All right. Good afternoon, everyone. Thanks for joining us here at the SVB Leerink Global Healthcare Conference. My name is Tom Smith. I'm one of the Senior Biotech Analysts here at SVB Leerink, and happy to welcome our next company to the virtual stage, Keros Therapeutics, and their CEO, Jas Seehra. Jas, thanks so much for joining us. Appreciate it. Thanks, Tom, okay, f or that introduction, and thank you for giving us the opportunity to present the exciting story of Keros. Cheers, cheers. Just before we jump into it, just a quick reminder for the audience here. If you have any questions for Keros, you can either ask them via the webcast portal, or you can send them directly to me via email. With that out of the way, Jas, I think you have a couple slides you're going to use to give us an overview of the company, and then we'll jump right into Q&A. Thank you, Tom. So, before I go into the slides, the usual disclaimer about forward-looking statements. What I'm going to tell you, that Keros is harnessing the powerful biology of the TGF-beta superfamily. We are a clinical stage company that targets the TGF-beta superfamily to develop novel therapeutics. What's really exciting about this biology is that this has been validated with marketed products, INFUSE, a bone morphogenetic protein-2 [inaudible] fusion, REBLOZYL, okay, a modified activin receptor for treatment of anemia in beta thalassemia, as well as in myelodysplastic syndrome. And really, what we are able to do at Keros is that we're able to leverage our extensive experience in the TGF-beta superfamily, knowing the structures of these proteins, their function, and couple that with protein engineering to generate a clinical pipeline of differentiated products. So you see, okay, the first three product candidates on this slide. 050 is a modified activin receptor IIA ligand trap, and it's designed to address ineffective hematopoiesis by modulating this pathway. Based upon our phase I data, we believe that it has the potential to correct multiple cytopenias in patients with ineffective hematopoiesis. That includes MDS and myelofibrosis. Our second product is a small molecule kinase inhibitor targeting another member of the TGF-beta superfamily, activin receptor-like kinase 2. This is being developed for treatment of anemias where iron imbalance is the underlying cause. That includes rare genetic diseases such as iron-refractory iron deficiency anemia, to the broad indication of iron deficiency anemia which includes chronic inflammation in there. And then, there is the opportunity for the same molecule in a very rare genetic disease called fibrodysplasia ossificans, where following injury, skeletal muscle and tendons turn to bone. We'll be starting phase II studies with this program in 2021. Then lastly, a preclinical asset that will be entering the clinic later. Another activin receptor ligand trap that is being developed for treatment of disorders associated with bone loss and pulmonary arterial hypertension. I think what you see here is the unique position that Keros is in, where our discovery pipeline is yielding us multiple product candidates and will continue to build the pipeline in bone, muscle, and pulmonary diseases. This is a pipeline that is shown on a chart, and you can see that 050 has completed its phase I studies. It's in its first phase II study in MDS patients and will report the initial data in mid 2021. We're also starting the phase II study in myelofibrosis this year. 047, which completed its phase I towards the end of last year, we'll be initiating two phase II trials, one in IDA and the other in IRIDA in 2021. Lastly, 012 will enter the clinic in the second half of 2021. You see a nice progression of assets, okay, advancing through the pipeline and new ones coming in to fill that pipeline. The anticipated milestones are shown on this slide, where we will be announcing the initial data from our phase II trial in MDS with 050 in middle of the year. Starting that myelofibrosis study in 2021. The two phase II trials with 047 in 2021 and we'll be presenting data from the 012 in PAH at a major conference in 2021 and starting the clinical trial in the second half of 2021. With that, I want to thank you, okay, for the opportunity and leave it to you, Tom. Great. Thanks, Jas. Really nice overview. You alluded to the mechanism. I want to start with the mechanism for KER-050. You talked about it a little bit, but maybe if you could just go into a little bit more detail around differentiation between KER-050 and say, luspatercept and sotatercept, the other ligand traps that most investors are familiar with. Yeah. I think all three molecules are derived from the activin receptors. The difference is that luspatercept, the approved product, is derived from activin receptor IIB with a single mutation, in that actually renders it incapable of binding to one ligand, activin A. sotatercept binds to multiple ligands, activin A, activin B, GDF8, GDF11, and those have all been published either in peer-reviewed journals or alternatively in the patents from Acceleron. The difference between sotatercept and luspatercept is that luspatercept does not bind activin A. KER-050 binds all the same ligands that sotatercept does, activin A, activin B, GDF8, and GDF11. And therefore, it has some of the same biology that has been observed with sotatercept. Sotatercept, based upon both preclinical and clinical studies that have been published, increases bone mineral density as well as increases red blood cells. On luspatercept, according to all of the published work, only increases red blood cells, does not have effects on the bone axis. KER-050 actually increases red blood cells, increases bone mineral density, and reduces fibrosis. You begin to see some of the impact of that it is indeed increasing red blood cells in healthy volunteer studies, but also increasing platelets. That's because there is an interplay, [inaudible], okay, right, that's ongoing in the bone marrow between red blood cells and platelet precursors, megakaryocyte. So 050 is the first molecule that has demonstrated increases in platelets in healthy volunteers. And we think this really could be a meaningful differentiation for 050, whereby in MDS patients, in the low risk to intermediate risk for developing to AML, 90% of those patients are anemic. 30%-40% of those patients have low platelets and/or low neutrophils, so they're thrombocytopenic or neutropenic. And therefore, an agent that works on multiple lineages has the potential to correct the cytopenias that are in those patients. Some of the first-line treatments in MDS are in fact the hypomethylating agents, such as azacitidine. Okay, right? They actually cause —they relieve the anemia, but they cause thrombocytopenia. Therefore, there is a significant unmet need for a molecule that actually does treat the thrombocytopenia in patients. We think that's a real advantage. Thinking about it in myelofibrosis, in myelofibrosis, the primary defect in the JAK-STAT pathway results in actually overproduction of megakaryocyte precursors that fail to develop to platelets. Those megakaryocyte precursors break down, cause inflammation in the bone marrow, and set off that vicious cycle where you don't get red blood cells, you don't get platelets, and then eventually you end up with extramedullary hematopoiesis in the spleen, which then causes all of the other symptoms, right? JAK inhibitors really only treat the symptoms. They're not disease modifying. So if you could actually allow those megakaryocyte precursors to mature all the way to platelets, you actually reduce the inflammatory signals in the bone marrow. Therefore, there's a potential you could reduce the fibrosis and therefore, [inaudible] start having a disease modifying role. That to me is very, very exciting. I think that's how 050 gets differentiated from other molecules of similar lineage. Okay. Yeah, no, I appreciate the distinction there. I guess I'd like to go through the phase I data in a little bit of detail, but if we could maybe focus on the erythropoiesis pathway and one of the potential benefits here is you're acting earlier in the pathway. You think about potential application of that, perhaps you see a broader impact or an impact on anemia in a broader segment of the population. What clinical data, Jas, or what preclinical data do you have that gives you confidence that 050 is actually acting on the earlier parts of the erythropoiesis pathway? Similarly, how is that differentiated from either sotatercept or luspatercept? Yeah, I think, okay. It's sporadic or small pieces of data that will give you that confidence. The first one is actually the preclinical data, where mechanistically you see what's going on. What you see is the treatment from a single dose of 050 results in rapid increases in red blood cells. And remember that in the journey of a common myeloid progenitor all the way to a red blood cell, in humans, that takes 21-28 days, whereas in rodents, it takes about seven to 10 days. If you see increases very quickly, that means that you're actually accelerating the progression of precursor cells that were already almost completing their journey. Preclinically, we see those increases as early as 12 hours. We've only looked at 12 hours. We haven't looked any earlier, but we see those increases. That means those red blood cells almost completed their journey. But then we continue to see that increase through day 14 that we publish in our ASH presentation. When you look at the clinical data, you see the same thing, right? That you see that there are increases on day two in hemoglobin levels, red blood cells, and reticulocyte, and those increases continue. The C-max of the drug in our phase I study is achieved on day four. On day four, you've got the maximum concentration of the drug. And by day 15, where the drug levels have gone significantly lower because of the half-life of the drug being 12 days, you still continue to see increases. In fact, those increases are occurring through day 29. And if you look carefully at the data that I've done, you see that some of the participants in that study had increases through day 42. That means there are cells that are coming out into the bloodstream that started their journey 21, 28 days earlier. By that point, there's no drug on board. That really does fit the preclinical data that you're acting at all stages of erythropoiesis. Furthermore, we've seen that with 050, we get increases in erythropoietin in circulation, in rodents. That again means that you're helping assist those cells at the earliest stages, which is being committed to go down that erythroid lineage. I think that all of that data tells me that there is potential of this molecule to be working at multiple stages. In patients with ineffective erythropoiesis, they're a heterogeneous population. They don't have interruption of these pathways at one spot. Okay, right? Otherwise, they would all be very homogeneous, and therefore you would get a treatment, okay, right? Where everybody's responding. And therfore-- as a consequence, I think of it as, well, you have the potential where REVLIMID showed really nice data, okay, right? with respect to increases in a small subset of the MDS patients, these RS patients, and they got a response of roughly 40%. Why aren't the other patients responding? They must have other perturbations in the pathway that are preventing the maturation of a red blood cell. I think given what we've seen, given the broad mechanism, there is a potential that we can treat RS, non-RS, and perhaps even have a different response rate. Okay. Yeah. No, I appreciate those comments. I guess shifting to the phase I results and maybe the phase II study design here, where you're going to have initial dose escalation data in MDS patients in mid-2021. You know, if you could talk a little bit about how you're thinking about dosing, and then a little bit about the design of this dose-ranging study and some expectations ahead of the initial data cut. How many patients, potentially how many dose levels? Obviously, you're still running the study and collecting the data, what's your sense for what investors should expect ahead of this data readout? Yeah. So, we haven't actually shared what our starting dose is, okay, right? But I think if you go back to our phase I study where the lowest dose in the SAD that we detected drug in circulation was the 0.5 mg/kg, and then the lowest drug that we tested in the MAD was 0.7. That gives you sort of the range for the starting dose, okay, right? We know that at those doses, we got some increases, okay, right, in reticulocytes and increases in red blood cells. Therefore, we were getting target engagement. Think of those two numbers as sort of the starting dose. After that, dose escalation is very simple. It's the usual Fibonacci design of a study. The first number is one, two, three, five. You're dose escalating where the next dose after the first one will be double, okay, right? Then the next one is triple, and so on. We have four cohorts planned in the dose escalation. It's an open label study where there are six participants in each dose level, three RS, three non-RS. Based upon the safety data, you'll escalate to the next dose. And in the RS, in each cohort, the patients will receive four doses at 28 days interval over a three-month period, then be followed for another three months. That's what the study is designed, and what we're looking for in part one of that study, okay, right? Which is dose escalation. We're looking for signals of activity. Does this drug show changes in red blood cell param— does this drug show changes in platelets? Okay, right? That's the lowest level. Okay. When you start thinking about beyond that, how durable is that effect? Okay, right? Is it a few weeks? Is it a month? Is it two months? Okay, right? Because then that starts giving you information about how you're going to treat the patients. Okay, is it a monthly dosing? Is it less frequent? We believe, based upon our phase I data, that we can dose on a monthly schedule. Could we dose on a less frequent schedule based upon the mechanism? We don't know. We have to find that out in the patient. What are the doses at which we begin to see responses in RS versus non-RS? Are they different? Okay, right? These are patients. They are— okay, right, already have the disease. Some of them would not have received any transfusion, but others would have received transfusion. What about those that have received transfusions? Are you seeing, okay, the same signals of activity on the red blood cell and on the platelet axis, right? And if you are seeing those, are some of them having to skip transfusion because they don't need? You're going to get a lot of data, but it's going to be little bits of data, because you're going to get some patients responding, others. That gives you clues about how you want to start thinking about your part two, which is the dose confirmations of the study, where you have a larger number of patients. You start to get experience in terms of what your response rates are, and also, okay, right, you're treating now in your open label extension, and you're getting your experience as to what the durability of the effects are. Right. Okay. Yeah, no, that's really helpful. Just in terms of patient numbers, Jas, and understand that you're still enrolling the study and you don't have perfect visibility into that, but it is an open label study. What's a reasonable expectation for number of patients at this initial data cut? Tom, we haven't shared that, okay, right? What I would tell you is that we started the study in the second half of last year. As I just shared, the actual treatment period of the study and follow-up is six months. Prior to being dosed, the patients have to be followed for two months so that you get their baseline characteristics for their hematology parameter. The patients are in the study essentially eight months, okay, right? So, you can imagine, here we are in February. We're moving on this. I can't tell you how many patients we'll have. I think what I can tell you is we're progressing along that study. It is an open-label study, so we'll share the first of the data in mid-year, we'll continue to give additional updates as we get more data. Okay, okay. Appreciate that. Maybe we can talk— you mentioned also initiating the study for 050 in myelofibrosis. Maybe if you can talk to a sense of timing for kicking that off, trial design, and then what you're looking for in terms of initial signals within the MF population, how that compares to MDS? Yeah. So, I think, first of all, okay, right? We'll be starting that study this year. We haven't given any more guidance than that, nor have we shared the design of the study. I think previous programs have actually hinted at what that study is going to look like. It's going to be a study as a monotherapy in myelofibrosis patients, as well as those that are on JAK inhibitors, okay? And again, here, the ability to modulate platelets could be a real advantage for us. In those patients, it's going to have to be a longer treatment, okay, right? Probably six months and longer. What you're looking for is, are you correcting the anemia? Are you actually correcting the thrombocytopenia in those patients? As you progress, what are the impacts on the bone marrow there? One of the things, okay, that happens in these patients is that as that aberrant bone marrow is expanding, the inflammatory signals are actually causing bone resorption. You get osteoporosis. Do you, okay, right, now begin to see a reduction in that? You can follow that through bone biomarkers and so on. We'll be looking at all of that. I think that's where, okay, right, returning that bone marrow to a more normal state has the potential to have disease-modifying properties to the molecule. Okay. Yeah. I mean, certainly, certainly a lot of potential as a pancytopenia agent, particularly within MF. Yeah, we'll be interested to see what the initial signals look like. Maybe I want to be conscious of time and make sure that we get through. You guys have a number of programs. KER-047 your ALK2 inhibitor. Maybe if you can talk a little bit about the plans here in IDA and IRIDA, and I guess also putting it in context with the phase I data that you announced late last year. Do you feel like you need further dose ranging here? Just walk us through how you're thinking about positioning 047 within these initial indications? Yeah. I think if you looked, our phase I data was really exciting data, because for me, seeing the biology once again translate from rodents to humans is really exciting. Our phase I data really showed, okay, right, that you can inhibit ALK2. As a consequence of inhibition, you get reductions in hepcidin, the master regulator of iron, and immediately you begin to see changes in serum iron, increases in serum iron. So you're mobilizing that. What was even more satisfying is that that iron is available for incorporation into red blood cells, where we saw increases in hemoglobin content of reticulocyte, the newly minted red blood cells that are found in circulation. That was all very exciting, okay, right? In our multiple ascending dose, our lowest dose was 50 mg dose. Even at the 50 mg dose, we saw decreases in hepcidin from that treatment. These are normal, healthy individuals. They already have low hepcidin. Yet we saw decreases in that hepcidin. As we think about going into the IRIDA, into the iron deficiency anemia, some of these patients are going to have much higher levels of hepcidin. So, when you have high hepcidin, do you actually get a bigger response? Do you actually need lower doses of the drug? In some ways, okay, right, knowing that we're seeing changes at 50 tells you you've got target engagement and you're initiating all of the downstream biology. I think, okay, right, exploring lower doses in patients that have high hepcidin is warranted for those reasons, because you may end up seeing similar or bigger effects in patients, even at lower doses. As you well know, the lower your dose, okay, right, the better in the long run in terms of safety observation. So I think it is warranted that we explore a lower dose in the patient. How many doses? I'm not sure that we need to explore too many doses there. It is likely that in the IRIDA patients that we'll do an open label study similar in concept to the MDS study, where you look at dose escalation so that you can see changes in these red blood cell parameters in a patient population that have high hepcidin. That then actually sort of feeds into what we do in the IDA as well. Okay, okay. That makes sense. Maybe if we could switch gears just quickly on KER-012. It's a compound I've been getting increasing number of inbound questions around, and we've seen some of the success with sotatercept in PAH, and I think there's obviously increasing enthusiasm around that approach. Maybe if you could just walk us through how 012 is differentiated versus sotatercept and you've talked about, I guess, filing an IND for 012 later this year, maybe how you're thinking about the initial clinical approach here? Yeah, so, look, 012 is based upon the activin receptor IIb, whereas sotatercept is based upon the activin receptor IIA. That doesn't mean too much. It's just they have different origins in that. I think where there is a difference is that when we were trying to come up with another molecule, we were looking for a molecule that would increase bone mineral density without increasing red blood cells. Sotatercept has demonstrated both preclinically as well as clinically that increases bone mineral density, but also increases red blood cells. In fact, the clinical development of that program was that in phase I study, okay, right? It was looking at bone, as well as red blood cells. The red blood cell effects was large enough that it became an anemia drug. More recently, it's pivoted to PAH because the biology suggests that it could have a benefit. But I think that's where the excitement is, because Acceleron did a really nice job of demonstrating the role of TGF-beta signaling in PAH. There is a limitation to the sotatercept. In their phase II study, they showed really nice response in terms of the six-minute walk test, the PVR, which is catalyzing their progression to a phase III study. They also saw increases in hemoglobin content in red blood cells. That increase at the highest dose they used, 0.7, was 1.5 g per dL, a mean change of 1.5 g per dL. And their lead investigator did say that she wants to use the higher dose. I asked the question, what about patients, okay right, that may need a higher dose? Could you dose them, okay, right, higher? Not with sotatercept. That I think is where 012 has the advantage, that it does not have that effect on red blood cell axis. We were looking for a molecule from day one that would have the effects of sotatercept on bone axis without the red blood cell. We published that data at ASBMR in September with respect to the bone and then in the updated S-1, we shared that we're seeing the activity in the PAH model. This makes me very excited. And I think given that we've got multiple indications that we could eventually go into, it again warrants us going into a healthy volunteer study, establishing what are the safe doses and the efficacious dose. The thing here is by having that effect on bone axis, we can measure that. We can measure that through bone biomarkers in circulation. We can measure that through imaging techniques and so on. You can actually get a feel for what the drug is doing pharmacologically, as well as determining the safety. That allows you to pivot in all the indications that you want to go into, PAH, okay, right, bone indications such as osteogenesis imperfecta. Got it. Yeah, we'll be looking for those initial early signals, it sounds like probably at some point in 2022. All right, Jas. Well, look, unfortunately, we're up against the time here, but thank you so much for spending the time sharing your insights, and we'll stay tuned here. Certainly, a lot of interest ahead of the initial MDS data coming up in mid-2021. Thank you everybody for joining us, and hope everyone has a great rest of the day. Thank you, Tom, for the opportunity. Appreciate it.