesting preclinical data. And what I like about our setup is that this is our lead program. We're well financed. We had over $450 million funding us in the 2029 as of the end of September. And this is Our Focus. We are building this company specifically to advance KRSA 028 and hopefully make a difference in how we treat Alzheimer's patients. I would say TFR shuttles probably are the most de-risked approach because Roche has shown some of this data and could soon be the most derisked because they have the positive data. The handful of ones that are TFR, including yourself, obviously validate that, but with potential better dosing and other profiles, that could be great. And again, it's a pretty big market and about to get bigger if Trilobasin-3 is positive. So thank you very much, Jonathan. I appreciate it. and I'm looking forward to the data next year. Thank you. Okay, thank you.
- Company
- KORSANA BIOSCIENCES INC
- Keywords
- KORSANA BIOSCIENCES INC