Great. Thanks, everybody, for joining us here on day one of Piper Sandler's annual healthcare conference. I'm Joe Catanzaro, one of the biotech analysts here at Piper. It's my pleasure to welcome Keros Therapeutics joining us as their CEO, Jas Seehra. Jas, thanks for joining us. Jas and I spent like two-plus hours just talking about PAH last night. So this 25 minutes is going to be a little tough, I think, but I think we're going to cover some good ground. Maybe first, Jas, I could give you a minute or two to just sort of introduce the company, what you guys have been up to, and what we have to look forward to.
Keros is a clinical-stage company focused entirely on TGF-beta pathway. We have three product candidates: one, elritercept, in hematology indications, and we just announced a partnership with Takeda on that this morning. It's a great partnership that truly captures the value of that program. Our second program, cibotercept, is in PAH, and we believe we have a differentiated mechanism profile from now approved sotatercept, WINREVAIR. Our third program is KER-065, which increases skeletal muscle, decreases fat mass. Therefore, it's a treatment for obesity, but also for a broad set of indications where muscle wasting is the underlying disease.
Perfect. Maybe before we jump into cibotercept and PAH, I could sort of give you the opportunity to speak to the deal with Takeda and elritercept, and why that deal made sense, what made it attractive to you guys, and what it allows you to do now moving forward.
Yeah. So, Joe, we've spoken about this on a number of occasions. In a partnership, it's really important that both parties see the same vision for the drug. And I think in our discussions with Takeda, it's very clear right from day one that based upon the data we had generated to date, they see a differentiated product from the sotatercept. In addition, they see the opportunity not just in MDS and MF, but in all indications where there is anemia. So they're willing to commit to getting the product registered in MDS and MF, but also doing exploratory studies to look for other opportunities beyond that. And that, I think, makes it exciting. And then finally, okay, on the economics, it's a good deal for us economics-wise, where there's $200 million upfront, more than $1.1 billion in milestones, which are split over development milestones, approval, and commercialization.
Royalties starting out in the low double digits to high teens as the commercialization of the product gets too high on IFN.
Great. So maybe with that, we could jump to cibotercept and PAH. And I want to start by saying I felt like last night in our conversation, your level of conviction and the ability to differentiate relative to sotatercept is greater than maybe where it was post-Stellar a year, year and a half ago. So maybe I'll just leave it to you to sort of lay out where you see the opportunity for cibotercept. And if I'm right in that your level of conviction of differentiating, particularly on safety, I think, has increased as maybe physicians have gotten more experienced with sotatercept. So I'll leave it to you.
Yeah. I think sotatercept did a nice job of demonstrating this biology and validating it. And the efficacy with sotatercept is very good. However, we always appreciated that there were risks associated with the biology of sotatercept that is not there in cibotercept, namely that there are bleeding events with sotatercept that were observed in the open-label extension, phase two open-label extension, and that then panned out to be the case in Stellar. We predicted that. I think my conviction has not changed, but how you see it is also reflected by how the community is seeing sotatercept. I think in the 20 months that have elapsed since Stellar data came out, the community has really started appreciating the risk of those AEs, right? And I think there is a better understanding of those AEs that they are due to sotatercept.
I believe that the mechanism that we have with cibotercept will allow us to differentiate. You can see that in the toxicology studies, which are now in the public domain from the summary basis of approval for sotatercept, but we also put out our data about two weeks ago at a scientific meeting. We have a very different profile from sotatercept when it comes to the safety.
So one thing I mentioned to you last night was some feedback I had gotten from some physicians who believe that if you don't see bleeding with sotatercept within the first year, then you should be okay. And you emphatically sort of believe that will not be the case. So maybe speak to that and maybe how we should think about the kinetics of those events and the potential when we see cibotercept data.
I think in PAH patients, that hypoxia puts them at higher risk for bleeding events. That hypoxia results in reduced signaling of the positive side of the TGF-beta pathway, which is the BMPs. BMP signaling is reduced. That results in remodeling in the vasculature. It results in thickening of the vessel walls. But in other tissue, like the gastrointestinal tract, where you've got constant damage, you will have a repair mechanism that leads to more fragile blood vessels. And then sotatercept on itself further inhibits BMP signaling. Therefore, it will, with time, continue to make the situation worse. On top of that, sotatercept increases red blood cells. And it's the increase in red blood cells, I believe, that results in some of the bleeding events showing up earlier. But that remodeling is going on all the time.
After all, there is the disease hereditary hemorrhagic telangiectasia, where patients have loss of 50% signaling of one part of the pathway. And it results in those patients having telangiectasias towards the first decade of life. In the second decade of life, that leads to more serious bleeding events like gastrointestinal bleeds and cranial bleeds that can lead to death. So the precedent is there that this biology is cumulative, right? So I think you see the increase in bleeding events over the first 12 months, but then it continues beyond that. It's just not large enough for the physicians to take note today. But once patients have been on it for two, three years, I think that biology is going to continue to compound.
I wanted you to elaborate a bit on a topic that I maybe didn't fully appreciate, and that's the implications of increased hemoglobin to bleeding that's being seen with sotatercept. Because I think that's interesting because if that is, in fact, the case, I think cibotercept has sort of definitively shown, right, in healthy volunteers, that there is no elevated hemoglobin effect. So what do you think is going on with the increase in hemoglobin that is contributing to some of this bleeding risk?
So I go to the tox data first and foremost. In the rat and in the monkeys, the major tox finding was hemorrhage in the kidneys and tubular necrosis following that hemorrhage. In the monkeys, the NOAEL was one milligram per kg, 10 milligrams per kg, there was blood in the urine. You can look at the patent literature. You'll come to the conclusion that 10 mg per kg in monkeys is about one and a half grams per deciliter increase in hemoglobin. So you're seeing this. You see this with ESAs, where you get hemorrhage in the kidney, not the tubular necrosis. You get inflammation. But this is the biology of the TGF-beta pathway playing out in terms of tubular necrosis. But the hemorrhage is there. So increase in red blood cells results in increase in viscosity. If you've got a blood vessel that's fragile, it can break.
Any increase in viscosity is going to result in that blood vessel being more susceptible to breakage.
I want to maybe stay on hemoglobin but change ever so slightly. I remember before Stellar, there was all this discussion of how many patients are going to be able to titrate up to 0.7 mg per kg. We saw the data. Everybody was able to get there, and people seem to have forgotten about that. But when I've spoken with physicians recently, it seems to be a real pressure point for them to adopt and use sotatercept because of the hemoglobin monitoring. That on its own, do you think that's sufficient enough for cibotercept? If it says, if you show, "We don't need to dose titrate. We don't have this elevated hemoglobin," it's much easier to sort of adopt and get a patient on it. That on its own, is that sufficient to get physicians to buy into the program?
I think the last four years since we've been public and having these conversations, we started off with exactly that thesis. Once-a-month dosing, no monitoring. It's convenient, okay? I think when you don't have an increase in hemoglobin, therefore you don't have to titrate, you have the potential for a fixed dose. That alone is sufficient. And we've heard this again and again from investigators. It's a burden on the patient. It's a burden on the healthcare provider to be doing constant monitoring. And what we've also seen and heard from physicians is the label says that you have to monitor for hemoglobin changes for the first five doses, and if it's stable, no longer. They're chronically monitoring hemoglobin changes. That's a burden and a cost to the providers and an inconvenience for the patient and the healthcare provider. So we think that is enough in itself.
But add to that is that you could get potentially better efficacy and safety profile.
Yeah. That's a perfect segue because I think we covered good ground in safety. So I want to move to efficacy. And another sort of topic of discussion that we covered a lot last night was efficacy, dose responsiveness, whether there should be expectations there, what we saw with sotatercept, what we know about cibotercept and its ligand-binding profile, the doses you're testing in the TROPOS-2 trial. So maybe you could just sort of speak to that. What sort of data points lead you to believe that your higher doses will translate to greater target engagement and that potentially can maybe translate to better efficacy?
Yeah. So let me start off with the sotatercept data. Pulsar phase 2 trial showed a dose response in efficacy where the PVR changed from 0.3 to the 0.7 of 150 dyne reduction and a 250 dyne reduction. So there's a dose response, right? At those low doses, you do not have maximal target engagement. Go back to sotatercept. At 3 milligrams per kilogram on follicle-stimulating hormone, you get 100% target engagement. You see changes in bone biomarkers as well, bone-specific alkaline phosphatase. With KER-012, you also see 100% target engagement at 4.5 mg per kg subcutaneous, whereas the sotatercept was 3 mg per kg IV, so equivalent doses. So you're seeing the same changes in FSH when you look at the changes in bone-specific alkaline phosphatase. At those doses, they're similar. So you're seeing that same biology there.
And yet you're only dosing at 0.7, not fully able to capture the dose response. We're going all the way up to 4.5. So we've got 1.5 mgs per kg, 3, and 4.5. At the 1.5 mgs per kg, you already get to a higher C-Max. Your trough levels are down to negligible drug levels at day 28 before you come with the next dose. But you've got greater target engagement for a short period of time with the 1.5 mgs per kg. With the 3 and 4.5 mgs per kg, you've got maximum target engagement for a period of time. And that period of time that you have maximum target engagement increases with the dose. So you're allowing the biology to be initiated and play out.
So we think we've set up the experiment to be able to capture more biology if there is more biology to be captured from this pathway in PAH patients.
So on that point, I think many point to the FDA review documents and the dose response analysis they do for sotatercept and this idea that sotatercept is hitting maybe a ceiling on the dose response curve. Why do you think that dose response curve is only reflective of sotatercept and wouldn't be reflective of what you might see with cibotercept?
So remember, in these patients, you've got increased signaling of activins, and you've got decreased signaling of BMPs. You want to increase the BMP signaling and decrease the activin signaling. And you do that with a ligand trap that binds the activins. But if a ligand trap is binding the BMPs as well, okay, right, you're still inhibiting that. So if your window is this, you want that balance to be restored to this. With sotatercept, you're decreasing the activin signaling, but you're further decreasing the BMP signaling. So the window is not there, whereas with a more selective BMP sparing molecule like cibotercept, that window is bigger. And therefore, you have the potential to actually capture more of the biology.
So I went back and looked at the Pulsar phase 2 top-line release, and I noted to you it was very limited. So I want to give you an opportunity to sort of lay out what you hope to be able to say about TROPOS at the time of its top-line readout, which you guys have guided towards as a 2Q25 event.
Yeah. You try to always present the meaningful data at a medical conference. It just so happens with the timing of the cibotercept readout, which is Q2 2025, there is no medical meeting that we can present the data in close proximity of that. American Thoracic Society meeting, the deadlines for that have gone. The late breaking is in January, so we will not be able to present there, so the earliest medical meeting is towards the end of Q3. I think for that reason, we are going to have to provide all of you a more substantive data set. I think that includes the PVR, the safety, the secondary endpoint of changes in NT-proBNP and six-minute walk test, so it will be a more expansive data that we will share.
We will still hold back some so that we can present it at a medical meeting later on, but we will be as comprehensive as we can in order to provide that.
Maybe you could just remind us, of course, TROPOS has three dosing arms. Your expectations of pooling, I think the primary endpoint is designed so that those doses all get pooled and compared against placebo versus being able to maybe look at each dose independently and then determine what will be the move-forward dose.
Yeah. So for statistical significance, it is the combined dose groups against the placebo. Nevertheless, we do anticipate being able to see a dose response. It will not be statistically significant, but there will be a trend, as you saw with sotatercept and Pulsar. Each dose of 0.3 and 0.7 was not statistically significant from the placebo. But when you pool the two doses, it was statistically significant. So I think that's our expectation. With Pulsar, it was 150 dyne and 250 dyne. So the average for the two dose levels comes out at roughly 200 dyne. I think if there is no more biology to be captured, then that's sort of the number that we think is possible.
Maybe on that point, maybe you kind of answered as I've been trying to go through the exercise. It seems like Stellar is the right comp as we look at safety. In Pulsar, they weren't really looking for bleeding events and telangiectasias. But on efficacy, should we look at Pulsar in that 150, 250, 200 midpoint, or is Stellar a fair comp or not as well?
Well, so in Stellar, okay, right, everybody got 0.3 for one dose and then 0.7 thereafter, okay? And you got about 250 dyne change there. So whether you're looking at Stellar or Pulsar, okay, numerically, you get the same change in PVR with both. So I think it's six of one and half a dozen of the other, right? Pulsar showed you the dose response. It was a smaller study, so I think it's an appropriate look.
I think something that was interesting that came up again last night was a future phase 3. I mean, everybody always thinks of the implications of sotatercept being on the market and what that means. Obviously, TROPOS enrolled very quickly. But maybe looking towards the future, contingent on TROPOS, what a planned phase 3 could look like in PAH?
I think the registration trial would look very similar to Stellar, okay? I think in these studies, more than 80% of the patients come from ex-U.S. So there is not going to be competition from sotatercept in many of those countries. Even when you've got approval in Europe, it takes a long time before the drug is commercially available. I think one of the concerns would have been that with the phase 2, if sotatercept was commercially available in the U.S., then A, patients would have been warehoused for that. And second, now that we see that physicians are treating the more advanced patients, those on triple therapy with sotatercept, there is always a possibility that in a phase 3 trial, there are fewer patients that are on triple therapy that come into a cibotercept phase 3. But that's only about 15% of the patients.
It doesn't really make a significant difference numerically in the study.
In these last couple of minutes, I do want to touch on 065, your Activin-Myostatin trap, but we'll get initial healthy volunteer obesity data in 1Q25. Maybe you could just speak to where you think 065 fits relative to the other efforts looking to sort of hit myostatin and Activin in the context of obesity and what we'll learn within that 1Q data set.
Yeah. I think 25 years of work in this field shows that there are multiple negative regulators of skeletal muscle. It's myostatin and Activin A. You can inhibit myostatin or Activin A on their own. You get small changes, go inhibit both of them, you get the biggest bang for your buck. And you can do that in numerous ways. You can go in with antibodies that bind myostatin and Activin together to get that big bang for your buck. You can do it with antibodies that inhibit the signaling by binding to the receptor, or you can do it through a ligand trap like we are doing it with 065, where you bind multiple ligands: Activin A, Activin B, myostatin, and some of the other players that are involved in not just in muscle, but also in the fate of fat cells, okay, right?
I think you'll see whether it's the Regeneron approach, the Lilly bimagrumab approach, whether it's 065, the magnitude of increase that you can get at highest target engagement is going to be similar. Myostatin antibodies on their own is going to be a small fraction of that. And you can see that in a poster presentation that's on our website where you see a head-to-head comparison of a myostatin antibody with bimagrumab with 065 and actually also of the Biohaven molecule.
Maybe last question. I wanted to get your quick thoughts on your read of Lilly having bimagrumab. Has run studies in combination with semaglutide. Just started a study with tirzepatide. But just recently in license, I think Activin 2AR antibody. What do you sort of make of that? What do you think their sort of read is?
I think they're putting lots of bets in this pathway, right? I think they see the value. I think the data is, in my mind, very clear that you need to inhibit both Activin A and myostatin in order to get the maximum increase in skeletal muscle. And Activin A and myostatin signal through both Activin receptor 2A and 2B. And both of those receptors are present on skeletal muscle. Therefore, you have to inhibit both receptors in order to see that. Inhibiting one on its own is not going to be sufficient. So they're putting lots of bets. And when you've got the resources that they have, they can afford to put bets because there's always surprises in this biology.
Yeah, that's for sure. Great. Well, with that, we're out of time. So, Jas, thanks for your time and thoughts. Thanks, everybody, for joining and enjoy the rest of your day here. Thank you.
Thanks, Joe.
Okay. Why don't we go ahead and get started with our next session? My name is Chris Raymond. I'm one of the biotech analysts here at Piper Sandler. Very pleased to have with us our next company, presenting company IO Biotech Inc. We have Qasim Ahmad, the CMO, and Amy Sullivan, who's the CFO, presenting for us this morning. Just a quick overview of the.