Joining us here on this our first day of B of A's 2025 Healthcare Conference. My name is Jason Zemansky. I'm one of the SMID-Cap Analysts here at the bank. For this slot, I'm very pleased to have join us on stage, Jasbir Seehra, Chief Executive Officer of Keros. Thank you so much for joining us. Thank you, Jason, for giving us the opportunity to have this conversation. Absolutely. Maybe let's start with some background just to help frame the discussion as we provide some context for those newer to the story. Can you provide a brief overview of Keros and its platform? I mean, why focus on the TGF-beta pathway? It's actually a very good question. The TGF-beta pathway is one of the key pathways that's involved in embryogenesis. It's involved in differentiation, leading to tissues, enduring growth of tissues, but also in the adult, in the maintenance of that tissue. It really controls the differentiation of stem cells into differentiated cells that make up the tissue. When this pathway is disregulated, it leads to many of the pathophysiologies of the diseases. Regardless of what may be the underlying cause of the disease, this pathway gets disregulated, and then you see the pathophysiology that is common to many, many diseases because of this pathway. Right? Importantly, the pathway is conserved through species, and therefore the biology is conserved. That makes it very easy to think about the safety and the efficacy profile because rodent studies and small animal studies often point you to the risk-benefit equation. Makes sense. And then maybe just real briefly, can you talk about your pipeline? Our pipeline has three assets in it. Most advanced is elritercept-050, which is a ligand trap in MDS and myelofibrosis. We partnered that with Takeda in December. We're great economics finding a partner that really shares the vision for the drug that we have, we had, and continue to have. I think that was a very good deal for us. Our next asset is cibotercept that was in PAH trial that we terminated early because we observed some safety signals, and we can surely talk more about that. We're going to have the top-line data in this quarter, and that will determine what the path forward for that molecule is. The third molecule is another ligand trap that now binds ligands that are negative regulators of skeletal muscle. As a consequence, treatment with that molecule increases skeletal muscle, increases muscle regeneration, decreases fat mass, and also improves bone mineral density. That makes it particularly suitable for multiple neuromuscular indications. Got it. Maybe let's address the elephant in the room. An activist investor has challenged Keros's leadership. That entity has published an open letter to shareholders, and Keros has responded. Maybe to address some of the points here, to the extent that you can say, at this point, are either cibotercept or 065, you know, drags on the PNL? I mean, at this point, TROPOS is winding down, 065's phase II is a ways away. I mean, what is spend like? Yeah, look, they've, you know, filed a copyright, the open letter, and then a slide deck, copyright, as of yesterday. We engage with investors all the time. I think I'm very active in engaging with investors. On average, since we went public, I think I hold 400 or more calls, meetings with investors. We listen to our investors, and we have that discussion at the board level. When we had the news with cibotercept, our stock declined. We are below our cash at this point in time. Based upon where we were, the feedback from some investors, the board decided to look at strategic alternatives to maximize the value of the pipeline for all shareholders. We're undergoing that process. It's unfortunate that ADR One is doing this because we should be focused on maximizing that value. That is through our pipeline as well as through looking at strategic alternatives. Cibotercept data will determine what the path forward is, but it is not an inactive drug. We demonstrated in phase I healthy volunteer studies that it improves bone mineral density in a very significant way. It was safe. There is potentially a path forward in PAH or in other indications, right? It is not a molecule that does not have any legs. With 065, we are going into boys with DMD as the next step. Subsequent to that, we will be looking at other neuromuscular indications that we have not as yet announced. I do believe we have the opportunity to generate data relatively rapidly, but it is not going to be immediate, right? I do believe in the value of these molecules. They have very powerful biology that can be leveraged in these indications. I'm one of the inventors on luspatercept and on sotatercept that are approved. 050 elritercept, we did a very nice deal with Takeda, and we moved it all the way through to phase III on our own. I think I do understand the biology, perhaps better than most people in this pathway, and I do believe there's value here. You know, it doesn't negate what we do in our research pipeline as well. Got it. Makes sense. Let's dive a little bit deeper into the pipeline. You know, appreciating that Takeda is now helming development of elritercept, but could you remind us of next steps? I mean, realistically, when could we see the asset enter the clinic for both myelodysplastic syndromes and/or myelofibrosis? It is in a phase III trial in MDS as we speak in a second-line setting. I think luspatercept showed that in a second-line setting, it achieved somewhere around $800 million on an annual basis. When it got approved and the label was expanded to frontline in 2025, it looks like the numbers are getting close to $2 billion, right? That's because you're starting the treatment early and therefore the patients are remaining longer. With elritercept, we demonstrated that the median duration of treatment is 130 weeks, whereas with luspatercept in the same setting, it was 30 weeks, right? I think you can treat the patient for a longer period of time. Takeda is very excited about the program. They continue to move it in second line, but they want to introduce it into the front line. They are also excited by myelofibrosis as well. I think we found the right partner that really shares the vision that we had for the drug in MDS myelofibrosis, but we would not have had the capital resources to start a front line until we had approval in the second line. It really accelerates the value proposition for the program through the partnership. What do you think elritercept needs to demonstrate to kind of reach that best-in-class mantle versus next-in-class? One of the things that is really critical is how patients feel. Patients spend a lot of time getting transfusions. When they go on a treatment that makes them feel better and gives them transfusion independence, that's the ideal solution. Luspatercept, in some patients, they don't tolerate it well. They feel fatigued, even though they are achieving transfusion independence. Now this becomes a discussion between the physician who says, "You're doing really well because you don't need transfusions." Transfusions are not benign, and yet the patient is not feeling great, and they want to come off the drug and go back to transfusions. If you have got a drug that can actually provide that benefit for a long duration of time, which for a physician is great, you put them on the treatment and then they just continue on it for an extended period of time, and the patient feels good, right? That I think is the ideal. We have demonstrated with elritercept that the MDS patient, they are not only becoming transfusion independent, but they are actually feeling better. Therefore, their quality of life is better. Makes sense. You've pledged to share the TROPOS phase II data with investors later this quarter. What are the thresholds or benchmarks for moving cibotercept forward, do you think? Yeah, I think we need to wait for the data to look at that data and see was there a set of patients based upon their baseline characteristics or their concomitant treatments because these patients are on multiple medications that results in them being prone to increased risk of pericardial effusion. If we can define that, then I think we can chart out a path forward with another phase II trial to test that hypothesis before endeavoring on an expensive phase III trial. I think that is the path that we can chart depending upon the data. Now, should there not be a path in PAH, as I said earlier, the drug is active. It was safe in healthy volunteers, and we don't have time to discuss it today, but we have learned over the years that in PAH patients, their vasculature is altered. We think of the vasculature of the pulmonary tissue where there's pruning and thickening, but in other tissue, that vasculature has become more fragile. What is pericardial effusion? It's the balance of fluid into the pericardium space and fluid out. Altering that could result in increased fluid. I don't know how that's happening at the moment. It wasn't described before, but it is a background in PAH patients. Since we announced pericardial effusion with cibotercept, we're hearing of more and more cases with sotatercept as well. This may well be biology in the patients that through this pathway, you're also providing benefit, and yet there is some additional biology in terms of the pericardial fluid retention. Got it. Maybe in the time we have left, 065, without getting too much into it, no instances of nose or gingival bleeding, telangiectasis or serious adverse events. That said, there was an increase in hemoglobin. Is this a cause for concern, and do you feel confident moving forward? Yes, I do feel confident moving forward because the increase in hemoglobin can be monitored, and therefore you can do dose adjustments accordingly. It is not as if there is no way of monitoring it, right? It is an easy lab measure. We saw in the SAD smaller changes than in the MAD. The MAD was overweight and obese individuals. They have inflammation, okay, right, to a greater extent. I think the differences that we saw in the obese individuals may partly be due to the differences in the patient population or the population that we had. We can titrate to the level of target engagement to see the efficacy. I'll just go back, okay, right, to a similar molecule, ACE-031, that went into boys with DMD, and at low doses of 0.5 mg per kg given monthly, you saw increase in skeletal muscle, decrease in fat mass, improvements in bone mineral density, and what looked like in the study, functional stabilization as measured in a six-minute walk test, right, with three months of treatment. That program was put on clinical hold for exactly the reasons of nosebleeds, gum bleeds, and telangiectasias. We have dialed that out. I do believe that we have the opportunity to be able to demonstrate efficacy and benefit in these patients while maintaining the hemoglobins at a level, okay, that is acceptable, which is less than 1.5 grams per deciliter. Perfect. I am afraid that's all we have time for, Jasbir Seehra. Thank you so much for joining us. Thank you for giving us the opportunity to have the discussion. Thank you very much. First day of BofA's 2025 Healthcare Conference. My name is Jason Zemansky. I'm one of the SMID-Cap Analysts here at the bank. For this slot, I'm very pleased to have join us on stage here, Todd Harris, CEO, and Doug Warner, Chief Medical Officer of Tyra Biosciences. Gentlemen, thank you. Thanks for having us. Thank you. Perfect. Maybe just to start broadly for investors who may not be as familiar with the story, you know, would you briefly describe Keros's platform? I guess when you think, what differentiates the elements of the platform from other developers focused on FGFR inhibitors? Thank you again for having us. At Tyra, we are a company focused on small molecule precision medicines and in the FGFR family, largely with our lead molecule that is an FGFR3 selective. We do all of our chemistry in-house. We use what we call our SNAP Chemistry Design Platform. It is a structure-based drug design platform that is highly iterative. Really, what is unique about that platform and capability is it allows us to solve very hard engineering problems like we did with making an FGFR3 selective molecule. Those FGFR inhibitors that have advanced into the clinic and even been commercialized are all pan-FGFR inhibitors. They hit FGFR1, 2, and 3 close family members very similarly. They do that because the active sites of those three family members are identical in that first shell. There is very little chemistry that you can take advantage of in terms of different amino acids. We leveraged our structure-based drug design approach to understand very fine differences in the active site that we ultimately exploited to make a selective FGFR3 inhibitor, which is really the first to go into the clinic and has now read out very positively in the clinic in our phase I MEC study and highlighted really the capability of the platform making something as selective as we did. Got it. I think that's a great segue into the first question here in NMIBC. You know, when you think about the space, I mean, ADCs have proven fairly effective for treating it, but what are some of the unmet needs? You know, are there distinct populations where you think an oral FGFR3 inhibitor would be especially attractive? Yeah, let me hand that over to Doug. Sure, thanks, Todd. In NMIBC, there are different risk categories. In low-grade intermediate risk NMIBC, that's about 70-80% of cases having FGFR3 alterations. While ADCs have shown success in the metastatic setting, in the earlier settings like NMIBC, the focus has really been on TURBT surgery and chemo, and there are very high recurrence rates in those settings. For instance, at one year with standard of therapy of surgery, TURBT and chemo, recurrence rates are about 30%. At two years, they're about 40%. While there are a lot of different drugs in development, all of those drugs are procedure-based and involve intravesical administration of agents, where TYRA-300 is an oral therapy, so it really solves a need there. Got it. You know, when you think about duration of therapy classically in NMIBC, it's historically been longer than other indications. I guess, you know, how sensitive do you think regulators are to side effects? I guess what gives you confidence about moving forward with 300? Yeah, so I think there certainly is an emphasis on safety and tolerability in that setting. First of all, because these patients are generally very healthy. Secondly, as you mentioned, there is long duration of treatment. Based on our work in MUC in the metastatic setting, where we tested much higher doses but evaluated lower doses, we feel the doses we are evaluating at 50 and 60 milligrams q.d. will be very well tolerated in this population. We do not foresee an issue there. Great. About how quickly do you think you'll be able to generate top-line results? Are there any signals that you think would be especially de-risking this early phase? Maybe I can take the latter part first. Complete response at three months would be an early signal of activity. That would give us a real understanding, early understanding of efficacy that'll help us de-risk going forward. I'll hand it over to Todd to talk. Yeah, and in terms of timelines, we're excited that, you know, we're activating our first sites this quarter. We're anticipating our first patient dose this quarter. We're really looking to fill these two cohorts of 50 and 60 milligrams as we get throughout the year, which should get us, you know, we haven't committed to a specific timeline, but as we get into early next year, an opportunity for that initial three-month CR rate to be discussed. Obviously, we'll want to follow those patients for longer, but that CR rate really is, I think, the critical top-line data. We're talking about a dose that can essentially chemo-ablate the tumor, which would very much likely be a dose if maintained, would keep those tumors at bay for very prolonged periods of time. Got it. Okay. Switching gears somewhat, you're also evaluating 300 and achondroplasia with phase II BH301 underway. I guess similar question as before, the CNP class has demonstrated solid efficacy, but where do they fall short? Yeah, it's a really exciting space, and it's really exciting to have options for kids. You've got VOXZOGO that has an accelerated approval. It's on the market. It's generating, you know, what would be about a 1.57 cm incremental benefit each year for these kids. Their phase II, phase III data, you know, read out approaching an analyzed height velocity of about 6 cm total versus what would be expected in the natural history of about 4 cm. Where that falls short is kids without achondroplasia are going to be growing about 7.6 cm per year from age one plus. And BioMarin just had some data over the weekend where they highlighted that if you were to treat a child from six months and inject them every single day with VOXZOGO up through those growth plates, they could expect optimistically that that kid might be able to grow 26 cm. The gap is more like 46 cm between typical height adult male and a male with achondroplasia. You're about addressing half of the final adult height benefit. All of these things are important as surrogates for the key clinical sequelae, which are spinal stenosis, foramen magnum stenosis, reach gait, and some of the pain and surgeries associated with the condition. Being able to truly modify the target, which is FGFR3, such that you're able to move the FGFR3 activity to something that's more physiologic could really close that gap. Even, you know, Ascendis, they're generating very similar data with CNP. There's really no more dose response curve with CNP. You can give more. It doesn't change that analyzed height velocity above 6 cm. That's different with FGFR3 inhibition. Now, we've proven the concept thanks to the great work that BridgeBio did with infragastrinib in a phase II. They also were able to get to about 6 cm AHV, but they're just beginning on the dose response curve. They're using a dose that's very low. It's one-sixth of a dose that you would use in an oncology setting trying to engage the target with an IC90. That gap of sort of one-sixth of lower dose highlights just how minimal their target engagement is. Kids that do go on full oncology doses, drugs like erdafitinib or Debiofarm had a drug as well, and they would do this because of brain tumors, when they have open growth plates, during that period of time, you see massive acceleration of growth, on average about 19 cm per year. That highlights the dose response curve we're talking about, which is a very, very wide dynamic range. Now, you wouldn't want to intentionally target that IC90 inhibition of FGFR3 with any of these kids because that level of and speed of growth can lead to overgrowth challenges, something called skiffy or fractures that would be problematic. However, we can do better than these existing therapies. You know, growth hormone, for example, in the first year typically takes kids to about 8.5 cm per year very safely in achondroplasia. VOXZOGO, when used in conditions where FGFR3 is not overexpressed, so FGFR3 wild type, that break is not in, they're able to move kids with short stature from 4 cm a year to 8.5 cm per year very safely. That's really where we're focusing. We want to get to that 8.5 cm per year. It will double the efficacy of what everyone else has seen in this indication, and there's really strong evidence that that can be done safely. We know we just need to hit the dose response curve. With the data we've already read out in the fall, we know we have the therapeutic index all the way up to half the oncology dose, which is as high as we'd want to go to avoid overgrowth. We don't see any hyperphosphatemia. We don't have any ALT or AST increases. We don't really have any AEs of significant concern. We're going to be able to engage the target beyond where infragastrinib was able to go with their very low dose and ideally hit that, you know, analyzed height velocity surrogate and ultimately make a much more meaningful impact on the clinical sequelae that are as most important. Got it. Maybe let's switch gears again. Obviously, as you mentioned earlier, we saw the SERF301 data last fall. You know, as you think about 300 and how that's differentiated from both J&J's erdafitinib, but also Lilly's Loxo, you know, when you think of that constellation there, where does 300 fit in? So, you know, first of all, because it's selective for FGFR3, we see clear differentiation when it comes to side effect profile against erdafitinib. So erdafitinib, you hear providers say it's as bad as chemotherapy because it really is. You have nail toxicity such as nail loss, nail infection at over 70%. It's single digits with TYRA-300. You have stomatitis or mouth ulcers over 55%, single digits with TYRA-300. So there's a real difference there, as well as efficacy where we saw a 55% response rate versus 35% with erdafitinib. With Loxo, it's BID dosing. And what we saw there is they haven't chosen a dose. They presented data looking at three different doses, but you have very high rates of diarrhea at around 76% and high rates of hyperphosphatemia at around 36%. So we see real differentiation there both on efficacy as well as safety. Are you winning on efficacy because your patients are able to stay on longer, i.e., the safety profile? Or are you winning on efficacy because you're winning on efficacy? Right. So it's likely a combination. You know, it's still early days, but certainly being able to stay on a highly active drug longer is going to result in more efficacy. Erta, its efficacy is pretty limited. Not only does it have a 36% response, 35% response rate, but the duration of response is actually less than five months. Some of that's due to non-selectivity. And with efficacy, other of that is due to non-selectivity resulting in toxicity. Got it. As we think about the updates as we move forward here, what are the benchmarks for success? Yeah, so our focus, you know, really prioritization is reflected in the phase II's that we're opening. NMIBC, ACH, these are our big and exciting opportunities, very large markets. For NMIBC, the benchmark, I think, has been really well set on efficacy when we look at both TAR-210, the locally delivered erdafitinib, and then erdafitinib given orally at a low dose of 6 milligrams. In that oral systemic exposure, they got an 83% CR rate. That's, you know, an excellent efficacy signal. However, it came with all of the toxicities. As we've tested the profile with KOLs, they have highlighted, especially for an oral, that even a 70% CR rate would be very attractive as long as you're choosing a dose that's really well tolerated. Part of our strategy here is evaluating multiple doses, trying to find really the most tolerable dose that can hit that 70% CR rate because we think that will get the most use and, you know, have the most long-term benefit for chronic dosing with patients. I think I already talked about the achondroplasia benchmark. Here, we're looking to exceed that 6 cm, get to 8.5 cm AHV in terms of our initial readout from higher doses in that phase II. Got it. In the brief time we have left, obviously you're well capitalized. I think it's at least through 2027. You've got three phase II underway, three different conditions. Then on top of that, you've got 200 in ICC and 430 in HCC. What are the puts and takes here in terms of prioritizing programs? Yeah, we've been really focused on making sure that we manage our cash effectively. We're in a great position with, you know, $318 million as of our last quarterly release, as you mentioned, being able to have money through 2027. A big part of that too is managing where do we want to prioritize that cash. We'll be able to get the key phase II readouts in NMIBC and in achondroplasia. Those are our top priorities. We'll be able to get our dose escalation and some initial dose expansion with TYRA-200 and 430. One of the things we're being intentional about in the metastatic urothelial setting is that we don't want to necessarily at risk rush into a phase III, which is going to be a very expensive spend while we're waiting to get this very meaningful phase II data out from these other indications. We're actually pulling back a little bit, not opening additional sites there, whereas we are opening a ton of sites for ACH and NMIBC. That's a very intentional exercise so that we've got the levers to make sure we can manage our cash through getting all of the data that we need to see. Perfect. Exciting story. Todd, Doug, thank you so much for joining us. Thank you. Appreciate it. Thank you.