Edgewise Therapeutics, Inc. (EWTX) Earnings Call Transcript & Summary

Good morning, and welcome to the Edgewise Therapeutics Research Update. As a reminder, this call is being recorded, and a replay will be made available on the EdgeWise website following the conclusion of the event. I'd now like to turn the call over to Michael Nophy, EdgeWise's Chief Financial Officer. Please go ahead, Michael.

Thank you, and good morning. Welcome to the Edgewise Therapeutics conference call to discuss the differentiated mechanism of action of our cardiac sarcomere modulator, EDG-7500. This morning, we issued a press release announcing these findings, along with the presentation slides for this call, is available in the Investors and News section of our website at www.edgewisex.com. A replay of the event will also be available as a webcast on our website. Joining me today are Dr. Kevin Koch, Chief Executive Officer; Dr. Alan Russell, Chief Scientific Officer; and Dr. Michael Ayers, a cardiologist and Executive Director, Clinical Development and EDG-7500 Program Leader. Before we begin, I would like to remind you that some of the statements made today during the call are forward-looking statements that are subject to a number of risks and uncertainties. These may cause our actual results to differ materially, including those described in our reports filed with the SEC. You are cautioned not to place undue reliance on these forward-looking statements, and Edgewise disclaims any obligation to update the statements. I will now turn the call over to Kevin.

Michael, thank you. So I want to provide some historical perspective on the company and where we're going as a company and our new focus on the cardiovascular pipeline. We'll spend most of the day discussing the mechanism of action of 7500 and how that fits in the therapeutic portfolio for the treatment of cardiovascular disease. And then I'll end up with some near-term milestones for the company. As you all know, we started back in 2017 as a muscle platform company looking to invent and develop drugs for the treatment of orphan and rare muscle disorders with a high unmet medical need. We started our program working on identifying a drug for the treatment of muscular dystrophies -- that target of interest there was a type 2 skeletal muscle myosin inhibitor that was really a novel concept at the time. We ran a high-throughput screen, identified some leads, produced thousands of molecules and ultimately selected sevaseemten as the clinical candidate and filed the IND in 2020. Very rapidly, we were able to mechanistically validate the hypothesis that we would prevent muscle damage in muscular dystrophies and move that forward into Phase III. That Phase III will be reading out in -- by the end of the year. Along the way, though, a company, Servier found our data really compelling and made us an offer that really changed the nature of the company, actually acquired for $2.65 billion last July with the deal closing last July, $1.55 billion upfront and $1.1 billion in milestones. So it was a great milestone for the company and provided us significant cash to be able to propel and fund our cardiovascular programs. Great work by that team. So I want to turn back to the cardiovascular portfolio because really, the initiation of the cardiovascular portfolio came out of a counter screen for the skeletal muscle program. So when you run a high-throughput screen, you'll get a number of hits wer through about 2 million molecules. We found dozens of lead series. And then we ran a counter screen to make sure that we had selective skeletal muscle inhibitors -- during that time, we found a set of leads that were actually more potent on cardiac tissue than skeletal muscle tissue. And so we found molecules that were quite similar to what were found previously, which were type 1 cardiac myosin inhibitors, very similar to the drugs, aficamten and mavacamten. But we chose not to just follow that path and make a third-generation type 1 myosin inhibitor. We found another lead series that was quite unique. And what was unique about them, they were pharmacologically active, partial inhibitors and when tested in vivo, they did not lower fractional shortening or ejection fraction in the rat. And so they had a unique profile -- and then we started -- after we completed the IND of sevaseemten, we started to really focus in on the cardiovascular program and that particular lead series. We made several thousand molecules in that space and ultimately chose EDgeY-7500 for development in 2023, where we filed the IND. Just in the past couple of months in June, we reported on the 12-week Part D data and showed really in a lot of ways, validation of much of what we've seen preclinically with the drug. We'll be talking extensively about that in the coming 45 minutes. And so turning to the next slide. 7500 is poised to go into Phase III by the end of the year. We found a second-generation molecule that has a somewhat different metabolic profile optimized for HFpEF, which is now you probably saw on clinicaltrials.gov has been -- enrollment is open for the HFpEF study. And we won't say too much about 003, but we do feel like we have a lead series and are in the candidate selection mode. So we're very excited about the portfolio in cardiovascular disease and look forward to talking to you about how we discovered the drug and some of the key biologic nuances of this value proposition. Next slide. So I just want to give you a quick executive summary of what you're going to hear today. 7500 regulates the cardiac cycle. So essentially, we are controlling the rate in which the cardiac cycle moves throughout different parts of the cardiac cycle through control of the lever arm, which is of the drug binding to the regulatory light chain. This targeting of the regulatory light chain can produce really diverse phenotypes and pharmacologic profiles. There are RLCs that look virtually identical to CMIs, and there are regulatory light chain modulators that have very strong diastolic effects and a completely sparing of systolic effects, which is exemplified by EDGY-7500. Preclinically, EDGY-7500 slows early systolic pressure speeds up mid- to late coli and enhances early diastolic relaxation, which is the key pathology in HCM and HFpEF. Clinical findings have been consistent with these preclinical observations, whereas SY-7500 has reduced gradients, preserved net systolic function and improved diastolic relaxation, and we'll be showing some of that data today. More interesting, we'll show you a bit of data on the biophysical arrangement of the heads associated with the Libra arm and that by maintaining availability of the heads to reengage the actin filament during exercise, we believe that we'll be having a strong effect on endpoints like peak VO2 and KCCQ when a patient exercises or exerts themselves. and we'll show you data to support that position. So with that, I'm going to hand it off to Michael Ayres and allow him to discuss some of the details of the story.

Thank you, Kevin. When I went from running a full-time HCM practice to part-time, I did so because I find this narrative and story so compelling for our HCM patients, and I am just over the moon excited to talk to you today about this mechanism. So when we go around and we talk to investors, cardiologists, knowledge leaders or just anyone about this drug, 4 questions come up with some constancy. One, how exactly does this thing alleviate gradients? And two, how does it do so? How does it modify contractility without ultimately suppressing ejection fraction, previously thought of as the holy grail for obstructive HCM treatment. How does it improve diastolic dysfunction? And is it doing so in a unique way relative to other agents? And lastly, is there something in this secret sauce that allows for meaningful symptom and functional improvements relative to other things in the class. So I'm really excited to go through those 4 questions with you today. Let's start on the next slide with a little bit of a disease state background to make sure we're all on the same page. HCM is the most common inherited heart condition affecting around 1 in 500 people. It's dramatically underdiagnosed, misdiagnosed, and there are delays in diagnosis that impact care. It's defined as having a heart muscle thicker than 1.5 centimeters in absence of some cause. And broadly speaking, we like to classify it into two buckets or phenotypes. While both of these patients have thick and heart muscles, they have slightly different pathologies or problems. The first is your obstructed group. This group has a thickened heart muscle right at the outflow tract, which results in a divorcing of the pressure inside the ventricle relative to the aorta. There is a pressure difference. So now to generate your 120 millimeters of mercury systolic blood pressure, you might have to do 170 millimeters of mercury at work. The second subtype is the non-obstructive subtype, which also has thick and heart muscle, but doesn't have that misplaced thick and heart muscle right at the outflow tract. Though I'll point out, both the obstructed and non-obstructed group suffer from diastolic dysfunction or abnormal relaxation. When it comes to current therapies in this area, to be honest, there are some effective agents at relieving outflow tract obstruction, but they do so directly via negative inotropy they decrease how hard the heart is squeezing and they keep that highway out of the heart open. What is less well treated at present is the diastolic dysfunction that is shared in both of these groups. We know that around 1/3 of patients with obstructive hypertrophic cardiomyopathy have residual symptoms despite gradient relief, and we think that's from the diastolic dysfunction. And we know that current therapies have some diastolic benefit, but that benefit is coming directly at systolic cost that's ultimately tied up to echo monitoring for ejection fraction excursions. Let's talk a little more about HCM pathology on the next slide. So no matter your shape of the HCM no matter where your thickness is, there's kind of a shared pathogenesis at the sarcomere or muscle cell level. Let's start with what a healthy cell looks like. In a healthy cell, there are myosin that are available to buying with actin. And at any given moment, you might have 40% to 60% of these myosin that are participating in a given heartbeat while the other 40% to 60% are in and off or relaxed state. When you exercise, that number goes up. When you take a nap, number goes down. And at every day life, you modulate myosin availability based on need. In HCM, this system is disregulated. And now you have a myosin over availability such that basically all the time around 80% to 90% of your myosin are engaged in activities with actin, that results in is a hyperdynamic EF. It results in thickening of the heart muscle and in about 2/3 of patients it results in that obstructive physiology we described on the previous slide. What it also resulted in diastolic dysfunction, really from 2 things: one, Too many myosin heads are on at the end of the Hertz squeeze. And so it takes longer to pull them off. And each myosin head is a little stickier than normal. And two, because the heart muscle is thick, when you get to the end of diastole filling, the pressure is higher than it would have been in a normal heart. Lastly, on overactivation impairs what's called the cardiac reserve. Now the cardiac reserve says at rest, I might need liters a minute of blood flow. But with activity, that number might go up to 10. So I need to be able to change blood flow based on need. That thick and heart when has to be faster, can fill well and the cardiac reserve gets diminished. In this environment, the approach from cardiac myosin inhibitors makes some sense. If to many minds and heads are on, why don't we turn them off. The issue is that while this at first glance looks like a normalized state where you've moved back to that 40% to 60% of myosin availability. The inhibited myosin heads become semi permanently off such that when you need to recruit those miles and heads, you can't. This is really a pseudo normalized state. And let's talk about what issues come from this pseudo normalized state throughout the cardiac cycle. The cardiac cycle is made up of two components: the squeezing and the relaxation of each heartbeat. The squeezing is known systole when actin is bound to myosin and the sarcomere shortening, pushing blood out of the heart. Diastole is when myosin is popping off of Acton and the heart is relaxing. The cardiac myosin inhibitor is going to turn myosin heads decreasing contractility, keeping that outflow tracked open, but the benefit is coming not just because of -- but due to myosin turning off. There is a systolic liability directly tied to the benefit of gradient growth. In diastole, cardiac myosin inhibitors don't actually speed up the release of those sticky myosin heads to allow the heart to fill. What myosin inhibitors do is they just say, well, at the end of systole fewer mics and heads are on, it's going to take less time to pull them off. There is no direct speeding of that association. It's kind of diastolic benefit happens. Moreover, that diastolic benefit again is handcuffed to system cost. When we think about HCM, it's not a whole sale problem of systole or diastally. And that's why this broad approach of just turning myosin heads off becomes further problematic. We think of obstruction as a systolic phenomenon, and that's true, but it turns out that obstruction is initiating in early sleep when the mitral valve is moving in the outflow tract. And you have to have that moment in order for obstruction to then propagate through mid to late system. But it's early systole that is the most important moment to target potentially therapeutically. When it comes to diastole, again, there's abnormal relaxation. But as the heart rate speeds up, mid- to late diastole shorten and early diastole becomes more and more important. Said another way, when it's time to vacuum and your heart rate picks up, you lose some of mid delayed diastole. And early diastole becomes the sweet spot for cardiac filling. This is particularly true in HCM. So as we take a step back and say, with that in mind, what would the ideal HCM medication look like, we can come up with some conclusions. It would obviate or alleviate obstruction, but it would do so by targeting the initiation of obstruction in early systole and by decreasing pressure generation in early systole, obstruction would never initiate and therefore, it wouldn't propagate in mid to late systole. You could then utilize mid-to-late systole to catch up on that myocardial work. You could have a form of systolic procrastination where you're just doing the ejection fraction towards the end. In this scenario, there's no gradient and EF is preserved, potentially helping you escape things like echo monitoring for titration. In diastole, I've emphasized for you that it's early diastole that is really important for those myosin heads to come off. and the ideal therapy would directly speed up this process, and it would do so without handcuffing your benefit to systolic liability. Again, this idea that you could have your cake needed to. Lastly, the ideal therapy would save your cardiac reserve so that when it was time to bring in the trash from the curve, you could pick up your cardiac output. You could tolerate those faster heart rates. And when needed, you could increase how much blood was squeezing out with each beat. To do this, you have to maintain myosin availability. So the ideal therapy would do that, too. This kind of fine-tuning, I'm really excited to tell you we think is possible, and I'm going to turn it over to Alan to tell you a little bit more about that.

Thank you very much, Michael. What you can see on this slide is a more detailed exploration of the sarcomere. And let me just orient you to what you're looking at here. As Michael showed there's these 3 different states of myosin, the of the ready-to-bid the bound states of myosin. What I want to show you now is walk down the rod domain of myosin. So what you can see just south of those myosin molecules is 2 units. It's the essential light chain and the regulatory like can they're bound to the hinge region of that rock domain, and they essentially control the compliance of the lever arm that's responsible for generating the shortening force once myosin attaches to active. So regulatory light chain is an adjuster of force. It doesn't turn it on or off, but it modifies the efficiency of the way that myosin binds to act in and develops force. Now the regulatory light chain can be controlled through phosphorylation, and it has two main functions. The first function is to adjust the stiffness of the lever arm to modify the efficiency of force development and relaxation. And the second function of myosin is to adjust that position of myosin relative to active these off states to these real-defined states. Moving on to the next slide. 7500 essentially modifies both of these functions of the RLC. So what I'll tell you over the next few slides is a story that kind of adds some flesh to the bone of these simple facts associated with 7,500 mechanism. Firstly, 7500 reduces the stiffness of the lean, and there's consequences to that reduction of stiffness primarily slowing that early contractual process that Michael told you about and also speeding relaxation. The second function that 7500 caries is it actually moves those myosins towards active, moving them away from this off-state to this ready to buying state. So in many ways, in that context, it looks more like an activator than it does an inhibitor. We go on to the next slide. In the discovery of EDG-7500, we synthesized more than 1,000 analogs and subjected them to detailed mechanistic profiling. During this process, we recorded surprising mechanistic diversity in this class of compounds. EDG-7500 was selected to maximize effects on early contraction and relaxation without carrying a systemic liability, but these properties are unusual. A good contrast would be other ROC modulators that have been described in literature and have often been labeled as myosin inhibitors. By that, we mean that there's strong negative inotropes and inhibit contraction under most conditions. They also exhibit a narrow therapeutic index in vivo with strong inhibition of cardiac contractility and an increased liability of systolic heart failure. One of the nice elements of the JCI paper story is that we highlight this diversity by describing the enantiomer, the mirror image of EDG-7500, a compound called EDG-7499, which also leaves in most respects, like a cardiac myosin inhibitor. Next slide. Okay. So what I'm going to do now is just walk you through some basic target identification and then we'll get on to the nuts and bolts of the mechanism. So first, I'd like you to focus on the left-hand side of this image. What we have here is isolated myosin subunits. They exist in a dimer here. You can see that Rod domain, the RLC ELC and the myosin, and we're measuring ATP consumption in the preparation of these proteins. You'll notice a couple of things here. One is that the slope of the inhibition with 7500 let protein is very gradual. The second that you notice is that it's not complete. So the most that you can inhibit this system is approximately 50%. Next, what we do is we strip off the RLC sub units with a detergent. It's relatively easy to do this and then you can replace those subunits subsequently. Once we remove those ROC subunits in the middle figure, you can see that the compound is inactive. It is no longer modulating the OTP function of this preparation. If we then put RLC back again, in this context, is a recombinant human version of the RLC, you see that activity returning. So the activity is completely dependent upon the presence of that ROC subunit. Next slide. A second element of the target identification is to take that full length myosin and use an enzyme to clip up the lever arm of myosin and just leave the myosin head itself and those 2 ROC ELC subunits. What you can see is in the full length version, that's the dark blue dots. The compound is inhibitory as we showed before, with that 50% maximum effect. Once you chop off the lever arm, you render that preparation inactive to 7500. 7500, no longer has in the effect what that's telling you is that the meat and potatoes of the inhibitory effect of 7500 is that lever arm region and not the enzymatic function of the myosin itself. Next slide. Okay. What I'm going to do now is break you through the various questions that Michael proposed that we've received multiple times about how 7500 works. So the first question, how does it reduce gradients without suppressing contraction. The first thing we're going to do is use a pictorial trick so that you can visualize the nature of the inhibition from EDG-7500. This is actually not in the paper, and it's some beautiful work done in our in-house labs downstairs. So what you're looking at here are glass cover slips coated in those full-length myosin molecules. Then what we do is we have fluorescent acting to that preparation. And you can actually see the myosin moving acting around on the cover slip by visualizing that fluorescent acting. And you can see on the left-hand side, those that will act in filament creeping around quite efficiently. If we add a high concentration of 7500 in the middle, what you see is these actin filaments are still moving around, but that is -- it's much more slowed approximately 50% compared to the left-hand figure. The right-hand figure is a video, but it's hard to see because when you're at a high concentration of a cardiac myosin inhibitor, you've essentially arrested those filaments completely. You've removed all of the actin enzymatic function, and those at infillments are now unable to move. They're just frozen in time. And you can see that represented by the figure on the right-hand side. So 7,500 slows but does not stop the motion of myosin. Next slide. You can also see this slowing in other simple preparations that we use in the lab. In this case, what you're looking at is fiber bundles isolated from pig cardiac ventricle. Now these fiber bundles are then treated with a detergent that removes the membranes and you can actually activate contraction in these fiber bundles directly just adding calcium. In this experiment, what we do is we rapidly add calcium, and we look at the speed of force development. So on the left-hand side, we've added calcium at that 0 time point. And then what you see is the contraction increasing up to a plateau, then we remove the calcium and then that contraction goes down. What you can see in the blue line there is addition of 7500 slows that force development. Now there's a concentration response. The K act as just the rate of that contraction. And you can see very much like the ATP consumption figures earlier, a partial inhibition of this system with concentration response. So the more 7500 that you add, the more you slow force development. Next slide. Just to add a little bit of extra complexity and this is really important for 7500, that slowing effect is contingent upon how much calcium is in the system. As you stimulate the heart, calcium is released into the muscle fiber and directly interact with the actin filament or elements on the actin filament to initiate contraction. The more calcium you add, the more contraction that you get. What we found with 7500 is at high levels of calcium, the slowing effect is very, very modest. That would be the dark blue tringles there. Whereas at low calcium, you see much more slowing, okay? And that would be the open triangles. So as you go through this calcium cycle, you exhibit more slowing at low calcium and less slowing at high calcium. Next slide. So the theory here would be in early contraction, in early sister, EDG 7500 should slow that contraction and then it should have a lesser effect in the mid- to late part of system. We've actually measured this early contraction using a high frame rate echocardiogram measure of pre-ejection time. Now pre ejection time is that period of the systolic cycle where rejection is not occurring and pressure is increasing. It's the period between electrical stimulation of the ventricle and opening of the aortic valve. So what you see in these cases is a 17.3% increase in that pre-ejection time with 7500. So you're essentially increasing that -- the time that, that pressure increases and you're delaying the early pressure that leads to gradients in abstract hypertrophic cardiomyopathy. It is a very selective way of relieving gradients with this medicine. Next slide, and next question. How much 7,500 modulate hypercontractility without suppressing LVEF? This is a big question. We received this all the time. How can you turn down contraction and not see decreases in LVEF. And this comes on the next slide to the other function of the regulatory life check. And as I mentioned before, 7500 moves myosin towards actin instead of a waste. So I'll show you a little bit of data over the next couple of slides and just illustrate how the RLC can modify the position of myosin and 7500 can intervene in this process. Now on to the next slide. Now one of the wonderful things about muscle is it so organized, you can actually use X-ray defraction and scattering patents from that to fraction to be able to tell where those myosin heads are in relation to actin. So you can actually tell whether you have a red head, a green head or a dark green head essentially. On the left-hand side is this measure put into graph form. And we have a resting us these are peak ventricular tissues. And what you're doing is at rest, you're measuring whereabouts are those myosins relative to the actions? Are they close to the acting or are they further away? If that number increases, they get close to actin and if it goes down, they get further away. So if I added mavacamten, for instance, that number would go down as you move those myosin heads away from actin and towards the fit. Interestingly, when you add 7500 without adding any calcium, you're not activating this system, you move those heads away from the thick filament and into this preactivated state. So 7500 is influencing the position of myosin bringing it closer to active. Next slide. So this is a physiological trick that you can use in these pig fibers that you can activate with calcium. But instead of looking at force development from rest, now what we're looking at is force redevelopment once these fibers are maximally contracted. So you add calcium, the fiber contractually and then you rapidly shorten that fiber, which causes the myosin heads to disengage. They pop off transiently. And then what you do is you relengthen the fiber and you measure how long it takes for that force to redevelop. So this is a surge for how efficient those myosins are in the fully activated state. What you can see is 7,500 treatment under these conditions accelerates that force redevelopment. Not only does it accelerate it overshoots. So treatment with 7500 and this positional change of Myos actually preloads these myosins once calcium is present, to develop more force faster. So if you go on to the next slide, you can kind of see this all at once. So instead of just early contraction slowing with 7,500, you actually have this continuum where slowing turns into speeding. So as you progress through early systole into late mid-systole, now you have a catch-up phenotype in which that early slowing is overridden by that speeding phenotype. And we believe that's the secret sauce for why you don't see LVEF changes that pretty much any concentration you look at with 7500, you get that slowing of contraction. That's very important for obstruction relief in obstructive type cardiomyopathy, but it doesn't lead to any transient decreases in LVEF because of this catch-up phenotype. Next slide. So the third question, how my 7500 improve relaxation? I'm returning now to these fiber bundles that we have from the peak ventricle that we've removed the membranes and we can activate ready with calcium. You add the calcium, the fibers contract. And now we're looking at the other side of the spectrum. We're going to rapidly remove calcium. We're going to look at how fast these fibers relax. Top left-hand figure, you can see that addition of 7,500 accelerate that relaxation process. The green line is going down faster than the black line. And you can measure a concentration response for that. So KRL or rate of relaxation. And you can see as you elevate the concentrations of 7500, you increase that speed of relaxation. One of the super interesting things about 7500 that's very defining is that if you look at that concentration response for relaxation, it's actually more potent at doing this than it is slowing the contraction right? So it's quite possible to engineer an exposure of 7500 that accelerates relaxation without touching system at all. You can kind of see this in action on the next slide. So I'm going to use a preclinical story, and then I'm going to hand it over to Michael to tell you a clinical one. So in this case, what we did is 12 -- 6 months of treatment with EDG 7500 in the pig model of nonobstructive hypertrophic cardiomyopathy. So this pig has a mutation in the myosin molecule itself, NY H7 that leads to a nonobstructive phenotype. There's no obstruction in this pig but they get hypertrophy and they've diastolic dysfunction. In this 6-month study, we treated the pigs. And you can see on the left-hand side, there's no change in LVEF. So much of the same as in the humans. We have a decreased contraction of the heart. At the end of this study, we instrumented these peak so that we could actually measure pressure volume relationships in the heart. And you can see 2 very clear phenotypes here. The first in the middle is that end diastolic pressure has decreased in the presence of 7500. It's elevated in the untreated peaks because they have diastole dysfunction and residual end diastolic pressure that's relieved with the compound. On the right-hand side, this is a matter of ventricular compliance, the end diastolic pressure volume relationship. And you can see that, that's also decreased. So you've improved ventricular compliance and the ability of that tissue to relax two great signs that you're directly affecting relaxation with this compound without cost on LVEF. So at this point, I'm going to hand it back to Michael to describe clinical effects and then talk about exercise capacity in cardiac reserve.

Thanks, Alan. So rather than speeding up relaxations by merely having fewer myosin heads on board attached to actin at the end of systole, we think we're directly speeding up that association. And that is manifest clinically. What we see here is what's called e-prime measurement. So your heart is shape somewhat like a bullet with the apex or tip, and the flat part of the to the bullet being the mitral valve. When the heart squeezes that bullet shortens, and then when the heart relaxes that bullet is going to lengthen as the mitral valve moves away from the tip. We actually directly measure the speed of this mitral valve movement in early diastole. And it's a way for us to capture how much of a suction cup effect you're seeing in early distally. How much is the ventricle actively pulling blood down through the mitral valve to fill for the next beat. In obstructive physiology, there are many changes occurring within the ventricle. Systolic pressures are dramatically changing, and that actually influences some of our e-prime changes. As such, as a clinician, I think the purest place to look for the lusitropic benefit of an HCM med would be in a non-obstructive cohort. So I'm going to draw your attention to the right side of this slide. As a point of reference, if you were to look at, say, Redwood Cohort 4 or what kind of e-prime changes you're seeing with myosin inhibition, that's going to give you e-prime augmentation speeding up of relaxation of around 16% to 24%. What you see here is the proof in the pudding of the lusitropic benefit of this medicine with a 37% increase in e-prime, this goes hand-in-hand with robust improvements we saw with another filling metric, over e-prime in this same cohort. Now keep in mind that Alan showed you that the lusitropic relaxation benefit of this medicine is more potent earlier than the changes in systolic contractility and changing how the ventricle squeezes, you get the relaxation benefit first. We think that this is manifest on this graph, where we have an obstructive patient who has given a low dose of 25 milligrams. And you can see at week 1, the gradient hadn't yet moved. You hadn't lowered early systolic pressure enough that you had initiated gradient relief. But despite the gradient still being present, the BNP, which is a measurement of wall tension is going down. We think this is reflective of the fact that even at 25 milligrams of this compound, you're getting a lusitropic benefit that is showing up in blood work prior to the gradient being relieved. So lastly, let's go through why we think that moving into a Phase III trial, 7500 might translate into really noteworthy meaningful improvements in symptoms as well as functional capacity like those seen in a cardiopulmonary exercise test. So we mentioned earlier at rest, your output is made up of how many times the heart is beating and how much is being pushed out with each beat. Add those two things together and you get your resting output. However, when we're not at rest, when we're living our everyday lives, you're changing cardiac output based on demand. And you're doing that with those two features. You're increasing heart rate, which requires normal filling of an HCM heart, which is more difficult at higher heart rates, again, why early diastole is so important. But you're also increasing stroke volume, so that the blood being delivered each beat goes up. This is particularly important in early exercise. While the first metric, heart rate is going to be dependent on diastolic reserve on how well your heart can fill at faster heart rates, the second component, augmenting your stroke volume is of equal importance, if not greater importance and early exercise. And this component requires that same concept of myosin recruitability. If your myosin heads are crammed into a permanently off state, when you ask your stroke volume to pick up, it won't. But if your myosin heads are closer to actin and available for business, when it's time for you to augment cardiac output, stroke volume will go up, and we think that will be meaningful for patients. You'll see a pictorial representation of that on this slide. With a cardiac myosin inhibitor, hardly starts to go up, but stroke volume remains relatively stagnant, a finding that you can pull out of cardiopulmonary exercise data in various HCM CMI trials. Compare and contrast that with the left-hand panel, where you see Well, I need more cardiac output, let's use more myosin heads let's augment ejection fraction. We think that this is going to be differentiating, particularly in the nonobstructive space. where you're not also benefiting from alleviation of obstruction. Myosin availability is normal and natural. And we think that the regulatory light chain is a way to fine-tune this system. We have preclinical data that does support this supposition. I'll start with the normal wrap on the left. This rate was given a normal dose of the cardiac myosin inhibitor and a super physiologic 10 to 15 fold higher than what we would use clinically dose of 7500 in order to induce systolic dysfunction at baseline. We then took these 3 different groups and gave them dobutamine, which is a compound that will speed up heart rate and increase ejection fraction or tried to a state that really mimics exercise. What you see in red is that the CMI RAD did see some increases with dobutamine, likely driven predominantly through heart rate changes. But your EDG-7500 sees a much larger augmentation of stroke volume, where you're starting to approach what the normal rat control did with respect to fractional shortening changes. On the right, we have a different model. This is a non-obstructive pig, NYH-7403q pig that was treated chronically with 7500. What you'll see first is that the blue triangles moved closer in resting stroke volume to the wild-type control. So you already started to resemble more of a normal or noneffective pig. When you gave dobutamine, again, you see the same result. There's myosin availability and recruitability with this compound. So when you ask the heart to do more, it does. Ejection fraction picks up, and stroke volume increases. Again, we think this is going to be differentiating in a Phase III trial and in everyday life with this compound. You can see that reflected with our KCCQ, albeit this was nonplus-controlled Part D data, where you see really robust improvements in KCCQ. And on the right, I want to draw your attention to the shape of that nonobstructive curve. It continues to uptrend through week 12, a result we're very excited to tell you about in the future. So let's take a step back and revisit those 4 questions. How exactly does this thing get rid of the great. Well, by changing lever arm stiffness, you decrease early systolic pressure, you prevent obstruction from happening, from initiating and gradients go down. Can you do it without changing the ES well by moving myosin closer to Acton by speeding up force development in mid- to late systole. You see ejection fraction, global longitudinal strain and global circumferential strain all remain static throughout treatment with 7500. You can alleviate Cadia without reducing net systolic function. How do you improve diastolic dysfunction? While rather than being diastolically beneficial by happenstance by just turning off myosin heads directly linking your mechanism of diastolic benefit to systolic liability, we directly at low doses potently speed up, Actimize and disengagement, creating a such ventricle in early diastole that is going to be meaningful, not just at rest, but when you ask the heart rate to pick up an exercise, as seen with our BNPs and KCCQs. Lastly, is maintaining myosin availability important. What are you going to do for functional status. Well, we think it is and we think that the ability to increase stroke volume when you exercise is going to be pivotal for how our patients feel and how they ultimately perform in a cardiopulmonary exercise stress test in Phase III. We -- with that, I'm going to hand it back to Alan to bring us on.

So I'm just -- this is just a little bit of fun here, a little history of how people have described factors that control the heart. So about 130 years ago back in 1897, and academic in Germany, Theodor Wilhelm Engelmann, described 4 physiological properties that define cardiac function. You can use these in the picture, if you like, they describe contractility, heart rate, conduction velocity and excitability. What you'll notice, though, is that relaxation was not defined by Engelmann at that point. And that was largely because for an 85-year period, people assume that relaxation was a passive process that could not be intervened. Now in 1982, a group led by Drik Brusa in the Netherlands change that thinking about relaxation and lucitropy through examination of adrenergic agonist and their effect on accelerating relaxation. So an 85-year gap between defining most of the functions of the heart and then the ability to relax and that's really through understanding that you could accelerate this process. Now another gap occurs as pharmacologists try and intervene in this relaxation process. In many ways, they are directly identifying and optimizing positive loose trucks has been the holy grail of pharmaceutical research for the last 40 years or so. So between 2015 and 2025, the community made some breakthroughs there. We have Sergagene therapies. We have pharmaceutical companies trying to identify direct activators of Circa. This is the channel that sucks calcium away from the muscle and should accelerate calcium removal and relaxation. The CMIs, of course, have their own lusitropic effects, but these are coupled to systolic deficits and all of the heart failure syndromes that you can get with large concentrations of those compounds. We believe we're the first people who have actually nailed this 1 down to directly cause lusitropic. It's been a long journey. If you look at that time line, but we believe in both 7500 and 15400, you have compounds that selectively and directly accelerate relaxation of the heart. We think this will have great therapeutic potential. I'm going to hand this back to Kevin just to wrap it up.

Thanks, Alan. Thanks, Michael. Really great presentation. Obviously, we've made great progress in the program, and this is 1 of the milestones of talking about the mechanistic aspects of 7500. Just to give you some milestones for the rest of the year and 2027, we expect to have regulatory feedback in the first half of 26 based on both the trial design and all aspects of moving 7500 into Phase III. I think this is quite important because we plan on having a Phase III profile where we would not use echo monitoring to get people to their target dose. And so that would be an important milestone for the company to create an ease of use for 7500 among the community cardiologists. 7500 Phase III initiation is probably by the end of the year. We're already have zeros in place and are talking to sites as we speak. And also, we just saw on clintrials.gov, that we've initiated the trial of 15400 in HFPEF and that is open for screening today. Milestones for '27 are we plan to provide 48-week data in the open-label extension of the serato the 50 so-on patients that we produced back in June. We plan on having Phase II data probably in the second half of '27, a little bit longer than we thought because the KOLs have essentially said, you need to have a large enough data set to be really able to interpret the data, and this is a placebo-controlled study. So I think we've moved that to more real second half '27. And then we've made great progress on 003. We've chosen a therapeutic indication of heart failure. And this molecule will be showing in future scientific meetings, some of the unique profile of this molecule. So with that, I think we can take some questions, and I look forward to hearing your thoughts.

Great. Thank you to our speakers. [Operator Instructions] So our first question comes from Joe Schwartz at Leerink.

Thanks for hosting this enlightening event, and congrats on the publication. My first question is a regulatory one. I think the FDA has wanted to go through all of the latest Phase II data to support dosing without routine echo monitoring. So does today's mechanistic package and the publication contribute to that conversation or do you think the agency only gives way to clinical exposure response data here?

I think that -- they've recognized that this is a differential mechanism. I should point to 15400 that we have an ejection fraction cut-off of which is typically normal for the typical HFpEF population. So I think they've looked at the 7,500 data and the 15400 data and with an understanding if it's a novel mechanism and perhaps have drawn the conclusion that we will not see excursions below 50% or at least outside of the range of normal variability of the ejection fraction measures. But I think that remains to be, I think, validated through the agency providing us feedback on the Phase III clinical trial design and moving forward. So stay tuned. I think it's important, but I think we have, at least from our initial feedback in January, February that our trial design seems reasonable. And I think they have all the data now and they'll make a decision.

Okay. Great. And then if I could just ask a follow-up on the isoform selectivity does 7,500 engage the atrial light chain isoform as well as the ventricular one. I'm just asking because of the atrial fibrillation questions some have raised and whether there's any direct atrial pharmacology at all or whether any AF is just background risk.

Thanks for the question, Joe, is an obvious one. just the same as the myosin inhibitors, 7500 has biochemical activity against both atrial and ventricular forms. And in many ways, the pharmacology that we described for the ventricle is kind of slowing speeding type of thing is intact in the atria. As such, there's no direct correlation between those effects and atrial fibrillation. We see it as important in HCM that you both changes at the heart to relieve stress. So that was our goal going in. And that looks like what we've got.

Just to be clear, though, the CMIs had exactly the same effect on both compartments with the Atrium and the metric.

Our next question comes from Yasmeen Rahimi, Piper Sandler.

Thank you for the excellent presentation, and we have been asking you many, many questions, and it was really great to put into context the and combine it to the clinical product profile that you're seeing. I guess the question that I have for you is you have, before this publication became available, started to share this mechanism with key opinion leaders. Maybe you could talk about the sentiment and the perception that they have versus the differentiation that they see versus the CMIs. And how important is this publication for enrollment of your Phase III and, of course, strategic discussions.

I'll take the second question is that we do think that it's important to provide the mechanism because I think it just provides context for recruitment. But I'll ask the question of Michael how the key opinion leaders have reacted to the mechanism and their understanding of the differentiation from the CMIs.

Yes. Thank you for the question. KOLs are just very excited about this because the two major problems right now or unmet needs are, one, how can we do this without laborious echo monitoring, which means how do we remove systolic liability from the equation? And two, can we provide diastolic benefit to people that don't have super normal ejection fractions of 70 or higher to begin with. And as feel like we address both of those issues head on, and so they're quite excited.

I should point out that Sharad proposed this effect on early statistically back in the mid-2000s. And this is actually something people learned in textbooks, but we're never able to demonstrate this either preclinically or clinically. And that actually -- I think we have validated those original hypotheses of the origin of the Cadia in HCM. That paper is in a footnote on one of the slides.

Our next question comes from Laura Chico at Wedbush.

I think this might be directed towards Alan. I guess -- and apologies if I missed this. Have you run 7,500 head-to-head versus CK-586 and some of these assays, ATPAs, motility, fiber assays. I guess I'm just trying to understand in the context of Slide 16, a where there's that continuum, where do 586 and 7,500 sits relative to each other. And I guess just one quick clarification. Kevin, on the base case assumption for Phase III, it sounds like not using it to get to a target dose, but what would be the expectation on Echo scheduling.

My bet first? Or do you want to do your bet.

Do your, I'll catch up.

Laura, it's a great question. What does 586 look like. And of course, we take that compound run it through the full ringer to see how it compares. What I would say is that these things have complicated profiles, but in many ways, 586 exhibit profiles on that spectrum towards the CMI end. And you can see that in action in the RA data that Cytokinetics have published. And indeed, in the clinical data showing LVEF PK/PD relationships pretty much indistinguishable from Aficounter. It's clearly possible to make RLC inhibitors that probably slow that early contraction so much that you start eating into LVEF. And I think that's probably what's going on with 586.

Yes. I think GCA people will be very clear where actually it's probably as closely related as you get the Inantimer of 7500 and 7499 has a very, very distinct profile. And we wanted to point that out is that this becomes based on our mechanism prothesis in our preclinical brining a choice, a choice of a molecule that does not affect isle and is diastolic bias versus the choice of a molecule that has a systolic liability if you believe that you need to have systolic decreases to have efficacy. We obviously had a different belief when we chose 7500 that we could drive benefit for the patient without systolic liability, and that was the hypothesis we made when we selected 7500. Now in regards to what we have provided to the agency in regards to the Phase III protocol and essentially echo monitoring. We've provided a baseline echo, a 12-week echo that is not read by the PI. It's taken by the PI, sent to a core lab and stored and a 24-week echo at the end of study. And we're using tolerability essentially to get patients to a target dose will help tell you about the target dose we've chosen later on. But essentially, an echo in the beginning and an echo at the end and no utilization of a PI read echo throughout the dose escalation process.

Our next question comes from Tessa Romero at JPMorgan.

Exciting day for Edgewise, and thank you for all this detail. Kevin, Alan, Michael, if we think about the RLC as a target, are there any liabilities based on your understanding of how the drug works in the heart and its impact on the cardiac cycle. And what specifically underpins your confidence that in larger studies with the molecule that you will continue to be safe and well tolerated and as you think about 7500 as a chronic medicine or HCM.

I can start liabilities of RLC, I mean if you look in the liters if you mutate RLC, much the same as if you take any sarcomeric component, you will get. So I guess that's a liability if you did that. Now of course, what we're doing here isn't doing that. We're just modulating the RLC in the small molecule. So as such, we're not aware of any hypothetical problems that could occur as a result of modulating RLC. It's a pretty subtle effect that we're looking at. that hopefully, you got a feel for in the data that we presented.

Yes, I think if you could go into the literature, you'll find knockouts or even human mutations. And all of that information goes back to -- it can be recapitulated in rat and mouse models. And we've, of course, treated in our tox studies, we've taken multiple measures either via ECG or structural changes. And we do not see any of these effects in the rat models on long-term dosing. And remember, we did the six-month health -- R4OQ pig model of nonobstructive ATM. And we actually see, I would say, a benefit on the train that model at 6 months where we see a decrease in real size. And one of the comments in one of the papers is that -- you see hypo phosphorylation of decreases in phosphorylation of the RLC. In our PAG studies, we see either neutral or increased levels of phospholation after 6 months of dosing. So benefits to the Atria. So I think that all fits together that there's really no tie to the mechanism. I think, again, like all things in HCM, whether you decrease the contraction or increase the contraction, hyperdynamic change in the heart leads to increased rates of part of atrial fibrillation. And so I think it's kind of a -- it's -- I think the whole thing is not type of the mechanism in any way.

Our next question comes from Leo Timashev from RBC Capital Markets.

Ask your question. really interesting mechanism, and I appreciate the clear presentation here. I wanted to ask kind of a multi-parter on calcium. If you can help just some tangles of what we're seeing there. I guess, given the impact on the sensitivity of the drug's effect to calcium, do HCM patients have normal calcium levels in the heart -- and then calcium blockers are also fairly common drugs in the cardiac space. I guess, do you anticipate any impacts of calcium blockers on a 7500 would work? And then if I can squeeze in 1 last one. Just given the impact of 7500 moving myosin closer to actin, can that also increase heart rate.

I will address the first 2, and then I'll hand it over to Michael to talk about kind of physiological stuff. So calcium and HCM is -- well, the dysfunction in HCM, of course, is sarcomeric largely, right? And calcium will make attempts to kind of overcome the dysfunction there. So he may get changes in calcium transient, but still, we will have systolic and diastolic levels of calcium. I don't anticipate the compound if particularly in an HCM patient in regards to calcium concentrations, but diastole you still have very low levels of calcium and assisted you have these maximal levels of calcium. Calcium channel blockers don't affect calcium transients per se. They affect the influx of calcium, but those calcium stores that I'm talking about coming from the circa, the big bag of calcium within the muscle, and that largely is unaffected by those types of therapies. So our calcium effects that we described at this kind of low to high housing should be tacked irrespective of whether you're in a casting channel blocker or HCM. And I will pass the other question over.

Maybe just preclinically. We've examined exhaustively in myocytes and other systems that we've not seen any effect on a calcium channel...

Yes, that's a good one.