At the Leerink Partners Global Healthcare Conference. My name's Tom Smith. I'm one of the senior biotech analysts here at Leerink, and it's my pleasure to welcome our next company to the stage, Keros Therapeutics, represented by their CEO, Jas Seehra. Jas, thanks so much for joining us. Tom, thank you for giving us the opportunity to have this conversation. Of course. A busy 2025, busy year of execution and kind of setting up 2026. Maybe just kind of walk us through some of the like highlights, execution highlights from 2025 and then what we're looking forward to most in the year ahead. Yeah. As you know, we did the Takeda deal at the end of 2024. Twenty-five was a year of transition of that program, eleriscept, to Takeda. We got that done. Going forward, they're responsible for the program in totality. We still get to advise, but they're responsible for it. I will say that partnership is one of the best, if not the best, I've ever seen, okay, right. The two teams really clicked and moved everything forward, very, very aggressively to the point where now Takeda is in the Phase three second line, but they're also starting the Phase three frontline study. It's on ClinicalTrials.gov. Later this year, they will make a decision on the myelofibrosis trial. I think really good progress on that front. Very shortly after that, we saw some safety signals with cibotercept, whereby we stopped dosing in the mid and high dose. In January, we stopped the trial completely, read it out, and it was very, very clear, okay, right, that we could not go forward with that program. That program was deprioritized. Around the same time, we read out our Phase one healthy volunteers study with rinvatercept 065, and it provided the proof of mechanism. Thereafter we decided planning to take it into neuromuscular indications. First of those is Duchenne's, and we hope to get that trial started any day, but it's moved into Q2. Earlier this year, we actually announced a second indication, which is ALS, and yesterday we announced that our program had been selected in the ALS MyMatch program after experts evaluated the merits of it. That's very exciting. Awesome. Great. Let's start just kind of high level like, the foundation of the Keros story here is really TGF-beta superfamily. Mm-hmm better ways to really harness that tremendous biology. Maybe if you could talk a little bit about just sort of the platform and pipeline that you have, and then we'll dive into rinvatercept. Yeah. I think when I started the company, it was going to be focused on TGF-beta pathway, and we've remained true to that. Our first program was a small molecule, and then thereafter, all of our programs have been ligand traps based upon the activin receptors. I think that's provided us with a rich and fertile ground to mine, where we can come up with variants that have different properties due to different ligand bindings, and therefore, they can be targeted to different indications. I think that's still a very fertile ground. I think what you'll see going forward is also addition of ligands with appropriate properties for systemic delivery, as well as multimodal inhibitors. I think the platform is there. It's rich in biology that can be captured for different indications. Great. Let's dive into rinvatercept, and this is your ActRII ligand trap. We're inhibiting activin A and myostatin. As you said, you've produced some healthy volunteer data that's giving you the confidence to go into DMD. Maybe just talk about kind of the underlying biology here and, like, why you're excited about taking that approach into DMD, and then talk a little bit about what you saw in the healthy volunteer data. Yeah. I think in DMD, the way I tend to think about it, Tom, is these patients get diagnosed between the ages of 18 months and three years because they are slow to start walking, but each and every one of them eventually walks, and then they go on to run before they start seeing that decline phase. Glucocorticoids are able to preserve okay function for a period of time, and that's beneficial for those patients. It's a delay in loss of ambulation by 2-3 years. Keeping that muscle regenerating allows that muscle to provide benefit to these patients. I think that fundamentally is what we're trying to do. By inhibiting the ligands activin and myostatin, we are inhibiting the negative regulators of skeletal muscle, and by inhibiting those, we're allowing muscle regeneration to occur. As a consequence, the muscle fibers are a little bit bigger, and therefore they're a little bit stronger and able to withstand use and not break down so readily. In these patients, as that muscle breaks down, it's replaced by fatty infiltrates and fibrotic tissue. Well, the TGF-beta pathway, in particular activin, is also a potent driver of fibrosis and adiposity. When the muscle gets infiltrated with fat itself makes Activin B, which signals through the same receptor. Now you've got a cascade of biology that's bringing other players that by using rinvatercept, you can inhibit Activin A, myostatin, but also Activin B, okay, right? You're really dealing with all of the consequences of pathology, which is a weak muscle due to lack of dystrophin. That makes sense. How, and there are a number of other myostatin activin agents that are in the clinic, like how is rinvatercept differentiated from those other approaches? Yeah. I think there's been the myostatin only antibodies which went into Duchenne's trials, and all of them showed signals of activity but failed to show efficacy in their primary endpoint. Okay. I think that's because myostatin is not the only negative regulators. If you've got two players acting through the same signal, right? It's like two soccer balls, okay, with two players and one goalie, okay, right? You're gonna get goals, right. You have to actually inhibit both of the ligands, right? I think that's where the failure of those molecules came from. You wanna inhibit both negative regulators, and you can do that through ligand traps that bind multiple ligands. You can do it through antibodies that bind each of the ligands, or you can actually inhibit the receptor, okay, right? That's the signaling receptor. Unfortunately, there are two receptors, activin receptor 2A and 2B, so you've got to inhibit both of those. If you inhibit both receptors or all the ligands that are negative regulators, you're gonna get to the same magnitude effect. Where the differences are in the safety signals that you get. The activin receptors are also used by other ligands, bone morphogenetic proteins, and if you inhibit the activin receptors, you begin to see the signals, safety signals associated with bone morphogenetic proteins. Bimagrumab, which falls into that, which is an antibody that targets the receptors. You see muscle spasm, diarrhea, and acne, which are all, I believe, due to BMP inhibition. The ligand traps have not shown any of those safety signals. They've shown other safety signals like bleeding and so on, but they have not shown those safety signal. I think it's really about trying to select the best molecule for the indication that you're going after. That makes sense. Turning to the Phase one healthy volunteer results, I think you saw, I mean, very good evidence of kind of like multi-tissue activity. Mm-hmm. We saw lean muscle, lean mass increases, good changes in bone biomarkers. Maybe just, like, what was the most, I guess, striking part of the healthy volunteer experience for you that drove you to advance in DMD? Yeah, it was the increases in lean mass. There was really nice improvements in bone mineral density, which are key to the patients with DMD, but also actually the decreases in fat mass, and it was the abdominal fat that showed the biggest decreases, right? And that's the unhealthy fat, okay, right? Therefore, okay, right, in boys with DMD, this could translate into because they have increased adiposity, because they have been on glucocorticoid for multiple years, they are actually beginning to develop metabolic syndrome. They have insulin resistance, high circulating A1C. So by reducing fat in those patients, you're not only providing metabolic benefit on the muscle, but you're probably providing benefit long-term in terms of cardiac health, okay, as well. That makes sense. Let's talk about your Phase two plans. Mm-hmm. Maybe just walk us through the design. I think you have a cohort targeting late ambulatory, you have a cohort targeting non-ambulatory. Yeah, just help us understand, like, how are we looking to tease out the signal in DMD? Yeah. First things, more than 60% of the patients are not eligible for exon skipping, right? They're on glucocorticoids, and they can either be on glucocorticoids alone or some may graduate to gene therapy, but very few are gonna be on gene therapy. We think that's the larger population that we can look at to evaluate the safety and get signals of activity from this molecule. Late ambulatory, why? Just as we were talking earlier, boys with DMD get diagnosed, but then they do all walk, and they get stronger. You don't want them in that early phase where they're getting stronger because it'll be easy to miss signals of functional benefit. On the other hand, when they get into that late ambulatory stage, that's when they're either stable or losing muscle function. It'll be easier to see those signals, okay, right. The non-ambulatory early non-ambulatory still have upper body function, and that's important because for these boys, every time they lose a ability to do something, it's a milestone they're never gonna get back. Therefore, being able to maintain upper body function is really critical for their quality of life, okay, right, once they're in a wheelchair. I think that's the two cohorts, right? When we think about what are the endpoint, of course it's safety. We need to establish the safety. Because it's an open label study, we will be getting data on a frequent basis. There'll be imaging every three months. When we have enough patients with enough data, we can share that, okay, right, with the broader community, the investors, the patient, and the KOL. I think that will be helpful to us as we are planning a registrational trial to get signals of activity that say, "Oh, this drug may be doing something," right? "And therefore it's worth me being on this trial." I think it's really looking at lean mass there, fat mass, bone mineral density, right? That you should start seeing a 3-6 months time changes in those. Then, with longer-term treatment, and changes in functional benefit, that can include improvements in forced vital capacity. If you're strengthening the intercostal muscles as well as the diaphragm, that may translate, okay, right, to improvements in your ability to breathe. I think with longer term, there's a whole host of exploratory endpoints that could translate to functional benefit that could be part of a registrational trial. Yep. Within the context of the open label nature of the study, yeah, like how do you think about what would be like a really convincing signal across some of the early biomarker data that you're gonna get? Like longer-term stabilization improvement on some of these, I mean, that is, I guess the dream goal. Yeah. Like what is a realistic sort of bar for success here in this study? Yeah. I think, look, on the quantitative measures, the imaging ones, you should start seeing some changes at three months. Okay. Right. Therefore I think they're gonna be, compared to baseline, they're gonna be easy to see. Okay. Right. It's small studies, so don't look for statistically significant, look for the trends, okay, right? The directionality. I think on the functional benefit, that's where it's gonna be challenging. Okay. It's again gonna be compared to their baseline. If the patients are already in a decline, okay, right, you may be able to stabilize those. You got to look relative to the baseline, is there a decline and how does that compare to natural history studies? Okay. Right. I know, it's gonna be an external study, therefore, no matter what you get, you have to take it with a pinch of salt. I'm looking for signals of activity that give me reasons to plan the Phase three registrational trial. When you talk about biomarkers, we see a lot of serum biomarkers change as a consequence of treatment. We saw it in the healthy volunteer study. Actually, just couple of days ago, we presented a poster at MDA that showed changes in biomarkers of inflammation, fibrosis in a healthy volunteer study. Not only did we see fat mass changes, bone mineral density, skeletal muscle, but also really some of the key drivers of the pathology in patients with DMD, which is inflammation and fibrosis. I'm excited for those reasons. We're seeing all this signal now. Can we put it all together to show a benefit in patients with this disease? Yep. Yeah, that's helpful. I suppose just timing-wise, like we need to get the study kind of off the ground, but just like from a high level without asking for explicit guidance, but it's open label. Yeah. 3-6 months, good sort of timeframe for biomarker changes. Like reasonable to think we could see some data in 2027. Yeah. I, you know, not official guidance, okay, right. We haven't got the first of the patients in the trial, but I think as we get more patient, we'll be able to provide better guidance. Today, I think it is reasonable to assume you're gonna get patient coming into the trial on a rolling basis. As soon as you have, you know, perhaps six patients or eight patients, okay, right, you begin to build confidence, okay, right, in the data that you can share externally. You've known me a long time, okay. I'm not gonna share data that I don't have confidence is, you know, you're not gonna ever hear me talk to you about one patient data. Right. Okay, let's shift gears. You highlighted also the obvious clinical interest in ALS. Mm-hmm. Different disease pathogenesis, but at the end of the day, you're trying to preserve the muscle function. Mm-hmm. Maybe just talk about, I guess, where the biology fits in specifically around ALS. I know you also have some preclinical data, I think a SOD1 mouse model. Just, yeah, what signals did you see there that give you the conviction to move it forward? Yeah in ALS? I think first of all, let's look at it at a high level. In ALS, it's the loss of neuromuscular junctions, okay, from the motor neuron that lead to the muscle atrophy. Okay. That's the driver of the pathology. But that muscle atrophy, okay, also leads to inflammation in the muscle, right? And therefore some of the pathology is consistent with what you see in the healthy volunteers as well as in DMD. We're not gonna attempt to preserve the neuromuscular junction, but each axon innervates many, many different myofibers and therefore, okay, right, when you lose one neuromuscular junction. You still got innervation of all of the other myofibers. If you can make those stronger and as a consequence maintain the functionality of that muscle, you can provide quality of life benefit to patients, right? They'll be able to preserve functionality of whatever muscles they have for a longer period of time. That's the key, okay, right here. We're gonna do nothing for the neuromuscular junction directly. Now, if you improve inflammation, maybe there are downstream effects, okay, right, but I'm not counting on those, okay, right, today. What have we observed in preclinical studies? We've observed as a consequence of treatment, these animals are able to maintain better weight than untreated animals. Animals, as they're getting sicker, they're losing weight, whereas treated animals are gaining weight. That also translates into better grip strength, right? So you've got. You can actually see they have more muscle, but in addition to having more muscle, it's stronger. Then if you stimulate the nerve, you can actually measure what the strength of the muscle as a whole is, and you can see how it declines. You can... we've shown that that is improved, okay, right, as a consequence of treatment. How this translates in everyday life for the mouse is that placebo mice are gradually getting weaker and weaker to the point where they're having more difficulty moving around their cages and so on. Treated animals are not, okay. They're jumping up and down, okay, and doing all the things that mice do when you put them in a new cage. However, at some point in their life, it gets to a point where there is enough loss of innervation in the muscle that now you have catastrophic failure of that muscle, and you'll find animals that have died in the cage that looked perfectly fine the previous day. I believe that's because of loss of innervation of the diaphragm. That muscle is critical for life. Yep. I think, you know, the animal data actually guides us, okay, as to what we could see in patients. That makes sense. You've guided to regulatory engagement, I think, in the second half of the year. Maybe just, like what are the key questions that you need to get alignment with FDA on? The patient population? Like, what aspects of the trial design are outstanding questions? Yeah, I think let's talk about the population first, okay? It's a very heterogeneous population, so you wanna be careful that you target the population where you think you can show some benefit, okay, in an early trial. I think you have to exclude the patients that are rapidly progressing in their disease. The bulbar patients are not going to be included, so it's gonna be the slow progressing, right? They can be the familial ALS, okay, right, with the SOD mutations, or, they can be the sporadic, okay, but it has to be the slow progressing. I think you've got to get alignment around that, okay, right? you know, it's gonna be an exploratory study, so you're gonna have all of the measures that we talked about in DMD, but now you're also gonna have the ALS functional score revised, okay, right, as another measure. Small study, you're not gonna get, again, enough data to convince yourself, but you can again see directionality. Are you seeing that? Okay, right. I think that's what we wanna achieve, okay, right. Again, proof of mechanism in this patient population with signals of activity that give you reason to go to a registrational trial. That makes sense. Maybe just for those in the audience who may be a little less familiar, you mentioned selection in the consortium study. Yes. Like, just walk through kind of like what was that process and, you know, how did they select you? The Healey Center at MGH has been involved in doing clinical trials in ALS for a very long time, three decades or more, okay? They've really perfected ALS trials. They have rapid recruitment, and because it's the consortium of sites, okay, right, they know what they're doing. They have a advisory committee that looks at all opportunities, and it's a, you know, competitive process, and we got selected. I think it really goes to that our mechanism is different. We're treating the muscle, and now we have expert advice, okay, right, in helping us conduct the trial. The way the program runs, okay, right, is that actually they have four sites currently that will do all the recruitment. They'll do. They know, okay, how to assess all of the measures, okay, right, that are there. It really is a way of, efficiently utilizing limited resources within the company and then use an expert group to help you, okay, move the program forward. I'm really excited about that. That's great. A very efficient way of getting the program off the ground and interrogating the biology. Yeah. That's great. Okay. Let's switch gears and talk about elritercept, which you partnered with Takeda. Phase three RENEW study is underway in second line. Just kind of remind us the data that you guys generated in Phase two that led to that licensing deal and how this differentiates and compares versus luspatercept or imetelstat as sort of the two contemporary advanced therapies for anemia and MDS. Yeah. I think in the Phase two trials, we showed a broad response in treating anemia and increasing platelets, thrombocytopenia in MDS patients, both RS and non-RS. You know, in the non-RS, our numbers went up and down as the number of patients in the trial went. Okay, there were days, okay, when everybody said, "It's great," and the other days, "it doesn't look any different." As the data has matured, you've seen about a 50% response rate in the RS and similar in the non-RS, which is consistent with the hypothesis that we had that this molecule is different from luspatercept. Luspatercept was designed not to bind to activin A. This molecule has activin A binding, and therefore we should see a broader response. The other thing, okay, right, that was always on our radar is a durability of response. If you're inhibiting activin A, you're actually making the bone marrow less inflamed, which means you are allowing the normal hematopoiesis to continue. So you're preserving that, and that should result in a greater durability of response that we saw in the Phase two trials. That's what led to a partnership. In the RENEW Phase three trial, it's in a second line setting, luspatercept was a sort of durability of response is about six to nine months. We've seen much longer, okay, right? That says, okay, right, that when patients are on treatment, they're gonna remain on treatment longer. In the frontline setting, with Reblozyl, you know, now you've got patients staying on it 18 months, okay, right, and therefore you see that show up in the revenues, okay. We're going from around three-quarters of a billion and now, okay, right, probably approaching $3 billion this year, okay, right, because you're treating more patients earlier. Imagine, okay, right, that with elritercept, you can continue that treatment beyond what luspatercept is able to do. You're starting them off and you're keeping them on the same treatment for a much longer period of time. That can get reflected in your revenues, okay, right. I think the other thing, okay, that luspatercept has consistently shown is bone pain, and patients just don't feel good. Even when they get transfusion independence, they still have all of the symptoms of being anemic, okay? That is not the case with elritercept. Patients feel better. The bone pain is disappearing, okay, right? Even patients that didn't achieve transfusion independence but were getting their transfusions reduced felt better, okay, right, on treatment. I think all of those observations that were there, both qualitative as well as quantitative, led to that partnership. Now they're starting the Phase three trial in frontline. It's called ARISE. It's on ClinicalTrials.gov. Then in myelofibrosis, we also showed increases in red blood cells as well as platelets leading to reductions in transfusions and also transfusion independence. They've continued that study. Data was presented at ASH consistent with what I just shared, and they are going to evaluate the opportunity in Phase three and make a decision this year. Yep. I think it's maybe building on 2 points that you've raised. One, the breadth of vision for Takeda and their willingness and desire to evaluate most every setting that this could potentially go into, which myelofibrosis is a logical next step. The other side of it, I guess, is the activin binding and the potential to maybe alter the underlying bone marrow. Like, both of those seem to lend themselves, in addition to the Phase two data, very well to advancing in myelofibrosis. Just remind us of. You mentioned this is one of the best deals you would know having been from with Acceleron when they partnered. Yes. luspatercept with Celgene back in the day. Just remind us of the economics and maybe help us think about, like, what an expansion into myelofibrosis could mean to revenues that then flow downstream to you in the form of royalties. Yeah. There's a $200 million upfront, and then $1.1 billion in biobucks. About $90 million of that was in development, and the rest is registration and then commercialization. The royalties start in low double digits and go up into the high teens. Okay, right? That's a good deal, okay, right? They're bearing all the costs. I think with MDS, we just talked about what the revenues are there with Reblozyl, so that's a competitive example. In myelofibrosis, okay, right, it's hard to say, but the patient population, if it's the same pricing, could also be equivalent to what Reblozyl was in second-line setting, okay, right? That's significant. I think you pointed out they actually are looking at broad set of indications because I do believe that this treatment could thereby provide benefit in patients with anemia regardless of what the underlying cause is. Their willingness to think broadly really got me excited, and I said, "This is the partner. Yep. That's excellent. Just in the last 30 seconds we have, wondering if you could give maybe a quick soundbite around the early pipeline and I guess plans to advance things from the early pipeline, just timing, what, like, when, you know, when can we expect to see something new here? In the past, we've always only shared what's in the pipeline when we're ready to actually start dosing healthy volunteers. I think you can always look at our corporate presentations at scientific conferences. Just at the MDA, we shared a neurotrophic factor, okay, right, in this pathway. I think we speak of our runway taking us into the first half of 2028. That includes the rinvatercept trials that we talked about, but also includes additional assets coming into the clinic. As the year progresses, we'll highlight some of those molecules. That's great. All right. Well, unfortunately, we're up against time, but thank you, Jas, for joining us and sharing the insights, and a lot to look forward to here in 2026, 2027. Thank you so much. As always, Tom, thank you very much. Okay, really appreciate the conversation.